Author response:

Thank you for the positive and constructive feedback on our manuscript. We appreciate you highlighting the importance of our work advancing our understanding of the molecular etiology of acquired immunodeficiency syndrome (AIDS). To extend and further substantiate the observation that the CARD8 inflammasome is activated in response to viral protease during HIV-1 cell-to-cell transmission, we are in the process of completing additional experiments that are responsive to reviewer feedback, including:

• Primary CD4+ T cell to monocyte-derived macrophage (MDM) transmission:  We have now repeated the cell-to-cell experiments with HIV-1 transfer from primary CD4+ T cells to primary monocyte-derived macrophages, and our findings are consistent with CARD8-dependent IL-1β release from HIV-1-infected macrophages in this more physiologic context. We are in the process of repeating these experiments with additional donors and will add these results to the revised manuscript.

• Heterogeneity amongst blood donors: We have now repeated the cell-to-cell transfer and CARD8 knockout in MDMs with additional donors. While we continue to observe heterogeneity amongst donors, the key observation that CARD8 is require for inflammasome responses to HIV-1 infection is consistent. We note that some donors, including the one individual reported in the first submission, have markedly diminished CARD8 activity (to both HIV-1 and VbP).

• Time course experiments: We did conduct a time course experiment when initially establishing these assays. We have now repeated these experiments with additional timepoints and in the presence or absence of the RT inhibitor nevirapine. The results of these experiments will be included in the revised manuscript.

• The role of Gasdermin D: We are mostly interested in the release of IL-1β from the infected macrophages due to its potential contribution to myeloid-driven inflammation in PLWH. To date, there is no evidence that any other pore-forming protein other than GSDMD can initiate IL-1β release (and pyroptosis) downstream of CARD8. Nonetheless, we will attempt this experiment with the Gasdermin D inhibitor, disulfiram. 

We believe these and other experiments will further support the importance of the CARD8 inflammasome in myeloid-driven inflammation in PLWH and look forward to submitting the revision.

Author response:

Public Reviews: 

Reviewer #1 (Public review): 

Summary: 

The authors test whether the archerfish can modulate the fast response to a falling target.

We have not tested whether archerfish can 'modulate the fast response'. We quantitatively test specific hypotheses on the rules used by the fish. For this the accuracy of the decisions is analyzed with respect to specific points that can be calculated precisely in each experiment. The ill-defined term 'modulate' does in no way capture what is done here. This assessment might explain the question, raised by the reviewer, of 'what is the difference of this study and Reinel, 2016' (i.e. Reinel and Schuster, 2016). In that study, all objects were strictly falling ballistically, and latency and accuracy of the turn decisions were determined when the initial motion was not only horizontal but had an additional vertical component of speed. The question of that study was if the need to account to an additional variable (vertical speed) in the decision would affect its latency or accuracy. The study showed that also then archerfish rapidly turn to the later impact point. It also showed that accuracy and latency (defined in this study exactly as in the present study) were not changed by the added degree of freedom. This is a completely different question and by its very nature does not leave the realm of ballistics.

By manipulating the trajectory of the target, they claim

that the fish can modulate the fast response.

While it is clear from the result that the fish can modulate the fast response, the experimental support for the argument that the fish can do it for a reflex-like behavior is inadequate. 

This is disturbing: The manuscript is full of data that directly report response latency (a parameter that's critical in all experiments) and there are even graphical displays of the distribution of latency (Figs. 2, 5). How fast the responses are, can also already be seen in the first video. Most importantly, the nature of the 40 ms limit has been discovered and has been reported by our group in 2008 (Schlegel and Schuster, 2008, Fig. 4). For easy reference, we attach Schlegel and Schuster, 2008 with the relevant passages marked in yellow. But later studies also using high speed video (ie. typically 500 fps) and simultaneously evaluating accuracy and kinematics (in the same ways as used here!) to address various questions repeatedly report and even graphically represent minimum latencies of 40 ms, e.g. Krupczynski and Schuster, 2013 (e.g. Fig. 2); Reinel and Schuster, 2014; Reinel and Schuster, 2016;  Reinel and Schuster, 2018a, b (e.g. see Fig. 7 in the first part) and report how latency is increased as urgency is decreased (if the fish are too close or time of falling is increased), as temperature is decreased or as viewing conditions and their homogeneity across the tank change. Moreover, even a field study is available (Rischawy, Blum and Schuster, 2015) that shows why the speed is needed. This is because of massive competition with at least some of the competitor fish also be able to turn to the later impact point. So, speed is an absolute necessity if competitors are around. Interestingly, when the fish are isolated, latency goes up and eventually the fish will no longer respond with C-starts (Schlegel and Schuster, 2008).

Another aspect: considering the introduction it would not even have mattered if not 40 ms but instead 150 ms were the time needed for an accurate start (which is not the case). That would still be faster than an Olympic sprinter responds to a gun shot. Moreoever, please also note that we carefully talk of reflex-speed not of a reflex-behavior (which is as easy to verify as any other if the false statements made).

Strengths: 

Overall, the question that the authors raised in the manuscript is interesting. 

Given the statement of no difference between the present study and Reinel and Schuster, 2016, it is not clear what this assessment refers to.

Weaknesses: 

(1) The argument that the fish can modulate reflex-like behavior relies on the claim that the archerfish makes the decision in 40 ms. There is little support for the 40 ms reaction time.

The 'little support' is a paper in Science in which this important aspect is directly analyzed (Fig. 4 of that paper) and that has been praised by folks like Yadin Dudai (e.g . in Faculty 1000). The support is also data on latency as presented in the present paper. Furthermore, additional publications are available on the reaction time (see above).

The reaction time for the same behavior in Schlegel 2008, is 60-70 ms, and in Tsvilling 2012 about 75 ms, if we take the half height of the maximum as the estimated reaction time in both cases. If we take the peak (or average) of the distribution as an estimation of reaction time, the reaction time is even longer. This number is critical for the analysis the authors perform since if the reaction time is longer, maybe this is not a reflex as claimed.

See above.

In addition, mentioning the 40 ms in the abstract is overselling the result.

See above.

Just for completeness: Considering a very interesting point raised by reviewer 2 we add an additional panel to further emphasize the exciting point that accuracy and latency are unrelated in the start decisions. That point was already made in Fig.4 of the paper in Science but can be directly addressed.  

The title is also not supported by the results. 

No: the title is clearly supported by the results that are reported in the paper.

(2) A critical technical issue of the stimulus delivery is not clear.

The stimulus delivery is described in detail. Most importantly we emphasize (even mentioning frame rate) that all VR setups require experimental confirmation that they work for the species and for the behavior at hand. Ideally, they should elicit the same behavior (in all aspects) as a real stimulus does that the VR approach intends to mimic. Whether VR works in a given animal and for the behavior at hand in that animal cannot be known or postulated a priori. It must be shown in direct critical experiments. Such experiments and the need to perform them are described in detail in Figure 2 and in the text that is associated with that figure.

The frame rate is 120 FPS and the target horizontal speed can be up to 1.775 m/s. This produces a target jumping on the screen 15 mm in each frame. This is not a continuous motion. Thus, the similarity between the natural system where the target experiences ballistic trajectory and the experiment here is not clear. Ideally, another type of stimulus delivery system is needed for a project of this kind that requires fast-moving targets (e.g. Reiser, J. Neurosci.Meth. 2008).

See above. It is quite funny that one of the authors of the present study had been involved in developing a setup with a complete panorama of 6000 LEDs (Strauss, Schuster and Götz, 1997; and appropriately cited in Reiser) that has been the basis for Reiser. This panorama was also used to successfully implement VR in freely walking Drosophila (Schuster et al., Curr. Biol., 2002). However, an LED based approach was abandoned because of insufficient spatial resolution (that, in archerfish, is very different from that of Drosophila).

But the crucial point really is this: Just looking at Figure 2 shows that our approach could not have worked better in any way - it provided the input needed to cause turn decisions that are in all aspects just as those with real objects. Achieving this was not at all trivial and required enormous effort and many failed attempts. But it allows addressing our questions for the first time after 20 years of studying these interesting decisions.

In addition, the screen is rectangular and not circular, so in some directions, the target vanishes earlier than others. It must produce a bias in the fish response but there is no analysis of this type. 

Why 'must' it produce a bias? Is it not conceivable that you can only use a circular part of the screen? Briefly, and as could have been checked by quickly looking into the methods section, this is what we did. But still, why would it have mattered in our strictly randomized design? It could have mattered only in a completely silly way of running the experiments in which exclusively long trajectories are shown in one condition and exclusively short ones in another.

(3) The results here rely on the ability to measure the error of response in the case of a virtual experiment. It is not clear how this is done since the virtual target does not fall.

Well, of course it does not fall!!! That is the whole point that enables the study, and this is explained in connection with the glass plate experiment of Fig. 1 and quite some text is devoted to say that this is the starting point for the present analysis. The ballistic impact point is calculated (just as explained in our very first paper on the start decisions, Rossel, Corlija and Schuster, 2002) from the initial speed and height of the target, using simple high-school physics and the justification for that is also in that paper. This has been done already more than 20 years ago. How else could that paper have arrived at the conclusion that the fish turned to the virtual impact point even though nothing is falling? We also describe this for the readers of the present study, illustrate how accuracy is determined in Figures, in all videos and in an additional Supplementary Figure. Consulting the paper reveals that orientation of the fish is determined immediately at the end of stage 2 of its C-start and the error directly reports how close continuing in that direction would lead the fish to the (real or virtual) impact point. This measure has also been used since the first paper in 2002 in our lab and it is very useful because it provides an invariant measure that allows pooling all the different conditions (orientation and position of responding fish as well as direction, speed and height of target).

How do the authors validate that the fish indeed perceives the virtual target as the falling target?

See above. The fish produce C-starts (whose kinematics are analyzed and reported in Figures), whose latency is measured (from onset of target motion to onset of C-start) and whose accuracy in aligning them to the calculated virtual impact point is measured (see above). Additionally, the errors are also analyzed to other points of interest, for instance landmarks, the ballistic landing point in the re-trained fish or points calculated on the basis of specific hypotheses in the generalization experiments.

Since the deflection is at a later stage of the virtual trajectory, it is not clear what is the actual physics that governs the world of the experiment.

As explained in the text what we need is substituting the ballistic connection with another fixed relation between initial target motion and the landing point. This other relation needs to produce a large error in the aims when they remain based on the ballistic virtual landing point. It is directly shown in the key experiments that the fish need not see the deflection but can respond appropriately to the initial motion after training (Figs. 3, 5 and corresponding paragraphs in the text as well as additional movies). Please also note that after training the decision is based on the initial movement. This is shown in the interspersed experiments in which nothing than the initial (pre-deflection) movement was shown.

Overall, the experimental setup is not well designed. 

It is obviously designed well enough to mimic the natural situation in every aspect needed (see Fig. 2) and well enough to answer the questions we have asked.

Reviewer #2 (Public review): 

Summary: 

This manuscript studies prey capture by archer fish, which observe the initial values of motion of aerial prey they made fall by spitting on them, and then rapidly turn to reach the ballistic landing point on the water surface. The question raised by the article is whether this incredibly fast decision-making process is hardwired and thus unmodifiable or can be adjusted by experience to follow a new rule, namely that the landing point is deflected from a certain amount of the expected ballistic landing point. The results show that the fish learn the new rule and use it afterward in a variety of novel situations that include height, side, and speed of the prey, and which preserve the speed of the fish's decision. Moreover, a remarkable finding presented in this work is the fact that fish that have learned to use the new rule can relearn to use the ballistic landing point for an object based on its shape (a triangle) while keeping simultaneously the 'deflected rule' for an object differing in shape (a disc); in other words, fish can master simultaneously two decision-making rules based on the different shape of objects. 

Strengths: 

The manuscript relies on a sophisticated and clever experimental design that allows changing the apparent landing point of a virtual prey using a virtual reality system. Several robust controls are provided to demonstrate the reliability and usefulness of the experimental setup. 

Overall, I very much like the idea conveyed by the authors that even stimuli triggering apparently hardwired responses can be relearned in order to be associated with a different response, thus showing the impressive flexibility of circuits that are sometimes considered mediating pure reflexive responses.

Thank you so much for this precise assessment of what we have shown!

This is the case - as an additional example - of the main component of the Nasanov pheromone of bees (geraniol), which triggers immediate reflexive attraction and appetitive responses, and which can, nevertheless, be learned by bees in association with an electric shock so that bees end up exhibiting avoidance and the aversive response of sting extension to this odorant (1), which is a fully unnatural situation, and which shows that associative aversive learning is strong enough to override preprogrammed responding, thus reflecting an impressive behavioral flexibility. 

That's very interesting, thanks.

Weaknesses: 

As a general remark, there is some information that I missed and that is mandatory in the analysis of behavioral changes. 

Firstly, the variability in the performances displayed. The authors mentioned that the results reported come from 6 fish (which is a low sample size). How were the individual performances in terms of consistency? Were all fish equally good in adjusting/learning the new rule? How did errors vary according to individual identity? It seems to me that this kind of information should be available as the authors reported that individual fish could be recognized and tracked (see lines 620-635) and is essential for appreciating the flexibility of the system under study. 

Secondly, the speed of the learning process is not properly explained. Admittedly, fish learn in an impressive way the new rule and even two rules simultaneously; yet, how long did they need to achieve this? In the article, Figure 2 mentions that at least 6 training stages (each defined as a block of 60 evaluated turn decisions, which actually shows that the standard term 'Training Block' would be more appropriate) were required for the fish to learn the 'deflected rule'. While this means 360 trials (turning starts), I was left with the question of how long this process lasted. How many hours, days, and weeks were needed for the fish to learn? And as mentioned above, were all fish equally fast in learning? I would appreciate explaining this very important point because learning dynamics is relevant to understanding the flexibility of the system. 

First, it is very important to keep the question in mind that we wanted to clarify: Does the system have the potential to re-tune the decisions to other non-ballistic relations between the input variables and the output? This would have been established if one fish was found capable of doing that. However, we do have sufficient evidence to say that all six fish learned the new law and that at least one (actually four) individual was capable of simultaneously handling the two laws. We will explain this much better (hopefully) in our revised version. We also have to stress that not all archerfish might actually be able to do this and that not all archerfish might learn in the same way, at the same speed, or using the same strategies. These questions are extremely interesting and we therefore definitely will include all evidence that we have. If some individuals are better than others in quickly adjusting, then even observational learning could become a part of the story. However, we needed to make and document the first steps. Understanding these is essential and apparently is difficult enough.

Reference: 

(1) Roussel, E., Padie, S. & Giurfa, M. Aversive learning overcomes appetitive innate responding in honeybees. Anim Cogn 15, 135-141, doi:10.1007/s10071-011-0426-1 (2012). 

Thanks for this reference!

Author response:

Public Reviews:

Reviewer #1 (Public review):

This is an interesting manuscript tackling the issue of whether subcircuits of the cerebellum are differentially involved in processes of motor performance, learning, or learning consolidation. The authors focus on cerebellar outputs to the ventrolateral thalamus (VL) and to the centrolateral thalamus (CL), since these thalamic nuclei project to the motor cortex and striatum respectively, and thus might be expected to participate in diverse components of motor control and learning. In mice challenged with an accelerating rotarod, the investigators reduce cerebellar output either broadly, or in projection-specific populations, with CNO targeting DREADD-expressing neurons. They first establish that there are not major control deficits with the treatment regime, finding no differences in basic locomotor behavior, grid test, and fixed-speed rotarod. This is interpreted to allow them to differentiate control from learning, and their inter-relationships. These manipulations are coupled with chronic electrophysiological recordings targeted to the cerebellar nuclei (CN) to control for the efficacy of the CNO manipulation. I found the manuscript intriguing, offering much food for thought, and am confident that it will influence further work on motor learning consolidation. The issue of motor consolidation supported by the cerebellum is timely and interesting, and the claims are novel. There are some limitations to the data presentation and claims, highlighted below, which, if amended, would improve the manuscript.

We thank the reviewer for the positive comments and insightful critics.

(1.1) Statistical analyses: There is too little information provided about how the Deming regressions, mean points, slopes, and intercepts were compared across conditions. This is important since in the heart of the study when the effects of inactivating CL- vs VL- projecting neurons are being compared to control performance, these statistical methods become paramount. Details of these comparisons and their assumptions should be added to the Methods section. As it stands I barely see information about these tests, and only in the figure legends. I would also like the authors to describe whether there is a criterion for significance in a given correlation to be then compared to another. If I have a weak correlation for a regression model that is non-significant, I would not want to 'compare' that regression to another one since it is already a weak model. The authors should comment on the inclusion criteria for using statistics on regression models.

Currently the Methods indeed explain that groups are compared by testing differences of distributions of residuals of treatment and control groups around the Deming regression of the control groups: “To test if treatments altered the relationship between initial performance vs learning or daily vs overnight learning, we compared the distribution of signed distance to the control Deming regression line between groups.” But this shall indeed be explained in more details.

The performance on a given day depends on a cumulative process, so that the average measure of performance is not fully informative on what is learned or what is changed by a treatment (this is further explained in the text p9-10).The challenge is to deal with the multivariate relationships where initial performance, daily learning, and consolidated learning are interdependent. While in control groups these quantities show linear relationships, this is far less the case in treatment groups; this may indeed be due to the variability of the effect of the treatment (efficacy of viral injections) which adds up to the intrinsic variability in the absence of treatment.

Our choice to see if there is a shift in these relationships following treatments, is to see to which extent treatment points in bivariate comparisons (initial perf x daily learning, daily learning x consolidated learning) are evenly distributed around the control group regression line. We take the presence of a significant difference in the distribution of residuals between the control and treatment group as an indication that the process represented in group is disrupted by the treatment: e.g. if the residuals of the treatment group are lower than those of the control group in the initial performance * daily learning comparison, it indicates that learning is slower (or larger). If the residuals of the treatment group are lower than those of the control group in the daily learning * consolidated learning comparison, it indicates that consolidation is lower. This shall be clarified in a revised version.

(1.2a) The introduction makes the claim that the cerebellar feedback to the forebrain and cortex are functionally segregated. I interpreted this to mean that the cerebellar output neurons are known to project to either VL or CL exclusively (i.e. they do not collateralize). I was unaware of this knowledge and could find no support for the claim in the references provided (Proville 2014; Hintzer 2018; Bosan 2013). Either I am confused as to the authors' meaning or the claim is inaccurate. This point is broader however than some confusion about citation.

The references are not cited in the context of collaterals: “They [basal ganglia and cerebellum] send projections back to the cortex via anatomically and functionally segregated channels, which are relayed by predominantly non-overlapping thalamic regions (Bostan, Dum et al. 2013, Proville, Spolidoro et al. 2014, Hintzen, Pelzer et al. 2018). ” Indeed, the thalamic compartments targeted by the basal ganglia and cerebellum are distinct, and in the Proville 2014, we showed some functional segregation of the cerebello-cortical projections (whisker vs orofacial ascending projections). We do not claim that there is a full segregation of the two pathways, there is indeed some known degree of collateralization (see below).

(1.2b) The study assumes that the CN-CL population and CN-VL population are distinct cells, but to my knowledge, this has not been established. It is difficult to make sense of the data if they are entirely the same populations, unless projection topography differs, but in any event, it is critical to clarify this point: are these different cell types from the nuclei?; how has that been rigorously established?; is there overlap? No overlap? Etc. Results should be interpreted in light of the level of this knowledge of the anatomy in the mouse or rat.

Actually, the study does not assume that CL-projecting and VAL-projecting neurons are entirely separate populations (actually it is known that there is an overlap), but states that inhibition of neurons following retrograde infections from the CL and VAL do not produce identical results.

There is indeed a paragraph devoted to the discussion of this point (middle paragraph p20). “Interestingly, both Dentate and Interposed nuclei contain some neurons with collaterals in both VAL and CL thalamic structures (Aumann and Horne 1996, Sakayori, Kato et al. 2019), suggesting that the effect on learning could be mediated by a combined action on the learning process in the striatum (via the CL thalamus) and in the cortex (via the VAL thalamus). However, consistent with (Sakayori, Kato et al. 2019), we found that the manipulations of cerebellar neurons retrogradely targeted either from the CL or from the VAL produced different effects in the task. This indicates that either the distinct functional roles of VAL-projecting of CL-projecting neurons reported in our study is carried by a subset of pathway-specific neurons without collaterals, or that our retrograde infections in VAL and CL preferentially targeted different cerebello-thalamic populations even if these populations had axon terminals in both thalamic regions.”. In other words, we actually know from the literature that there is a degree of collateralization (CN neurons projecting to both VAL and CL, see refs cited above), but as the reviewer says, it does not seem logically possible that the exact same population would have different effects, which are very distinct during the first learning days. The only possible explanation is the CN-CL and CN-VAL retrograde infections recruit somewhat different populations of neurons. This could be due to differences in density of collaterals in CL and VAL of neurons with collaterals in both regions, or presence of CL-projecting neurons without collaterals in VAL, and VAL-projecting neurons without collaterals in CL in addition to the (established) population of neurons with collaterals in both regions. The lesional approach of CN-thalamus neurons in Sakayori et al. 2019 also observed separate effects for CL and VL injections consistent with the differential recruitment of CN populations by retrograde infections.

This should be improved in a revised version of the manuscript.

(1.3) It is commendable that the authors perform electrophysiology to validate DREADD/CNO. So many investigators don't bother and I really appreciate these data. Would the authors please show the 'wash' in Figure 1a, so that we can see the recovery of the spiking hash after CNO is cleared from the system? This would provide confidence that the signal is not disappearing for reasons of electrode instability or tissue damage/ other.

We do not have the wash data on the same day, but there is no significant change in the baseline firing rate across recording days.

(1.4) I don't think that the "Learning" and "Maintenance" terminology is very helpful and in fact may sow confusion. I would recommend that the authors use a day range " Days 1-3 vs 4-7" or similar, to refer to these epochs. The terminology chosen begs for careful validation, definitions, etc, and seems like it is unlikely uniform across all animals, thus it seems more appropriate to just report it straight, defining the epochs by day. Such original terminology could still be used in the Discussion, with appropriate caveats.

This shall be indeed corrected in a revised version.

(1.5) Minor, but, on the top of page 14 in the Results, the text states, "Suggesting the presence of a 'critical period' in the consolidation of the task". I think this is a non-standard use of 'critical period' and should be removed. If kept, the authors must define what they mean specifically and provide sufficient additional analyses to support the idea. As it stands, the point will sow confusion.

This shall be indeed corrected in a revised version

Reviewer #2 (Public review):

Summary:

This study examines the contribution of cerebello-thalamic pathways to motor skill learning and consolidation in an accelerating rotarod task. The authors use chemogenetic silencing to manipulate the activity of cerebellar nuclei neurons projecting to two thalamic subregions that target the motor cortex and striatum. By silencing these pathways during different phases of task acquisition (during the task vs after the task), the authors report valuable findings of the involvement of these cerebellar pathways in learning and consolidation.

Strengths:

The experiments are well-executed. The authors perform multiple controls and careful analysis to solidly rule out any gross motor deficits caused by their cerebellar nuclei manipulation. The finding that cerebellar projections to the thalamus are required for learning and execution of the accelerating rotarod task adds to a growing body of literature on the interactions between the cerebellum, motor cortex, and basal ganglia during motor learning. The finding that silencing the cerebellar nuclei after a task impairs the consolidation of the learned skill is interesting.

We thank the reviewer for the positive comments and insightful critics below.

Weaknesses:

(2.1) While the controls for a lack of gross motor deficit are solid, the data seem to show some motor execution deficit when cerebellar nuclei are silenced during task performance. This deficit could potentially impact learning when cerebellar nuclei are silenced during task acquisition.

One of our key controls are the tests of the treatment on fixed speed rotarod, which provides the closest conditions to the ones found in the accelerating rotarod (the main difference between the protocols being the slow steady acceleration of rod rotation [+0.12 rpm per s]- in the accelerating version).

In the CN experiments, we found clear deficits in learning and consolidation while there was no effect on the fixed speed rotarod (performance of the DREAD-CNO are even slightly better than some control groups), consistent with a separation of the effect on learning/consolidation from those on locomotion on a rotarod. However, small but measurable deficits are found at the highest speed in the fixed speed rotarod in the CN-VAL group; there was no significant effect in the CN-CL group, while the CN-CL actually shows lower performances from the second day of learning; we believe this supports our claim that the CN-CL inhibition impacted more the learning process than the motor coordination. In contrast the CN-VAL group only showed significantly lower performance on day 4 of the accelerating rotarod consistent with intact learning abilities. Of note, under CNO, CN-VAL mice could stay for more than a minute and half at 20rpm, while on average they fell from the accelerating rotarod as soon as the rotarod reached the speed of ~19rpm (130s).

The text currently states “The inhibition of CN-VAL neurons during the task also yielded lower levels of performance in the Maintenance stage,[[NB: day 5-7]] suggesting that these neurons contribute also to learning and retrieval of motor skills, although the mild defect in fixed speed rotarod could indicate the presence of a locomotor deficit, only visible at high speed.” Following the reviewers’ comment, we shall however revise the sentence above in the revised version of the MS to say that we cannot fully disambiguate the execution / learning-retrieval effect at high speed for these mice.

(2.2a) Separately, I find the support for two separate cerebello-thalamic pathways incomplete. The data presented do not clearly show the two pathways are anatomically parallel.

As explained above (point 1.2a), it is already known that these pathways overlap to some degree (discussion p 20), but yet their targeting differentially affects the behavior, consistent with separate contributions. A similar finding was observed for a lesional (irreversible) approach in Sakayori et al. 2019.

(2.2b) The difference in behavioral deficits caused by manipulating these pathways also appears subtle.

While we agree that after 3-4 days of learning the difference of performance between the groups becomes elusive, we respectfully disagree with the reviewer that in the early stages these differences are negligible and the impact of inhibition on "learning rate" (ie. amount of learning for a given daily initial performance) and consolidation (i.e. overnight retention of daily gain of performance) exhibit different profiles for the two groups (fig 3h vs 3k).

Reviewer #3 (Public review)

Summary:

Varani et al present important findings regarding the role of distinct cerebellothalamic connections in motor learning and performance. Their key findings are that:

(1) cerebellothalamic connections are important for learning motor skills

(2) cerebellar efferents specifically to the central lateral (CL) thalamus are important for short-term learning

(3) cerebellar efferents specifically to the ventral anterior lateral (VAL) complex are important for offline consolidation of learned skills, and

(4) that once a skill is acquired, cerebellothalamic connections become important for online task performance.

The authors went to great lengths to separate effects on motor performance from learning, for the most part successfully. While one could argue about some of the specifics, there is little doubt that the CN-CL and CN-VAL pathways play distinct roles in motor learning and performance. An important next step will be to dissect the downstream mechanisms by which these cerebellothalamic pathways mediate motor learning and adaptation.

Strengths:

(1) The dissociation between online learning through CN-CL and offline consolidation through CN-VAL is convincing.

(2) The ability to tease learning apart from performance using their titrated chemogenetic approach is impressive. In particular, their use of multiple motor assays to demonstrate preserved motor function and balance is an important control.

(3) The evidence supporting the main claims is convincing, with multiple replications of the findings and appropriate controls.

We thank the reviewer for the positive comments and insightful critics below.

Weaknesses:

(3.1) Despite the care the authors took to demonstrate that their chemogenetic approach does not impair online performance, there is a trend towards impaired rotarod performance at higher speeds in Supplementary Figure 4f, suggesting that there could be subtle changes in motor performance below the level of detection of their assays.

This is also discussed in point 2.1 above. In our view, the fixed speed rotarod is a control very close to the accelerating rotarod condition, with very similar requirements between the two tasks (yet unfortunately rarely tested in accelerating rotarod studies). We do not exclude the presence of motor deficits, but the main argument is that these do not suffice to explain the differences observed in the accelerating rotarod. No detectable deficit was found in the CN group while very clear deficits in learning/consolidation were observed. A mild deficit is only significant in the CN-VAL group, while the deficit is not significant in the fixed-speed rotarod for the CN-CL group which shows the strongest deficit in accelerating rotarod during the first days: e.g. on day 2, the CN-CL group is already below the control group with latencies to fall ~100s (corresponding to immediate fall at ~15rpm) while the fixed speed rotarod performances at 15s of the control and CNO-treated groups show an ability to stay more than 1 min at this speed. The text shall be improved to clarify this point.

(3.2) There is likely some overlap between CN neurons projecting to VAL and CL, somewhat limiting the specificity of their conclusions.

There is indeed published evidence for some degree of anatomical overlap, but also for some differential contribution of CN-VAL and CN-CL to the task. The answer to this point is developed in the points 1.2a 2.2a above. Although this point was exposed in the discussion (p20), the text shall be improved in a revised version of the MS to clarify our statement.

Author response:

Reviewer #1 (Public review):

Summary:

The authors successfully detected distinct mechanisms signalling prediction violations in the auditory cortex of mice. For this purpose, an auditory pure-tone local-global paradigm was presented to awake and anaesthetised mice. In awake rodents, the authors also evaluated interneuron cell types involved in responses to the interruption of the regularity imposed by local-global sequences. By performing two-photon calcium imaging and single-unit electrophysiology, the authors disentangled three phenomena underlying responses to violations of the distinct local-global regularity levels: Stimulus-specific adaptation, surprise and surprise adaptation. Both stimulus-specific adaptation and surprise-or deviant-evoked responses are observable under anaesthesia. Altogether, this work advances our understanding of distinct predictive processes computing prediction violations upon the complexity of the regularity imposed by the auditory sequence.

Strengths:

it is an elegant study beautifully executed.

Weaknesses:

No weaknesses were identified by this reviewer.

Reviewer #2 (Public review):

Summary:

Oddball responses are increases in sensory responses when a stimulus is encountered in an unexpected location in a sequence of predictable stimuli. There are two computational interpretations for these responses: stimulus-specific adaptation and prediction errors. In recent years, evidence has accumulated that a significant part of these sequence violation responses cannot be explained simply by stimulus-specific adaptation. The current work elegantly adds to this evidence by using a sequence paradigm based on two levels of sequence violations: "Local" sequence violations of repetitions of identical stimuli, and "global" sequence violations of stimulus sequence patterns. The authors demonstrate that both local and global sequence violation responses are found in L2/3 neurons of the mouse auditory cortex. Using sequences with different inter-stimulus intervals, they further demonstrate that these sequence violation responses cannot be explained by stimulus-specific adaption.

Strengths:

The work is based on a very clever use of a sequence violation paradigm (local-global paradigm) and provides convincing evidence for the interpretation that there are at least two types of sequence violation responses and that these cannot be explained by stimulus-specific adaption. Most of the conclusions are based on a large dataset, and are compelling.

Weaknesses:

The final part of the paper focuses on the responses of VIP and PV-positive interneurons. The responses of VIP interneurons appear somewhat variable and difficult to interpret (e.g. VIP neurons exhibit omission responses in the A block, but not the B block). The conclusions based on these data appear less solid.

We agree with the referee that the response modulations observed in  VIP and PV-Positive interneurons are weak and variable. This is indicated in the manuscript. Probably, larger scale recordings are necessary to ascertain fully the presence and distribution of omission responses.

Reviewer #3 (Public review):

Summary:

In their manuscript entitled "Parallel mechanisms signal a hierarchy of sequence structure violations in the auditory cortex", Jamali et al. provide evidence for cellular-level mechanisms in the auditory cortex of mice for the encoding of predictive information on different temporal and contextual scales. The study design separates more clearly than previous studies between the effects of local and global deviants and separates their respective effects on the neuronal responses clearly through the use of various contextual conditions and short and long time scales. Further, it identifies a contribution from a small set of VIP interneurons to the detection of omitted sounds, and shows the influence of isofluorane anesthesia on the neural responses.

Strengths:

(1) The study provides a rather encompassing set of experimental techniques to study the cellular level responses, using two complementary recording techniques in the same animal and similar cortical location.

(2) Comparison between awake and anesthetized states are conducted in the same animals, which allows for rather a direct comparison of populations under different conditions, thus reducing sampling variability.

(3) The set of paradigms is well developed and specifically chosen to provide appropriate and meaningful controls/comparisons, which were missing from previous studies.

(4) The addition of cell-type specific recordings is valuable and in particular in combination with the contrast of awake and anesthetized animals provides novel insights into the cellular level representation of deviant signals, such as surprise, prediction error, and general adaptation.

(5) The analysis and presentation of the data are clear and quite complete, yet remain succinct and perform insightful contrasts.

(6) The study will have an impact on multiple levels, as it introduces important variations in the paradigm and analytical contrasts that both human and animal researchers can pick up and improve their studies. The cell-type-specific results are particularly intriguing, although these would likely require replication before being completely reliable. Further, the study provides a substantial and diverse dataset that others can explore.

Weaknesses:

(1) The responses from cells recorded via Neuropixel and 2p differ qualitatively, as noted by the authors, with NP-recorded cells showing much more inhibited/reduced responses between acoustic stimulations. The authors briefly qualify these differences as potentially indicating a sampling issue, however, this matter deserves more detailed consideration in my opinion. Specifically, the authors could try to compare the different depths at which these neurons were sampled or relate the locations in the cortex to each other (as the Neuropixel recordings were collected in the same animals, a subset of the 2p recordings could be compared to the Neuropixel recordings.).

We agree with the referee that the sampling issue, which we propose as a possible explanation for the large difference between our Neuropixel and 2P imaging recordings, must be investigated more thoroughly. This is, however, largely outside of the scope of this study. We have reported the depth and location of Neuropixel recordings in Figure S2. The depth is similar for both techniques covering mostly layers 2, 3 and 4. The location spans mostly the primary auditory cortex for two photon imaging and Neuropixel recordings. One Neuropixel recording is located in the ventral secondary auditory cortex. We could not find any evidence that the response to global violations in Neuropixel data stems specifically from this particular recording. 

(2) The current study did not monitor the attentional state of the mouse in relation to the stimulus by either including a behavioral component or pupil monitoring, which could influence the neural responses to deviant stimuli and omissions.

As reported by Bekinschtein et al. 2009, the attentional state influences responses to global violation in human subjects. It is extremely difficult to precisely compare attentional states in mice and human subjects. We have performed recordings in mice that had to attend to sound to detect a white noise sound in between the sequence to obtain a reward. This did not lead to increased global violation response. However, as the sequence themselves did not predict reward in this context they may divert attention. Therefore, this result is inconclusive and not worth including in our manuscript. If the sequence predicts rewards, there is a potential confound between violation responses and reward expectations or motor preparation signals. Pupil monitoring could be an alternative which we did not investigate.

(3) Given the complexity and variety of the paradigms, conditions, and analyzed cell-types, the manuscript could profit from a more visual summary figure that provides an easy-to-access overview of what was found.

This is an excellent suggestion, although given the complexity and diversity of our observations it may be hard to fit everything in one understandable figure.

Author response:

We appreciate the insightful comments and suggestions, which will significantly improve our work. We will revise the manuscript to address the reviewer’s concerns. Here, we list some of the key aspects of those concerns and our preliminary plans to address them.

Both reviewers pointed out that we did not sufficiently justify the chosen optogenetic stimulation frequencies. We acknowledge and concur with their assessment, and will discuss it more extensively from a biological perspective (e.g., the neural firing rates in the olfactory bulb, OB, anterior olfactory nucleus, AON, and piriform cortex, Pir, under natural odor stimulation and respiration rhythm). Reviewer #1 suggested using beta values (b) rather than the area under the BOLD signal profile (AUC) to quantify the fMRI activations as they are more conventional for general linear model (GLM) analysis. We are aware of b and have used them for quantification of the amplitude of fMRI activations in our previous rodent fMRI studies1-3. However, in this study, we chose to utilize AUC as it offers a more comprehensive measure of BOLD signal change over time, including shape, duration, and magnitude, thereby capturing the bulk of neural activities and their dynamics throughout the stimulation period. b primarily represents the peak amplitude of BOLD responses (i.e., the % BOLD signal change)4 and can be constrained by the assumptions and limitations of the GLM analysis, such as the shape of the hemodynamic response function (HRF). AUC provides greater flexibility in capturing different aspects of neural responses across various brain regions, such as transient peaks and sustained responses.

As mentioned by reviewer #1, correlating the adaptation of BOLD and electrophysiology signals at the brain region level would better signify our findings. We will pursue additional analysis to address this in our forthcoming responses. Reviewer #2 would like us to clarify the image and signal quality of our echo planar imaging (EPI)-based fMRI data, especially in the regions close to the air-tissue interface such as OB, Pir, entorhinal cortex and amygdala, and the methodology for some of the experimental protocols implemented in our study. We will show the raw EPI fMRI images from a representative animal and revise the results, discussion, and methods sections of the manuscript to address reviewer #2's concerns.

In our forthcoming detailed responses to the reviewers' comments and recommendations, we will revise the text, figures, and captions accordingly to address and clarify the questions brought up by both reviewers.

References

(1) Gao, P.P., Zhang, J.W., Chan, R.W., Leong, A.T.L. & Wu, E.X. BOLD fMRI study of ultrahigh frequency encoding in the inferior colliculus. Neuroimage 114, 427-437 (2015).

(2) Leong, A.T.L., Wong, E.C., Wang, X. & Wu, E.X. Hippocampus Modulates Vocalizations Responses at Early Auditory Centers. Neuroimage 270, 119943 (2023).

(3) Gao, P.P., Zhang, J.W., Fan, S.J., Sanes, D.H. & Wu, E.X. Auditory midbrain processing is differentially modulated by auditory and visual cortices: An auditory fMRI study. Neuroimage 123, 22-32 (2015).

(4) Goddard, E. & Mullen, K.T. fMRI representational similarity analysis reveals graded preferences for chromatic and achromatic stimulus contrast across human visual cortex. Neuroimage 215, 116780 (2020).

Author response:

The following is the authors’ response to the original reviews.

Public Reviews:

Reviewer #1 (Public Review):

Summary:

The emergence of Drosophila EM connectomes has revealed numerous neurons within the associative learning circuit. However, these neurons are inaccessible for functional assessment or genetic manipulation in the absence of cell-type-specific drivers. Addressing this knowledge gap, Shuai et al. have screened over 4000 split-GAL4 drivers and correlated them with identified neuron types from the "Hemibrain" EM connectome by matching light microscopy images to neuronal shapes defined by EM. They successfully generated over 800 split-GAL4 drivers and 22 split-LexA drivers covering a substantial number of neuron types across layers of the mushroom body associative learning circuit. They provide new labeling tools for olfactory and non-olfactory sensory inputs to the mushroom body; interneurons connected with dopaminergic neurons and/or mushroom body output neurons; potential reinforcement sensory neurons; and expanded coverage of intrinsic mushroom body neurons. Furthermore, the authors have optimized the GR64f-GAL4 driver into a sugar sensory neuron-specific split-GAL4 driver and functionally validated it as providing a robust optogenetic substitute for sugar reward. Additionally, a driver for putative nociceptive ascending neurons, potentially serving as optogenetic negative reinforcement, is characterized by optogenetic avoidance behavior. The authors also use their very large dataset of neuronal anatomies, covering many example neurons from many brains, to identify neuron instances with atypical morphology. They find many examples of mushroom body neurons with altered neuronal numbers or mistargeting of dendrites or axons and estimate that 1-3% of neurons in each brain may have anatomic peculiarities or malformations. Significantly, the study systematically assesses the individualized existence of MBON08 for the first time. This neuron is a variant shape that sometimes occurs instead of one of two copies of MBON09, and this variation is more common than that in other neuronal classes: 75% of hemispheres have two MBON09's, and 25% have one MBON09 and one MBON08. These newly developed drivers not only expand the repertoire for genetic manipulation of mushroom body-related neurons but also empower researchers to investigate the functions of circuit motifs identified from the connectomes. The authors generously make these flies available to the public. In the foreseeable future, the tools generated in this study will allow important advances in the understanding of learning and memory in Drosophila.

Strengths:

(1) After decades of dedicated research on the mushroom body, a consensus has been established that the release of dopamine from DANs modulates the weights of connections between KCs and MBONs. This process updates the association between sensory information and behavioral responses. However, understanding how the unconditioned stimulus is conveyed from sensory neurons to DANs, and the interactions of MBON outputs with innate responses to sensory context remains less clear due to the developmental and anatomic diversity of MBONs and DANs. Additionally, the recurrent connections between MBONs and DANs are reported to be critical for learning. The characterization of split-GAL4 drivers for 30 major interneurons connected with DANs and/or MBONs in this study will significantly contribute to our understanding of recurrent connections in mushroom body function.

(2) Optogenetic substitutes for real unconditioned stimuli (such as sugar taste or electric shock) are sometimes easier to implement in behavioral assays due to the spatial and temporal specificity with which optogenetic activation can be induced. GR64f-GAL4 has been widely used in the field to activate sugar sensory neurons and mimic sugar reward. However, the authors demonstrate that GR64f-GAL4 drives expression in other neurons not necessary for sugar reward, and the potential activation of these neurons could introduce confounds into training, impairing training efficiency. To address this issue, the authors have elaborated on a series of intersectional drivers with GR64f-GAL4 to dissect subsets of labeled neurons. This approach successfully identified a more specific sugar sensory neuron driver, SS87269, which consistently exhibited optimal training performance and triggered ethologically relevant local searching behaviors. This newly characterized line could serve as an optimized optogenetic tool for sugar reward in future studies.

(3) MBON08 was first reported by Aso et al. 2014, exhibiting dendritic arborization into both ipsilateral and contralateral γ3 compartments. However, this neuron could not be identified in the previously published Drosophila brain connectomes. In the present study, the existence of MBON08 is confirmed, occurring in one hemisphere of 35% of imaged flies. In brains where MBON08 is present, its dendrite arborization disjointly shares contralateral γ3 compartments with MBON09. This remarkable phenotype potentially serves as a valuable resource for understanding the stochasticity of neurodevelopment and the molecular mechanisms underlying mushroom body lobe compartment formation.

Weaknesses:

There are some minor weaknesses in the paper that can be clarified:

(1) In Figure 8, the authors trained flies with a 20s, weak optogenetic conditioning first, followed by a 60s, strong optogenetic conditioning. The rationale for using this training paradigm is not explicitly provided.

These experiments were designed to test if flies could maintain consistent performance with repetitive and intense LED activation, which is essential for experiments involving long training protocols or coactivation of other neurons inside a brain.

In Figure 8E, if data for training with GR64f-GAL4 using the same paradigm is available, it would be beneficial for readers to compare the learning performance using newly generated split-GAL4 lines with the original GR64f-GAL4, which has been used in many previous research studies. It is noteworthy that in previously published work, repeating training test sessions typically leads to an increase in learning performance in discrimination assays. However, this augmentation is not observed in any of the split-GAL4 lines presented in Figure 8E. The authors may need to discuss possible reasons for this.

As the reviewer pointed out, many previous studies including ours used the original Gr64f-GAL4 in olfactory conditioning. Figure 1H of Yamada et al., 2023 (https://doi.org/10.7554/eLife.79042) showed such a result, where the first and second-order olfactory conditioning were assayed. Indeed, the first-order conditioning scores were gradually augmented over repeated training. In this experiment, we used low red LED intensity for the optogenetic activation. In the Figure 8E of the present paper, the first memory test was after 3x pairing of 20s odor with five 1s red LED without intermediate tests. Therefore, flies were already sufficiently trained to show a plateau memory level in “Test1”. In the revision of another recent report (Figure 1C-F of Aso et al., 2023; https://doi.org/10.7554/eLife.85756), we included the learning curve data of our best Gr64f-split-GAL4, SS87269. Under a less saturated training conditioning, SS87269 did show learning augmentation over repeated training.

(2) In line 327, the authors state that in all samples, the β'1 compartment is arborized by MBON09. However, in Figure 11J, the probability of having at least one β'1 compartment not arborized is inferred to be 2%. The authors should address and clarify this conflict in the text to avoid misunderstanding.

The chance of visualizing MBON08 in MCFO images was 21/209 in total (Figure 11I). If we assume that each of four cells adopt MBON08 development fate at this chance, we can calculate the probability for each case of MBON08/09 cell type composition. From this calculation, we inferred approximately 2% of flies would lack innervations to β'1 compartment in at least one hemisphere. However, we didn't observe a lack of β'1 arborizations in 169 sample flies. If these MBONs independently develop into MBON08 at 21/209 odds, the chance of never observing two MBON08s in either hemisphere of all 169 samples is 3.29%. Therefore, some developmental mechanisms may prevent the emergence of two MBON08 in the same hemisphere.

In the revised manuscript, we displayed these estimated probability for each case separately, and annotated actual observation on the right side.

(3) In general, are the samples presented male or female? This sample metadata will be shown when the images are deposited in FlyLight, but it would be useful in the context of this manuscript to describe in the methods whether animals are all one sex or mixed sex, and in some example images (e.g. mAL3A) to note whether the sample is male or female.

The samples presented in this study are mixed sex, except for Figure 11I, where genders are specified. We provided metadata information of the presented images in Supplemental File 7, and we added a paragraph in the in the method section:

“Most samples were collected from females, though typically at least one male fly was examined for each driver line. While we noticed certain lines such as SS48900, exhibited distinct expression patterns in females and males, we did not particularly focus on sexual dimorphism, which is analyzed elsewhere (Meissner et al. 2024). Therefore, unless stated otherwise, the presented samples are of mixed gender.

Detailed metadata, including gender information and the reporter used, can be found in Supplementary File 7.”

Reviewer #2 (Public Review):

Summary:

The article by Shuai et al. describes a comprehensive collection of over 800 split-GAL4 and split-LexA drivers, covering approximately 300 cell types in Drosophila, aimed at advancing the understanding of associative learning. The mushroom body (MB) in the insect brain is central to associative learning, with Kenyon cells (KCs) as primary intrinsic neurons and dopaminergic neurons (DANs) and MB output neurons (MBONs) forming compartmental zones for memory storage and behavior modulation. This study focuses on characterizing sensory input as well as direct upstream connections to the MB both anatomically and, to some extent, behaviorally. Genetic access to specific, sparsely expressed cell types is crucial for investigating the impact of single cells on computational and functional aspects within the circuitry. As such, this new and extensive collection significantly extends the range of targeted cell types related to the MB and will be an outstanding resource to elucidate MB-related processes in the future.

Strengths:

The work by Shuai et al. provides novel and essential resources to study MB-related processes and beyond. The resulting tools are publicly available and, together with the linked information, will be foundational for many future studies. The importance and impact of this tool development approach, along with previous ones, for the field cannot be overstated. One of many interesting aspects arises from the anatomical analysis of cell types that are less stereotypical across flies. These discoveries might open new avenues for future investigations into how such asymmetry and individuality arise from development and other factors, and how it impacts the computations performed by the circuitry that contains these elements.

Weaknesses:

Providing such an array of tools leaves little to complain about. However, despite the comprehensive genetic access to diverse sensory pathways and MB-connected cell types, the manuscript could be improved by discussing its limitations. For example, the projection neurons from the visual system seem to be underrepresented in the tools produced (or almost absent). A discussion of these omissions could help prevent misunderstandings.

We internally distributed efforts to produce split-GAL4 lines at Janelia Research Campus. The recent preprint (Nern et al., 2024; doi: https://doi.org/10.1101/2024.04.16.589741) described the full collection of split-GAL4 driver lines in the optic lobe including the visual projection neurons to the mushroom body. We cited this preprint in the revised manuscript by adding a short paragraph of discussion.

“Although less abundant than the olfactory input, the MB also receives visual information from the visual projection neurons (VPNs) that originate in the medulla and lobula and are targeted to the accessory calyx (Vogt et al. 2016; Li et al. 2020). A recent preprint described the full collection of split-GAL4 driver lines in the optic lobe, which includes the VPNs to the MB (Nern et al. 2024).”

Additionally, more details on the screening process, particularly the selection of candidate split halves and stable split-GAL4 lines, would provide valuable insights into the methodology and the collection's completeness.

The details of our split-GAL4 design and screening procedures were described in previous studies (Aso et al., 2014; Dolan et al., 2019). Available data and tools to design split-GAL4 changed over time, and we took different approaches accordingly. Many of split-GAL4 lines presented in this study were designed and screened in parallel to the lines for MBONs and DANs in 2010-2014 when MCFO images of GAL4 drivers and EM connectome were not yet available. With knowledge of where MBONs and DANs project, I (Y.A.) manually examined and annotated thousands of confocal stacks (Jenett et al., 2012; https://doi.org/10.1016/j.celrep.2012.09.011) to find candidate cell types that may concat with them.

Later I used more advanced computational tools (Otsuna et al., 2018; doi: https://doi.org/10.1101/318006) and MCFO images aligned to the standard brain volume (Meissner et al., 2023; DOI: 10.7554/eLife.80660.). Now, if one needs to further generate split-GAL4 lines for cell type identified in EM connectome data, neuron bridge website (https://neuronbridge.janelia.org/) can be very helpful to provide a list of GAL4 drivers that may label the neuron of interest.

Reviewer #3 (Public Review):

Summary:

Previous research on the Drosophila mushroom body (MB) has made this structure the best-understood example of an associative memory center in the animal kingdom. This is in no small part due to the generation of cell-type specific driver lines that have allowed consistent and reproducible genetic access to many of the MB's component neurons. The manuscript by Shuai et al. now vastly extends the number of driver lines available to researchers interested in studying learning and memory circuits in the fly. It is an 800-plus collection of new cell-type specific drivers target neurons that either provide input (direct or indirect) to MB neurons or that receive output from them. Many of the new drivers target neurons in sensory pathways that convey conditioned and unconditioned stimuli to the MB. Most drivers are exquisitely selective, and researchers will benefit from the fact that whenever possible, the authors have identified the targeted cell types within the Drosophila connectome. Driver expression patterns are beautifully documented and are publicly available through the Janelia Research Campus's Flylight database where full imaging results can be accessed. Overall, the manuscript significantly augments the number of cell type-specific driver lines available to the Drosophila research community for investigating the cellular mechanisms underlying learning and memory in the fly. Many of the lines will also be useful in dissecting the function of the neural circuits that mediate sensorimotor circuits.

Strengths:

The manuscript represents a huge amount of careful work and leverages numerous important developments from the last several years. These include the thousands of recently generated split-Gal4 lines at Janelia and the computational tools for pairing them to make exquisitely specific targeting reagents. In addition, the manuscript takes full advantage of the recently released Drosophila connectomes. Driver expression patterns are beautifully illustrated side-by-side with corresponding skeletonized neurons reconstructed by EM. A comprehensive table of the new lines, their split-Gal4 components, their neuronal targets, and other valuable information will make this collection eminently useful to end-users. In addition to the anatomical characterization, the manuscript also illustrates the functional utility of the new lines in optogenetic experiments. In one example, the authors identify a specific subset of sugar reward neurons that robustly promotes associative learning.

Weaknesses:

While the manuscript succeeds in making a mass of descriptive detail quite accessible to the reader, the way the collection is initially described - and the new lines categorized - in the text is sometimes confusing. Most of the details can be found elsewhere, but it would be useful to know how many of the lines are being presented for the first time and have not been previously introduced in other publications/contexts.

We revised the text as below.

“Among the 828 lines, a subset of 355 lines, collectively labeling at least 319 different cell types, exhibit highly specific and non-redundant expression patterns are likely to be particularly valuable for behavioral experiments. Detailed information, including genotype, expression specificity, matched EM cell type(s), and recommended driver for each cell type, can be found in Supplementary File 1. A small subset of 40 lines from this collection have been previously used in studies (Aso et al., 2023; Dolan et al., 2019; Gao et al., 2019; Scaplen et al., 2021; Schretter et al., 2020; Takagi et al., 2017; Xie et al., 2021; Yamada et al., 2023). All transgenic lines newly generated in this study are listed in Supplementary File 2 (Aso et al., 2023; Dolan et al., 2019; Gao et al., 2019; Scaplen et al., 2021; Schretter et al., 2020; Takagi et al., 2017; Xie et al., 2021; Yamada et al., 2023).”

And where can the lines be found at Flylight? Are they listed as one collection or as many?

They are listed as one collection - “Aso 2021” release. It is named “2021” because we released the images and started sharing lines in December of 2021 without a descriptive paper. We added a sentence in the Methods section.

“All splitGAL4 lines can be found at flylight database under “Aso 2021” release, and fly strains can be requested from Janelia or the Bloomington stock center.”

Also, the authors say that some of the lines were included in the collection despite not necessarily targeting the intended type of neuron (presumably one that is involved in learning and memory). What percentage of the collection falls into this category?

We do not have a good record of split-GAL4 screening to calculate the chance to intersect unintended cell types, but it was rather rare. Those unintended cell types can still be a part of circuits for associative learning (e.g. olfactory projection neurons) or totally unrelated cell types. For instance, among a new collection of split-LexA lines using Gr43a-LexADBD hemidriver (Figure 7-figure supplement 2), one line specifically intersected T1 neurons in the optic lobe despite that the AD line was selected to intersect sugar sensory neurons. We suspect that this is due to ectopic expression of Gr43a-LexADBD. Nonetheless, we included it in the paper because cell-type-specific Split-LexA driver for T1 will be useful irrespective of whether the expression of Gr43a gene is expressed in T1 or not.

And what about the lines that the authors say they included in the collection despite a lack of specificity? How many lines does this represent?

For a short answer, there are about 100 lines in the collection that lack the specificity for behavioral experiments.

We ranked specificity of split-GAL4 drivers in the Supplementary File 1. Rank 2 are the ideal lines, Rank 1 are less ideal but acceptable, and Rank 0 is not suitable for activation screening in behavioral experiments. Out of the 828 split-GAL4 lines reported here, there are 413, 305 and 103 lines in rank2, rank1 and rank0 categories respectively. 7 lines are not ranked for specificity because only flipout expression data are available.

Recommendations for the authors:

Reviewer #2 (Recommendations For The Authors):

As mentioned elsewhere and in addition to the minor points below, it is advisable for the authors to elaborate on the details of the screening process. Furthermore, a discussion about the circuits not targeted by their research, such as the visual projection neurons, would be beneficial.

See the response above to Reviewer #2’s public review.

Line 32-33: The citations are very fly-centric. the authors might want to consider reviews on the MB of other insect species regarding learning and memory.

We additionally cited Rybak and Menzel 2017’s book chapter on honey bee mushroom body.

Line 43-44: Citations should be added, e.g. Séjourné et al. (2011), Pai et al. (2013), Plaçais et al. (2013).

Citation added

Line 50-52: Citation Hulse et al. (2021) should be added.

Citation added

Line 162: In this part, it might be valuable for the reader to understand which of these PNs are actually connecting with KCs.

A full list of cell types within the MB were provided in Supplementary File 4 of the revised manuscript. See also response to Reviewer 3, Lines 150-1.

Line 179: Citation Burke et al. (2012) should be mentioned.

Citation added

Line 181: Thermogenic might be thermogenetic.

Corrected

Line 189: Citations add Otto et al. (2020) and Felsenberg et al. (2018).

Citations added

Line 208ff: The authors should consider discussing why they did not use other GR and IR promoters. For example, Gr5a is prominent in sugar-sensing, while Ir76b could be a reinforcement signal related to yeast food (Steck et al., 2018; Ganguly et al., 2017; see also Corfas et al., 2019 for local search).

We focused on the Gr64f promoter because of its relatively broad expression and successful use of Gr64f-GAL4 for fictive reward experiment. We added the Split-LexA lines with Gr43a and Gr66a promoters (Figure 7-figure supplement 2). Other gustatory sensory neurons also have the potential to be reinforcement signals, but we just did not have the bandwidth to cover them all.

Line 319: Consider citing Linneweber et al. (2020) for a neurodevelopmental account of such individuality.

We added a sentence and cited this reference.

“On the other hand, the neurodevelopmental origin of neuronal morphology appeared to have functional significance on behavioral individuality (Linneweber et al. 2020).”

Line 352: Citation add Hulse et al. (2021).

Citations added

Line 356ff: The utility and value of Split-LexA may not be apparent to non-expert readers. Moreover, how were LexADBDs chosen for creating these lines?

We have added an introductory sentence at the beginning of the paragraph and explained that these split-LexA lines were a conversion of split-GAL4 lines that were published in 2014 and frequently used in studying the mushroom body circuit.

“Split-GAL4 lines enable cell-type-specific manipulation, but some experiments require independent manipulation of two cell types. Split-GAL4 lines can be converted into split-LexA lines by replacing the GAL4 DNA binding domain with that of LexA (Ting et al., 2011). To broaden the utility of the split-GAL4 lines that have been frequently used since the publication in 2014 (Aso et al., 2014a), we have generated over 20 LexADBD lines to test the conversions of split-GAL4 to split-LexA. The majority (22 out of 34) of the resulting split-LexA lines exhibited very similar expression patterns to their corresponding original split-GAL4 lines (Figure 12).”

Line 374: Italicize Drosophila melanogaster.

Revised as suggested.

Reviewer #3 (Recommendations For The Authors):

Major Comments:

As mentioned in the Public Review, the drivers are nicely classified in the various subsections of the manuscript, but the statements in the text summarizing how many lines there are in specific categories are often confusing. For example, line 129 refers to "drivers encompassing 111 cell types that connect with the DANs and MBONs", but Figure 1E indicates that 46 new cell types downstream of MBONs and upstream of DANs have been generated. This seems like a discrepancy.

The 46 cell types in Figure 1E consider only the CRE/SMP/SIP/SLP area, where MBON downstreams and DAN upstreams are highly enriched, while the 111 cell types include all. To avoid confusion, we removed the “MBON downstream and DAN upstream” counting in Figure 1E in the revised manuscript.

Also, at line 75 the MBON lines previously generated by Rubin and Aso (2023) are referred to as though they are separate from the 828 described "In this report." Supplementary file 1 suggests, however, that they are included as part of this report.

Twenty five lines generated in Rubin and Aso (2023) were initially included in Supplementary file 1 for the convenience of users, but they were not counted towards the 828 new lines described in this report. To avoid confusion, we removed these 25 lines in the revised manuscript. Now all lines listed in Supplementary file 1 were generated in this study (“Aso 2021” release), and if a line has been used in earlier studies, or introduced in other contexts, for example the accompanying omnibus preprint (Meissener 2024, doi: 10.1101/2024.01.09.574419), the citations are listed in the reference column.

More generally, in lines 94-102 "828 useful lines based on their specificity, intensity and non-redundancy" are referred to, but they are subsequently subdivided into categories of lines with lower specificity (i.e. with off-target expression) and lines that did not target intended cell types (presumably ones unlikely to be involved in learning and memory). It would be useful to know how many lines (at least roughly) fall into these subcategories.

See the response above to Reviewer #3’s public review.

Finally, Figures 3B & C indicate cell types connected to DANs and MBONs and the number for which Split-Gal4 lines are available. The text (lines 136-7) states that the new collection covers 30 of these major cell types (Figure 3C)," but Figure 3C clearly has more than 30 dots showing the drivers available. Presumably existing and new driver lines are being pooled, but this should either be explained or the two should be distinguished.

“(Figure 3C)” was replaced with “(Supplementaryl File 3)” in the revised manuscript to correct the reference. Figure 3B & C are plots of all MB interneurons, not just the major cell types.

Minor Comments:

Although the paper is generally well written there are minor grammatical errors throughout (e.g. dropped articles, odd constructions, etc.) that somewhat detract from an otherwise smooth and enjoyable reading experience. A quick editing pass by a native speaker (i.e. any of several of the authors) could clean up these and numerous other small mistakes. A few examples: line 138 "presented" should be present; line 204: "contain off-targeted expressions" should be "have off-target expression;" line 219: "usage to substitute reward" is awkward at best and could be something like "use in generating fictive rewards"; line 326 "arborize[s]"; l. 331 "Based on the likelihood" should be something like "based on these observations"'; line 349 "[is] likely to appear"; l. 352 "extensive connection[s]"; line 353 "has [a] strong influence;" l. 963 "Projections" should be singular; etc.

All the mentioned examples have been corrected, and we have asked a native speaker to edit through the revised manuscript.

Lines 81-3: Is the lookup table referred to Suppl. File 1? A reference is desirable.

Yes, the lookup table referred to “Supplementary File 1” and a reference was added.

Lines 111-2: what is a "non-redundant set of...cell types?" Cell types that are represented by a single cell (or bilateral pair)? Or does this sentence mean that of the 828 lines, 355 are specific to a single cell type, and in total 319 cell types are targeted? The statement is confusing.

We revised the text as below.

“Figure 1E provides an overview of the categories of covered cell types. Among the 828 lines, a subset of 355 lines, collectively labeling at least 319 different cell types, exhibit highly specific and non-redundant expression patterns are likely to be particularly valuable for behavioral experiments. Detailed information, including genotype, expression specificity, matched EM cell type(s), and recommended driver for each cell type, can be found in Supplementary File 1. A small subset of 40 lines from this collection have been previously used in studies (Aso et al.,

2023; Dolan et al., 2019; Gao et al., 2019; Scaplen et al., 2021; Schretter et al., 2020; Takagi et al., 2017; Xie et al., 2021; Yamada et al., 2023). All transgenic lines newly generated in this study are listed in Supplementary File 2 (Aso et al., 2023; Dolan et al., 2019; Gao et al., 2019; Scaplen et al., 2021; Schretter et al., 2020; Takagi et al., 2017; Xie et al., 2021; Yamada et al., 2023).”

Line 148: "MB major interneurons" is a confusing descriptor for postsynaptic partners of MBONs.

We added a sentence to clarify the definition of the “MB major interneurons”.

“In the hemibrain EM connectome, there are about 400 interneuron cell types that have over 100 total synaptic inputs from MBONs and/or synaptic outputs to DANs. Our newly developed collection of split-GAL4 drivers covers 30 types of these ‘major interneurons’ of the MB (Supplementary File 3).”

Lines 150-1: Not sure what is meant by "have innervations within the MB." Sounds like cells are presynaptic to KCs, DANS, and MBONs, but Figure 3 Figure Supplement 1 indicates they include neurons that both provide and receive innervation to/from MB neurons. Please clarify.

For clarification, in the revised manuscript we have included a full list of cell types within the MB in Supplementary File 4. Included are all neurons with >= 50 pre-synaptic connections or with >=250 post-synaptic connections in the MB roi in the hemibrain (excluding the accessory calyx). The cell types include KCs, MBONs, DANs, PNs, and a few other cell types. The coverage ratio was updated based on this list.

Also, in line 152, what does it mean that they "may have been overlooked previously?" this seems unnecessarily ambiguous. Were they overlooked or weren't they?

Changed the text to “These lines offer valuable tools to study cell types that previously are not genetically accessible. Notably, SS85572 enables the functional study of LHMB1, which forms a rare direct pathway from the calyx and the lateral horn (LH) to the MB lobes (Bates et al., 2020). ”

Line 158 refers to PN cells within the MB, which are not mentioned in any place else as MB components.

What are these PNs and how do they differ from MBONs?

See responses to Lines 150-1 for clarification of cell types within the MB.

Line 188: not clear what is meant by "more continual learning tasks".

We rephrase it as “more complex learning tasks” to avoid jargon.

Line 235: Not clear why "extended training with high LED intensity" wouldn't promote the formation of robust memories. Is this for some reason unexpected based on previous experiments? Please explain.

See responses to weakness #1 of the same reviewer

Lines 317-9: It would be useful to state here that MB0N08 and MB0N09 are the two neurons labeled by MB083C.

Revised as suggested.

Line 368: Presumably the "lookup table" referred to is Supplementary File 1, but a reference here would be useful.

Yes, Supplementary File 1 and a reference was added.

Comments on Figures:

Figure 1C The "Dopamine Neurons" label position doesn't align with the Punishment and Reward labels, which is a bit confusing.

They are intentionally not aligned, because dopamine neurons are not reward/punishment per se. We intend to use the schematic to show that the punishment and reward are conveyed to the MB through the dopamine neuron layer, just as the output from the MB output neuron layer is used to guide further integration and actions. To keep the labels of “Dopamine neurons” and “MB Output Neurons” in a symmetrical position, we decide to keep the original figure unchanged. But we thank the reviewer for the kind suggestion.

Figure 1F and Figure 1 - Figure Supplement 1: the light gray labels presumably indicate the (EM-identified) neuron labeled by each line, but this should be explicitly stated in the figure legends. It would also be useful in the legends to direct the reader to the key (Supplementary File 1) for decoding neuronal identities.

Revised as suggested.

Figure 2: For clarity, I'd recommend titling this figure "LM-EM Match of the CRE011-specific driver SS45245". This reduces the confusion of mixing and matching the driver and cell-type names. Also, it would be helpful to indicate (e.g. with labels above the figure parts) that A & B represent the MCFO characterization step and C & D represent the LM-EM matching step of the pipeline. Revised as suggested.

Figure 6: For clarity, it would be useful to separately label the PN and sensory neuron groups. Also, for the sensory neurons at the bottom, what is the distinction between the cell names in gray and black font?

Figure 6 was updated to separate the non-olfactory PN and sensory neuron groups. The gray was intended for olfactory receptor neuron cell types that are additionally labeled in the driver lines. To avoid confusion, the gray cell types were removed in the revised figure, and a clarification sentence was added to the legend.

“Other than thermo-/hygro-sensory receptor neurons (TRNs and HRNs), SS00560 and MB408B also label olfactory receptor neurons (ORNs): ORN_VL2p and ORN_VC5 for SS00560, ORN_VL1 and ORN_VC5 for MB408B.”

Figure 7A: It's unclear why the creation of 6 Gr64f-LexADBD lines is reported. Aren't all these lines the same? If not, an explanation would be useful.

These six Gr64f-LexADBD lines are with different insertion sites, and with the presence or absence of the p10 translational enhancer. Explanation was added to legend. Enhanced expression level with p10 can be helpful to compensate for the general tendency that split-LexA is weaker than split-GAL4. Different insertions will be useful to avoid transvections with split-GAL4s, which are mostly in attP40 and attP2.

Figure 8F: It would help to include in the legend a brief description of each parameter being measured-essentially defining the y-axis label on the graphs as in Figure Supplement 2. Also, how is the probability of return calculated and what behavioral parameter does the change of curvature refer to?

We added a brief description to the behavioral parameters in the legend of Figure 8F.

“Return behavior was assessed within a 15-second time window. The probability of return (P return) is the percentage of flies that made an excursion (>10 mm) and then returned to within 3 mm of their initial position. Curvature is the ratio of angular velocity to walking speed.”

Figure 9E: What are the parenthetical labels for lines SS49267, SS49300, and SS35008?

They are EM bodyIDs. Figure legend was revised.

Author response:

The following is the authors’ response to the original reviews.

eLife Assessment

This study compiles a wide range of results on the connectivity, stimulus selectivity, and potential role of the claustrum in sensory behavior. While most of the connectivity results confirm earlier studies, this valuable work provides incomplete evidence that the claustrum responds to multimodal stimuli and that local connectivity is reduced across cells that have similar long-range connectivity. The conclusions drawn from the behavioral results are weakened by the animals' poor performance on the designed task.This study has the potential to be of interest to neuroscientists.

We thank the editor and the reviewers for their feedback on our work, which we have incorporated to help improve interpretation of our findings as outlined in the response below. While we agree with the editor that further work is necessary to provide a comprehensive understanding of claustrum circuitry and activity, this is true of most scientific endeavors and therefore we feel that describing this work as “incomplete” unfairly mischaracterizes the intent of the experiments performed which provide fundamental insights into this poorly understood brain region. Additionally, as identified in the main text, methods section, and our responses to the comments below, we disagree that the behavioral results are “weakened” by the performance of the animals. Our goal was to assess what information animals learned and used in an ambiguous sensory/reward environment, not to shape them toward a particular behavior and interpret the results solely based on their accuracy in performing the task.

Public Reviews:

Reviewer #1 (Public Review):

Summary:

The paper by Shelton et al investigates some of the anatomical and physiological properties of the mouse claustrum. First, they characterize the intrinsic properties of claustrum excitatory and inhibitory neurons and determine how these different claustrum neurons receive input from different cortical regions. Next, they perform in vitro patch clamp recordings to determine the extent of intraclaustrum connectivity between excitatory neurons. Following these experiments, in vivo axon imaging was performed to determine how claustrum-retrosplenial cortex neurons are modulated by different combinations of auditory, visual, and somatosensory input. Finally, the authors perform claustrum lesions to determine if claustrum neurons are required for performance on a multisensory discrimination task

Strengths:

An important potential contribution the authors provide is the demonstration of intra-claustrum excitation. In addition, this paper provides the first experimental data where two cortical inputs are independently stimulated in the same experiment (using 2 different opsins). Overall, the in vitro patch clamp experiments and anatomical data provide confirmation that claustrum neurons receive convergent inputs from areas of the frontal cortex. These experiments were conducted with rigor and are of high quality.

We thank the reviewer for their positive appraisal of our work.

Weaknesses:

The title of the paper states that claustrum neurons integrate information from different cortical sources. However, the authors did not actually test or measure integration in the manuscript. They do show physiological convergence of inputs on claustrum neurons in the slice work. Testing integration through simultaneous activation of inputs was not performed. The convergence of cortical input has been recently shown by several other papers (Chia et al), and the current paper largely supports these previous conclusions. The in vivo work did test for integration because simultaneous sensory stimulations were performed. However, integration was not measured at the single cell (axon) level because it was unclear how activity in a single claustrum ROI changes in response to (for example) visual, tactile, and visual-tactile stimulations. Reading the discussion, I also see the authors speculate that the sensory responses in the claustrum could arise from attentional or salience-related inputs from an upstream source such as the PFC. In this case, claustrum cells would not integrate anything (but instead respond to PFC inputs).

We thank the reviewer for raising this point. In response, we have provided a definition of “integration” in the manuscript text (lines 112-114, 353-354):

“...single-cell responsiveness to more than one input pathway, e.g. being capable of combining and therefore integrating these inputs.”

The reviewer’s point about testing simultaneous input to the claustrum is well made but not possible with the dual-color optogenetic stimulation paradigm used in our study as noted in the Results and Discussion sections (see also Klapoetke et al., 2014, Hooks et al., 2015). The novelty of our paper comes from testing these connections in single CLA neurons, something not shown in other studies to-date (Chia et al., 2020; Qadir et al., 2022), which average connectivity over many neurons.

Finally, we disagree with the reviewer regarding whether integration was tested at the single-axon level and provide data and supplementary figures to this effect (Fig. 6, Supp. Fig. S14, lines 468-511) . Although the possibility remains that sensory-related information may arise in the prefrontal cortex, as we note, there is still a large collection of studies (including this one) that document and describe direct sensory inputs to the claustrum (Olson & Greybeil, 1980; Sherk & LeVay, 1981; Smith & Alloway, 2010; Goll et al., 2015; Atlan et al., 2017; etc.). We have updated the wording of these sections to note that both direct and indirect sensory input integration is possible.

The different experiments in different figures often do not inform each other. For example, the authors show in Figure 3 that claustrum-RSP cells (CTB cells) do not receive input from the auditory cortex. But then, in Figure 6 auditory stimuli are used. Not surprisingly, claustrum ROIs respond very little to auditory stimuli (the weakest of all sensory modalities). Then, in Figure 7 the authors use auditory stimuli in the multisensory task. It seems that these experiments were done independently and were not used to inform each other.

The intention behind the current manuscript was to provide a deep characterisation of claustrum to inform future research into this enigmatic structure. In this case, we sought to test pathways in vivo that were identified as being weak or absent in vitro to confirm and specifically rule out their influence on computations performed by claustrum. We agree with the reviewer’s assessment that it is not surprising that claustrum ROIs respond weakly to auditory stimuli. Not testing these connections in vivo because of their apparent sparsity in vitro would have represented a critical gap in our knowledge of claustrum responses during passive sensory stimulation.

One novel aspect of the manuscript is the focus on intraclaustrum connectivity between excitatory cells (Figure 2). The authors used wide-field optogenetics to investigate connectivity. However, the use of paired patch-clamp recordings remains the ground truth technique for determining the rate of connectivity between cell types, and paired recordings were not performed here. It is difficult to understand and gain appreciation for intraclaustrum connectivity when only wide-field optogenetics is used.

We thank the reviewer for acknowledging the novelty of these experiments. We further acknowledge that paired patch-clamp recordings are the gold standard for assessing synaptic connectivity. Typically such experiments are performed in vitro, a necessity given the ventral location of claustrum precluding in vivo patching. In vitro slice preparations by their very nature sever connections and lead to an underestimate of connectivity as noted in our Discussion. Kim et al. (2016) have done this experiment in coronal slices with the understanding that excitatory-excitatory connectivity would be local (<200 μm) and therefore preserved. We used a variety of approaches that enabled us to explore connectivity along the longitudinal axis of the brain (the rostro-caudal, e.g. “long” axis of the claustrum), providing fresh insight into the circuitry embedded within this structure that would be challenging to examine using dual recordings. Further, our optogenetic method (CRACM, Petreanu et al., 2007), has been used successfully across a variety of brain structures to examine excitatory connectivity while circumventing artifacts arising from the slice axis.

In Figure 2, CLA-rsp cells express Chrimson, and the authors removed cells from the analysis with short latency responses (which reflect opsin expression). But wouldn't this also remove cells that express opsin and receive monosynaptic inputs from other opsin-expressing cells, therefore underestimating the connectivity between these CLA-rsp neurons? I think this needs to be addressed.

The total number of opsin-expressing CLA neurons in our dataset is 4/46 tested neurons. Assuming all of these neurons project to RSP, they would have accounted for 4/32 CLARSP neurons. Given the rate of monosynaptic connectivity observed in this study, these neurons would only contribute 2-3 additional connected neurons. Therefore, the exclusion of these neurons does not significantly impact the overall statistical accuracy of our connectivity findings.

In Figure 5J the lack of difference in the EPSC-IPSC timing in the RSP is likely due to 1 outlier EPSC at 30 ms which is most likely reflecting polysynaptic communication. Therefore, I do not feel the argument being made here with differences in physiology is particularly striking.

We thank the reviewer for their attention to detail about this analysis. We have performed additional statistics and found that leaving this neuron out does not affect the significance of the results (new p-value = 0.158, original p-value = 0.314, Mann-Whitney U test). We have removed this datapoint from the figure and our analysis.

In the text describing Figure 5, the authors state "These experiments point to a complex interaction ....likely influenced by cell type of CLA projection and intraclaustral modules in which they participate". How does this slice experiment stimulating axons from one input relate to different CLA cell types or intra-claustrum circuits? I don't follow this argument.

We have removed this speculation from the Results section.

In Figure 6G and H, the blank condition yields a result similar to many of the sensory stimulus conditions. This blank condition (when no stimulus was presented) serves as a nice reference to compare the rest of the conditions. However, the remainder of the stimulation conditions were not adjusted relative to what would be expected by chance. For example, the response of each cell could be compared to a distribution of shuffled data, where time-series data are shuffled in time by randomly assigned intervals and a surrogate distribution of responses generated. This procedure is repeated 200-1000x to generate a distribution of shuffled responses. Then the original stimulus-triggered response (1s post) could be compared to shuffled data. Currently, the authors just compare pre/post-mean data using a Mann-Whitney test from the mean overall response, which could be biased by a small number of trials. Therefore, I think a more conservative and statistically rigorous approach is warranted here, before making the claim of a 20% response probability or 50% overall response rate.

We appreciate the reviewer's thorough analysis and suggestion for a more conservative statistical approach. We acknowledge that responses on blank trials occur about 10% of the time, indicating that response probabilities around this level may not represent "real" responses. To address this, we will include the responses to the blank condition in the manuscript (lines 505-509). This will allow readers to make informed decisions based on the presented data.

Regarding Figure 6, a more conventional way to show sensory responses is to display a heatmap of the z-scored responses across all ROIs, sorted by their post-stimulus response. This enables the reader to better visualize and understand the claims being made here, rather than relying on the overall mean which could be influenced by a few highly responsive ROIs.

We apologize to the reviewer that our data in this figure was challenging to interpret. We have included an additional supplemental figure (Supp. Fig. S15) that displays the requested information.

For Figure 6, it would also help to display some raw data showing responses at the single ROI level and the population level. If these sensory stimulations are modulating claustrum neurons, then this will be observable on the mean population vector (averaged df/f across all ROIs as a function of time) within a given experiment and would add support to the conclusions being made.

We appreciate the reviewer’s desire to see more raw data – we would have included this in the figure given more space. However, the average df/f across all ROIs is shown as a time series with 95% confidence intervals in Fig. 6D.

As noted by the authors, there is substantial evidence in the literature showing that motor activity arises in mice during these types of sensory stimulation experiments. It is foreseeable that at least some of the responses measured here arise from motor activity. It would be important to identify to what extent this is the case.

While we acknowledge that some responses may arise from motor-related activity, addressing this comprehensively is beyond the scope of this paper. Given the extensive number of trials and recorded axonal segments, we believe that motor-related activity is unlikely to significantly impact the average response across all trials. Future studies focusing specifically on motor activity during sensory stimulation experiments would be needed to elucidate this aspect in detail.

All claims in the results for Figure 6 such as "the proportion of responsive axons tended to be highest when stimuli were combined" should be supported by statistics.

We have provided additional statistics in this section (lines 490-511) to address the reviewer’s comment.

In Figure 7, the authors state that mice learned the structure of the task. How is this the case, when the number of misses is 5-6x greater than the number of hits on audiovisual trials (S Figure 19). I don't get the impression that mice perform this task correctly. As shown in Figure 7I, the hit rate is exceptionally low on the audiovisual port in controls. I just can't see how control and lesion mice can have the same hit rate and false alarm rate yet have different d'. Indeed, I might be missing something in the analysis. However, given that both groups of mice are not performing the task as designed, I fail to see how the authors' claim regarding multisensory integration by the claustrum is supported. Even if there is some difference in the d' measure, what does that matter when the hits are the least likely trial outcome here for both groups.

We thank the reviewer for their comments and hope the following addresses their confusion about the performance of animals during our multimodal conditioning task.

Firstly, as pointed out by the reviewer, the hit-rate (HR) is lower than false-alarm-rate (FR) but crucially only when assessed explicitly within-condition (e.g. just auditory or just visual stimulation). Given the multimodal nature of the assay, HR and FR could also be evaluated across different trials, unimodal and multimodal, for both auditory and visual stimuli. Doing so resulted in a net positive d', as observed by the reviewer. From this perspective, and as documented in the Methods (Multimodal Conditioning and Reversal Learning) and Supplemental Figures, mice do indeed learn the conditioning task and perform at above-chance levels.

Secondly, as raised in the Discussion, an important caveat of this assay was that it was unnecessary for mice to learn the task structure explicitly but, rather, that they respond to environmental cues in a reward-seeking manner that indicated perception of a stimulus. "Performance" as it is quantified here demonstrates a perceptual difference between conditions that is observed through behavioral choice and timing, not necessarily the degree to which the mice have an understanding of the task per se.

In the discussion, it is stated that "While axons responded inconsistently to individual stimulus presentations, their responsivity remained consistent between stimuli and through time on average...". I do not understand this part of the sentence. Does this mean axons are consistently inconsistent?

The reviewer’s interpretation is correct – although recorded axons tended to have a preferred stimulus or combination of stimuli, they displayed variability in their responses (response probability), though little or no variability in their likelihood to respond over time (on average).

In the discussion, the authors state their axon imaging results contrast with recent studies in mice. Why not actually do the same analysis that Ollerenshaw did, so this statement is supported by fact? As pointed out above, the criteria used to classify an axon as responsive to stimuli were very liberal in this current manuscript.

While we appreciate this comment from the reviewer, we feel that it was not necessary to perform similar analyses to those of Ollerenshaw et al in order to appreciate that methodological differences between these studies would have confounded any comparisons made, as we note in the Discussion.

I find the discussion wildly speculative and broad. For example, "the integrative properties of the CLA could act as a substrate for transforming the information content of its inputs (e.g. reducing trial-to-trial variability of responses to conjunctive stimuli...)". How would a claustrum neuron responding with a 10% reliability to a stimuli (or set of stimuli) provide any role in reducing trial-to-trial variability of sensory activity in the cortex?

We thank the reviewer for their feedback. We acknowledge the reviewer's concern regarding the speculative nature of our discussion. To address the specific point raised, while a neuron with a 10% reliability might appear limited in reducing trial-to-trial variability in sensory activity, it's possible that such neurons are responsive to a combination of stimuli or conditions not fully controlled or recorded in our current setup. For instance, variables like the animal’s attentional or motivational states could influence the responsiveness of claustrum neurons, thus integrating these inputs could theoretically modulate cortical processing. We have refined this section to clarify these points (now lines 810-813).

Reviewer #2 (Public Review):

Summary:

In this manuscript, Shelton et al. explore the organization of the Claustrum. To do so, they focus on a specific claustrum population, the one projecting to the retrosplenial cortex (CLA-RSP neurons). Using an elegant technical approach, they first described electrophysiological properties of claustrum neurons, including the CLA-RSP ones. Further, they showed that CLA-RSP neurons (1) directly excite other CLA neurons, in a 'projection-specific' pattern, i.e. CLA-RSP neurons mainly excite claustrum neurons not projecting to the RSP and (2) receive excitatory inputs from multiple cortical territories (mainly frontal ones). To confirm the 'integrative' property of claustrum networks, they then imaged claustrum axons in the cortex during singleor multi-sensory stimulations. Finally, they investigated the effect of CLA-RSP lesion on performance in a sensory detection task.

Strengths:

Overall, this is a really good study, using state-of-the-art technical approaches to probe the local/global organization of the Claustrum. The in-vitro part is impressive, and the results are compelling.

We thank the reviewer for their positive appraisal of our work.

Weaknesses:

One noteworthy concern arises from the terminology used throughout the study. The authors claimed that the claustrum is an integrative structure. Yet, integration has a specific meaning, i.e. the production of a specific response by a single neuron (or network) in response to a specific combination of several input signals. In this study, the authors showed compelling results in favor of convergence rather than integration. On a lighter note, the in-vivo data are less convincing, and do not entirely support the claim of "integration" made by the authors.

We thank the reviewer for their clarity on this issue. We absolutely agree that without clear definition in the study, interpretation of our data could be misconstrued for one of several possible meanings. We have updated our Introduction, Results, and Discussion text to reflect the definition of ‘integration’ we used in the interpretation of our work and hope this clarifies our intent to the reader.

Reviewer #3 (Public Review):

The claustrum is one of the most enigmatic regions of the cerebral cortex, with a potential role in consciousness and integrating multisensory information. Despite extensive connections with almost all cortical areas, its functions and mechanisms are not well understood. In an attempt to unravel these complexities, Shelton et al. employed advanced circuit mapping technologies to examine specific neurons within the claustrum. They focused on how these neurons integrate incoming information and manage the output. Their findings suggest that claustrum neurons selectively communicate based on cortical projection targets and that their responsiveness to cortical inputs varies by cell type.

Imaging studies demonstrated that claustrum axons respond to both single and multiple sensory stimuli. Extended inhibition of the claustrum significantly reduced animals' responsiveness to multisensory stimuli, highlighting its critical role as an integrative hub in the cortex.

However, the study's conclusions at times rely on assumptions that may undermine their validity. For instance, the comparison between RSC-projecting and non-RSC-projecting neurons is problematic due to potential false negatives in the cell labeling process, which might not capture the entire neuron population projecting to a brain area. This issue casts doubt on the findings related to neuron interconnectivity and projections, suggesting that the results should be interpreted with caution. The study's approach to defining neuron types based on projection could benefit from a more critical evaluation or a broader methodological perspective.

We thank the reviewer for their attention to the methods used in our study. We acknowledge that there is an inherent bias introduced by false-negatives as a result of incomplete labeling but contend that this is true of most modern tracing experiments in neuroscience, irrespective of the method used. Moreover, if false-negative biases are affecting our results, then they likely do so in the direction of supporting our findings – perfect knowledge of claustrum connectivity would likely enhance the effects seen by increasing the pool of neurons for which we find an effect. For example, our cortico-claustal connectivity findings in Figure 3 likely would have shown even larger effects should false-negative CLARSP neurons have been positively identified.

Where appropriate we have provided estimates of variability and certainty in our experimental findings and do not claim any definitive knowledge of the true rate and scope of claustrum connectivity.

Nevertheless, the study sets the stage for many promising future research directions. Future work could particularly focus on exploring the functional and molecular differences between E1 and E2 neurons and further assess the implications of the distinct responses of excitatory and inhibitory claustrum neurons for internal computations. Additionally, adopting a different behavioral paradigm that more directly tests the integration of sensory information for purposeful behavior could also prove valuable.

We thank the reviewer for their outlook on the future directions of our work. These avenues for study, we believe, would be very fruitful in uncovering the cell-type-specific computations performed by claustrum neurons.

Recommendations for the authors:

Reviewing Editor (Recommendations for the Authors):

The editor recommends addressing the issues raised by the reviewers about the statistical significance of sensory response with respect to blank stimuli, and solving the issue generated by the exclusion of monosynaptically connected neurons in the connectivity study, to raise the assessment strength of evidence from incomplete to solid. Moreover, as the reported result stands, the behavioral task does not seem to be learned by the animals as the animals are above chance for visual and auditory but largely below chance level for multisensory. It seems that the animals do not perform a multisensory task. The authors should clarify this.

Reviewer #1 (Recommendations For The Authors):

Several references were missing from the manuscript, where mouse CLA-retrosplenial or CLA-frontal neurons were investigated and would be highly relevant to both the discussion of claustrum function and the context of the methodologies used here. (Wang et al., 2023 Nat Comm; Nair et al., 2023 PNAS, Marriott et al. 2024 Cell Reports ; Faig et al., 2024 Current

Biology).

Reviewer #2 (Recommendations For The Authors):

Let me be clear, this is an excellent study, using state-of-the-art technical approaches to probe the local/global organization of the Claustrum. However, the study is somehow disconnected, with a fantastic in-vitro part, and, in my opinion, a less convincing in-vivo one.

As stated in the public review, I'm concerned about the use of the term "integration", as, in my opinion, the data presented in this study (which I repeat are of excellent level) do not support that claim.

Below are my main points regarding the article:

(1) My main comment relates to the use of the term 'integration'. It might be a semantic debate, but I think that this is an important one. In my opinion, neural integration is the "summing of several neural input signals by a single neuron to produce an output signal that is some function of those inputs". As the authors state in the discussion, they were not able to "assess the EPSP response magnitude to the conjunction of stimuli due to photosensitivity of ChrimsonR opsins to blue light". Therefore, the authors did not specifically prove integration, but rather input convergence. This does not mean that the results presented are not important or of excellent quality, but I encourage the authors to either tone down the part on integration or to give a clear definition of what they call integration.

(2) The in vivo imaging data are somehow confusing. First, the authors image two claustral populations simultaneously (the CLA-RSP and the CLA-ACA axons). I may be missing the information, but there is no evidence that these cells overlap in the CLA (no data in the supplement and existing literature only support partial overlap). Second, in the results part, the authors claim that 96% of the sensory-responsive axons displayed multisensory response. This, combined with the 47% of axons responsive to at least one stimulus should lead to a global response of around 45% of the axons in multisensory trials. Yet, in Figures 6F-G, one can see that the response probability is actually low (closer to 20%). To be honest, I cannot really understand how to make sense of these results. At first, I thought that most of the multisensory responsive axons show no response during multisensory stimulus (but one in the unimodal stimulus). This hypothesis is however unlikely, as response AUC is biased toward positivity in Figure 6H. Overall, I'm not totally convinced by the imaging data, and I think that the authors should be more cautious about interpreting their results (as they are in the discussion part, but less in the results part).

(3) The TetTox approach used in the study ablates all neurons expressing the CRE in the CLA. If the hypothesis proposed by the authors is true, then ablating one subpopulation should not impact that much the functioning of the whole CLA, as other neurons will likely "integrate" information coming from multiple cortices (Figures 3 and 4), the local divergence (Figure 1) will then allow the broadcasting of this information back to multiples cortices. Do the authors think that such an approach deeply modified intra-claustral network connectivity? If this is not the case, shouldn't we expect less effect after lesioning a specific sub-population of CLA neurons?

(4) The behavioral protocol is also confusing. If I understand correctly, the aim of the task was to probe the D-Prime factor, as all trials, whatever the response of the animal are rewarded. From the Figure 7I, one can see that the mice cannot properly answer to the audiovisual cues, clearly indicating that both groups show impaired response to this type of trial. The whole conclusion of the authors is therefore drawn from the D-Prime calculation. However, even if D-Prime should represent a measure of sensitivity (i.e. is unaffected by response bias), two assumptions need to be met: (1) the signal and noise distributions should be both normal, and (2) the signal and noise distributions should have the same standard deviation. However, these assumptions cannot be tested in the task used by the authors (one would need rating tasks). The authors might want to use nonparametric measures of sensitivity such as A' (see Pollack and Norman 1964).

Reviewer #3 (Recommendations For The Authors):

While the study is comprehensive, some of its conclusions are based on assumptions that potentially weaken their validity. A significant issue arises in the comparison between neurons that project to the retrosplenial cortex (RSC) and those that do not. This differentiation is based on retrograde labeling from a single part of the RSC. However, CTB labeling, the technique used, does not capture 100% of the neurons projecting to a brain area. The study itself demonstrates this by showing that injecting the dye into three sections of the RSC results in three overlapping populations of neurons in the claustrum. Therefore, limiting the injection to just one of these areas inevitably leads to many false negatives-neurons that project to the RSC but are not marked by the CTB. This issue recurs in the analysis of neurons projecting to both the RSC and the prelimbic cortex (PL), where assumptions about interconnectivity are made without a thorough examination of overlap between these populations. The incomplete labeling complicates the interpretation of the data and draws firm conclusions from it.

Minor.

There is a reference to Figure 1D where claustrum->cortical connections are described. This should be 5D.

This is a correct reference pointing back to our single-cell characterizations of CLA morphoelectric types.

End of Page 22. Implies should be imply.

This has been resolved in the manuscript text.

Author response:

The following is the authors’ response to the original reviews.

Public Reviews:

Reviewer #1 (Public Review):

Summary:

This is an interesting and valuable study that uses multiple approaches to understand the role of bursting involving voltage-gated calcium channels within the mediodorsal thalamus in the sedative-hypnotic effects of alcohol. Given its unique functional roles and connectivity pattern, the idea that the mediodorsal thalamus may have a fundamental role in regulating alcohol-induced transitions in consciousness state would be both important for researchers investigating thalamocortical dynamics and more broadly interesting for understanding brain function. In addition, the author's examination of the role of the voltage-gated calcium channel Cav3.1 provides some evidence that burst-firing mediated by this channel in the thalamus is functionally important for behavioral-state transitions. While many previous studies have suggested an analogous role for sleep-state regulation, the evidence for an analogous role of this type of bursting in sedative-induced transitions is more limited. Despite the importance of these results, however, there is some concern that the manipulations and recording approaches employed by the authors may affect other thalamic nuclei adjacent to the MD, such as the central lateral nucleus, which has also been implicated in controlling state transitions. The evidence for a specific role of the mediodorsal thalamus is therefore somewhat incomplete, and so additional validation is needed.

Strengths:

This study employs multiple, complementary research approaches including behavioral assays, sh-RNAbased localized knockdown, single-unit recordings, and patterned optogenetic interventions to examine the role of activity in the mediodorsal thalamus in the sedative-hypnotic effects of alcohol. Experiments and analyses included in the manuscript generally appear well conceived and are also generally well executed. Sample sizes are sufficiently large and statistical analysis appears generally appropriate though in some cases additional quantification would be helpful. The findings presented are novel and provide some interesting insight into the role of the thalamus as well as voltage-gated calcium channels within this region in controlling behavioral state transitions induced by alcohol. In particular, the observed effects of selective knockout along with recordings in total knockout of the voltage-gated calcium channel, Cav3.1, which has previously been implicated in bursting dynamics as well as state transitions, particularly in sleep, together suggest that the transition of thalamic neurons to a bursting pattern of firing from a more constant firing is important for transition to the sedated state produced by ethanol intoxication. While previous studies have similarly implicated Cav3.1 bursting in behavioral state transitions, the direct optogenetic interventions and single-unit recordings provide valuable new insight. These findings may also have interesting implications for the relationship between sleep process disruption associated with ethanol dependence, although the authors do not appear to examine this directly or extensively discuss these implications of their findings.

Weaknesses:

A key claim of the study is that the mediodorsal thalamus is specifically important for the sedative-hypnotic effect of ethanol and that a transition to a bursting pattern of firing in this circuit facilitates these effects due to a loss of a more constant tonic firing pattern. Despite the generally clear observed effects across the included experiments, however, the evidence presented does not fully support that the mediodorsal thalamus, in particular, is involved. This distinction is important because some previous studies have suggested that another thalamic nucleus which is very close to the mediodorsal thalamus, the central-lateral thalamus, has previously been suggested to play a role in preventing sedative-induced transitions. Despite its proximity to the mediodorsal thalamus, the central-lateral thalamus has a substantially different pattern of connectivity so distinguishing which region is impacted is important for understanding the findings in the manuscript. While sh- RNA knockdown appears to be largely centered in the mediodorsal thalamus in the example shown, (Figure 2) this is rather minimal evidence and it is also not well explained (indeed, the relevant panels do not even appear to be referenced in the text of the manuscript) and the consistency of the knockdown targeting is not quantified. Additional evidence should be provided to validate this approach. Similarly, while an example is shown for the expression of ChR2 (Fig. 5) there seems to be some spread of expression outside of the mediodorsal thalamus even in his example raising a concern about how regionally specific this effect.

The recordings targeting the mediodorsal thalamus could provide evidence of a direct association between changes in activity specifically in this part of the thalamus with the behavioral measures but there are currently some issues with making this link. One difficulty is that, although lesions are shown in Figure S5 to validate recording locations, this figure is relatively unclear and the examples appear to be taken from a different anterior/posterior location compared to the reference diagram. A larger image and improved visualization of the overall set of lesion locations that includes multiple anterior/posterior coronal sections would be helpful. Moreover, even for these example images, it is difficult to evaluate whether these are in the mediodorsal thalamus, particularly given the small size of the image shown. Ideally, an example image that is more obviously in the mediodorsal thalamus would also be included. Finally, an assessment of the relationship between the approximate locations of recorded neurons across the tetrode arrays and the behavioral measures would be very helpful in supporting the unique role of the mediodorsal thalamus. The lack of these direct links, in combination with the histological issues, reduces the insight that can be gained from this study.

In addition to the key experimental issues mentioned above, there are often problems in the text of the manuscript with reasoning or at least explanation as well as numerous minor issues with editing. The most substantial such issue is the lack of clarity in discussing the mediodorsal thalamus and other adjacent thalamic nuclei, such as the central-lateral nucleus, in the author's discussion of previous findings. Given that at last one of the manuscripts cited by the authors (Saalman, Front. Sys. Neuro. 2014) has directly claimed that central-lateral, rather than the mediodorsal, thalamus is important for arousal regulation related to a conscious state, this distinction should be addressed clearly in the discussion rather than papered over by grouping multiple thalamic nuclei as being medial. As part of this discussion, it would be important to consider additional relevant literature including Bastos et al., eLife, 2021 and Redinbaugh et al., Neuron, 2020 which are quite critical but currently do not appear to be cited. Considering additional literature relevant to the function of the mediodorsal thalamus would also be beneficial. While the methods employed generally seem sound, the description in the methods section is lacking in detail and is often difficult to follow. Analysis methods such as the burst index appear to only be given a brief explanation in the text and appear not to be mentioned in the methods section. Similarly, the staining method used in Figure 2 does not appear to be described in the methods section. The most substantial case is for the UMAP approach used in Figure 4-E which does not appear to be described in the methods or even described in the main text. The lack of detailed descriptions makes it difficult to evaluate the applicability and quality of the experimental and analytical approaches. Citations justifying the use of methods such as the approach to separate regular spiking and narrow spiking neuron subtypes are also needed.

Beyond the problems with content and reasoning discussed above, there are also some relatively minor issues with the clarity of writing throughout the paper (for example, in the abstract the authors refer to "the ethanol resistance behavior in WT mice" but it is difficult to parse what they mean by this statement. Similarly, the next sentence "These results support that the maintenance..." while clearer, is not well phrased. Though individually minor, issues like this re-occur throughout the manuscript and sometimes make it difficult to follow so the text should be revised to correct them. There are also some problems with labels such as the labels of A1/A2 in Figure 4, which appear to be incorrect. Also, S7 has no label] on the B panels. Finally, some references are not included (only a label of [ref]).

Reviewer #2 (Public Review):

In the current study, Latchoumane and collaborators focus on the Cav3.1 calcium channels in the mediodorsal thalamic nucleus as critical players in the regulation of brain-states and ethanol resistance in mice. By combining behavioural, electrophysiological, and genetic techniques, they report three main findings. First, KO Cav3.1 mice exhibit resistance to ethanol-induced sedation and sustained tonic firing in thalamocortical units. Second, knocked-down Cav3.1 mice reproduce the same behaviour when the mediodorsal, but not the ventrobasal, thalamic nucleus is targeted. Third, either optogenetic or electric stimulation of the mediodorsal thalamus reduces ethanol-induced sedation in control animals.

Overall, the study is well designed and performed, correctly controlled for confounds, and properly analysed. Nonetheless, it is important to address some aspects of the report. The results support the conclusions of the study. These results are likely to be relevant in the field of systems neuroscience, as they increase the molecular evidence showing how the thalamus regulates brain states.

Reviewer #1 (Recommendations For The Authors):

Aside from the additional quantification and clarification of the analysis discussed in the weakness section, in general, the experiments included in the manuscript seem reasonable. However, I would suggest one additional experiment as well as one control, both of which are relatively straightforward optogenetic experiments, that I feel would be helpful to further improve the study. First, as the authors note, the optogenetic interventions used do not directly address the relevance of the changes in bursting patterns observed in the knockout (KO), which are by far the most robust effect, with the changes in alcohol sensitivity. One approach that could help address this would be to use patterned suppression via inhibitory opsins (e.g. halorhodopsin) to "rescue" the periods of inhibition associated with bursting in the KO. Localizing this inhibition to the mediodorsal thalamus would also lend further credence to their claim that this nuclei is the relevant circuit for their observed effects. For the control, tonic activation of the ventrobasal nucleus, as the authors did for the mediodorsal nucleus, would be beneficial to rule out the possibility that the observed effect would occur with any thalamic nucleus. In addition to these experiments, I did not note the strategy for sharing data obtained through this study so this should be added.

R1 – 1: A key claim of the study is that the mediodorsal thalamus is specifically important for the sedative-hypnotic effect of ethanol and that a transition to a bursting pattern of firing in this circuit facilitates these effects due to a loss of a more constant tonic firing pattern. Despite the generally clear observed effects across the included experiments, however, the evidence presented does not fully support that the mediodorsal thalamus, in particular, is involved. This distinction is important because some previous studies have suggested that another thalamic nucleus which is very close to the mediodorsal thalamus, the central-lateral thalamus, has previously been suggested to play a role in preventing sedative-induced transitions. Despite its proximity to the mediodorsal thalamus, the central-lateral thalamus has a substantially different pattern of connectivity so distinguishing which region is impacted is important for understanding the findings in the manuscript.

R1-A1: The reviewer is right that CL has been pointed as another candidate structure with causal influence on arousal and consciousness. We have focused our efforts in including only recording single units that were from tetrode located in the MD specifically using the lesion code we explain in the method section and in response to R1 question#3. We also produced a quantification of Cav3.1 knock-down that clearly demonstrates that the KD experiment was itself specific to MD, bilaterally, and that CL to CM were minimally impacted by the knock-down process (Fig. 2C and D). Moreover, the optogenetic  (fiber incidence was 30 degrees guaranteeing a central coverage rather than lateral; Fiber optic NA = 0.22) and electric stimulation (bipolar twisted electrodes, 50uA) experiments were also very selective and specific to the MD (Fig.S5). It remains clear that MD might not be the sole structure involved in the brain state control towards sedation and “anesthetic states”, and CL might be a significant contributor as well, however, we show that CL manipulations were rather irrelevant in our experiments  (Fig. 2, S5, S9 and S11).

R1-2: While sh-RNA knockdown appears to be largely centered in the mediodorsal thalamus in the example shown, (Figure 2) this is rather minimal evidence and it is also not well explained (indeed, the relevant panels do not even appear to be referenced in the text of the manuscript) and the consistency of the knockdown targeting is not quantified. Additional evidence should be provided to validate this approach.

R1-A2: In order to address this important question, we have created an additional panel quantification to fig2D. We have then quantified the intensity per area of Cav3.1 expression in sub zones of 4 regions of interest: MD (left, right; 2 subzones each), Centro Medial (CM; 1 subzones in total), Centrolateral/Paraventricular nucleus (CL/PCN; left, right; 2 subzones each) and the submedial nucleus (SMT; left, right; used as a control for the intensity normalization; 1 subzones in total). This panel clearly illustrates that MD was knocked-down bilaterally (p<0.001). Moreover, CM (p<0.05) and CL (p<0.01) were also partially and unilaterally knocked down, as well. This analysis confirms that our KD had a high specificity to MD.

We added the relevant figure caption and text:

[Result section, Cav3.1 silencing in the MD, but not VB, increased ethanol resistance in mice, paragraph 3]

“We then characterized the change in Cav3.1 expression following the shControl and shCav3.1 knockdown injections in three test regions MD (left and right), CM (centromedial nucleus) and CL (centrolateral nuclei, left and right side) and a negative control region SMT (submedial thalamic nuclei, left and right side). The average intensity was obtained from two coronal brain slices for each mice used in the experiment (see Methods sections, Cav3.1 Intensity quantification). Our results show that the targeting of the knockdown was very specific to the bilateral MD (p<0.001; Fig. 2D). We noted that the CM (p<0.05) and a marginal unilateral knock-down of the CL were also observed (p<0.01). Notably, we tested the correlation between the level of knock-down in MD and the total time in LOM and observed a significant association (Fig. 2D inset; R = 0.599, p = 0.018). This result highlights that the Cav3.1 knock-down was specific to MD and with an intensity associated with ethanol-induced loss of motion.”

R1-3: One difficulty is that, although lesions are shown in Figure S5 to validate recording locations, this figure is relatively unclear and the examples appear to be taken from a different anterior/posterior location compared to the reference diagram. A larger image and improved visualization of the overall set of lesion locations that includes multiple anterior/posterior coronal sections would be helpful. Moreover, even for these example images, it is difficult to evaluate whether these are in the mediodorsal thalamus, particularly given the small size of the image shown. Ideally, an example image that is more obviously in the mediodorsal thalamus would also be included. Finally, an assessment of the relationship between the approximate locations of recorded neurons across the tetrode arrays and the behavioral measures would be very helpful in supporting the unique role of the mediodorsal thalamus.

R1-A3: Related to fig.S5, we re-distributed the position of the recordings from the tetrode electrode burned positions over 3 representative coronal planes that best represent the implant positions. We also provided additional snapshots of tetrode location. To identify the positions of four tetrodes in each animal, we encoded the positions with different electrical lesion strategies as follows: 1 lesion(tetrode 1), 2 lesions while we redrew the tetrode with 100 um interval (tetrode 2), 3 lesions with 200um interval (tetrode 3), 4 lesions with 50um intervals (tetrode4). Tetrodes that were found outside of the MD delimited region were discarded post analysis. A straight relationship between the closeness of the electrode is unfortunately not possible for tetrode recording, a straight silicone probe which maintains the spatial spacing in recording would have been a better approach in that case, but unfortunately, it was not performed in our study.

R1-4: In addition to the key experimental issues mentioned above, there are often problems in the text of the manuscript with reasoning or at least explanation as well as numerous minor issues with editing. The most substantial such issue is the lack of clarity in discussing the mediodorsal thalamus and other adjacent thalamic nuclei, such as the central-lateral nucleus, in the author's discussion of previous findings. Given that at last one of the manuscripts cited by the authors (Saalman, Front. Sys. Neuro. 2014) has directly claimed that central-lateral, rather than the mediodorsal, thalamus is important for arousal regulation related to a conscious state, this distinction should be addressed clearly in the discussion rather than papered over by grouping multiple thalamic nuclei as being medial. As part of this discussion, it would be important to consider additional relevant literature including Bastos et al., eLife, 2021 and Redinbaugh et al., Neuron, 2020 which are quite critical but currently do not appear to be cited. Considering additional literature relevant to the function of the mediodorsal thalamus would also be beneficial.

R1-A4: We thank the reviewer for his comments and suggestions. We agree that the added references mentioned by the reviewers are highly relevant and should be integrated in the manuscript. We have integrated the above-mentioned references and further developed on the discussion on the role of MD relative to other thalamic nuclei (ILN and CL in particular). We believe that this better-referenced and clarified text does improve the manuscript greatly.

[introduction section, paragraph 3]

“The centrolateral (CL) thalamic nucleus has been implicated in the modulation of arousal, behavior arrest 31, and improvement of level of consciousness during seizures 32. Notably, the direct electrical stimulation of the intralaminar nuclei (ILN) and, in particular CL, promoted hallmarks of arousal and awakening in primate under propofol and ketamine propofol anesthesia.”

[Discussion section, paragraph 1]

“In this work, we identified that the neural activity in MD plays a causal role in the maintenance of consciousness. Whole body Cav3.1 KO and MD-specific Cav3.1 KD mice showed resistance to loss of consciousness induced by hypnotic dose of ethanol. In WT mice, MD neurons demonstrated a reduced firing rate in natural (sleep) and ethanol-induced unconscious states compared to awake states. This neural activity reduction was impaired in KO mice. In particular, transition to an unconscious state was accompanied with a switch of firing mode from tonic firing to burst firing in WT mice whereas this modeshift disappeared in KO mice. Finally, optogenetic or electric stimulations of the MD after ethanol injection were sufficient to induce a resistance to loss of motion, supporting that the level of neural firing in the MD is critical to maintain conscious state and delay unconscious state. We showed that the expression of Cav3.1 t-type calcium channels in MD is a cellular modulator associated with this effect.”

[Discussion section, MD is a modulator of consciousness, paragraph 2 and 3]

“The MD is known to innervate limbic region, basal ganglia and medial prefrontal cortex 50 and increased activity in MD might modulate the stability of cortical UP states (e.g. awaken, aroused and attentive states) and synchronization 9,26. Thus, MD might be a major hub involved in cortical state control and brain state stabilization.

Supporting the brain state stabilization theory and the ethanol resistance of Cav3.1 mutants, Choi et al.34 demonstrated that the loss of Cav3.1 T-type calcium channel reduced the bilateral coherence between PFC and MD under ketamine anesthesia and ethanol hypnosis, especially in the delta frequency bands. More importantly, under propofol anesthesia, Bastos et al.35 showed that intralaminar nucleus and MD stimulation lead to increased wake-up subscore and arousal, together with an increased in cortico-cortico and thalamo-cortical slow (delta) frequency power.

In the present study, we observed that MD KD (Fig. 2A), but not VB KD (Fig. S3) of Cav3.1 increased and is associated (Fig. 2D) with ethanol resistance in mice. We found that MD neurons in Cav3.1 mutant mice exhibited tonic firing within range of wakefulness (Fig. 3 and 4), indicative of resistance to ethanol and wake-like brain state. In addition, we found a strong association between the normalized tonic firing in MD and the arousal through brain states (i.e. walk to wake to sleep states), supporting that MD tonic firing could be interpreted both as a thalamic readout and a modulator of the brain state 11 (Fig. 3). Finally, direct optogenetic and electric MD stimulation increased resistance to loss of consciousness in WT mice (Fig.5 and Fig. S10). To our knowledge, this is the first report demonstrating the causal involvement of mediodorsal thalamic nucleus in the modulation of wakefulness and the resistance to ethanol-induced loss of consciousness in mice.”

R1-5: While the methods employed generally seem sound, the description in the methods section is lacking in detail and is often difficult to follow. Analysis methods such as the burst index appear to only be given a brief explanation in the text and appear not to be mentioned in the methods section.

R1-A5: We have added a clear definition in the supplementary method following the original work used:

[Supplementary Method section, Single Unit recording, sorting and analysis, last paragraph]

“The bursting index was derived as described in (Royer et al. 2012). Namely, the burst index was estimated from the spike auto-correlogram (1-ms bin size) by subtracting the mean value between 40 and 50 ms (baseline) from the peak measured between 0 and 10 ms. Positive burst amplitudes were normalized to the peak and negative amplitudes were normalized to the baseline to obtain indexes ranging from −1 to 1.” We also edited its mention in the text for clarity:

[Result section, Lack of Ca3.1 in MD neurons removes thalamic burst in NREM sleep, paragraph 2]

“[…] and a clear reduction in total bursting represented as bursting index (Fig. 3-B; ratio of spikes count <10 ms and >50 ms based on auto-cross-correlogram).”

R1-6: Similarly, the staining method used in Figure 2 does not appear to be described in the methods section.

R1-A6: The staining method can be found in the supplementary method of the paper. [supplementary method, Immunohistochemistry]

R1-7: The most substantial case is for the UMAP approach used in Figure 4-E which does not appear to be described in the methods or even described in the main text.

R1-A7: Regarding the method, the UMAP approach is described in the supplementary method document [Uniform Manifold Approximation and Projection (UMAP)]. We believe that only a succinct description was needed here considering the extent of the analysis. Regarding the inserts in the main text, we agree that the main text was lacking the description of these results and we have amended the main text to reflect a clear description of this result and what it entails. The following paragraph was added:

[Result section, Under ethanol, MD neurons lacking Cav3.1 show no burst and a wake state-like neural activity, second to last paragraph]

“Finally, we asked whether the firing modes and properties (tonic firing rate, burst firing rate; see supplementary methods) of single MD neurons would form distinct qualitative representation of “brain stages” using a lowered dimensional UMAP representation (Uniform Manifold Approximation and Projection42 ). We observed that for awake and active (i.e. walk), the brain state representation formed two adjacent clusters that confounded both wild and mutant neurons (Fig. 4E, left panel). The REM and NREM states, the wild type neurons formed 2 additional interconnected clusters, whereas the mutant neurons tend to overlap with the clusters attributed to the “awake” brain state (Fig. 4E, second to left panel). Ethanol induced fLOM, similarly to REM and NREM clusters, was distinct from awake clusters in wild type mice and overlapped with the NREM clusters (Fig. 4E, third to left panel). Here also, mutant MD neurons showed overlap with the awake clusters rather than the “low consciousness” brain states. These results indicate that the firing mode and properties could define a brain state representation that shows distinctions in levels of consciousness. Moreover, the mutant showed a representation of “low consciousness” states overlapping with wild type “awake” states consistent with the hypothesis of resistance to loss of consciousness.”

R1-8: Citations justifying the use of methods such as the approach to separate regular spiking and narrow spiking neuron subtypes are also needed.

R1-A8: We have added two references related to the observation of the two subpopulations of spiking neurons [Schiff and Reyes, 2012; Destexhe, 2008].

R1-9: Beyond the problems with content and reasoning discussed above, there are also some relatively minor issues with the clarity of writing throughout the paper (for example, in the abstract the authors refer to "the ethanol resistance behavior in WT mice" but it is difficult to parse what they mean by this statement.

R1-A9: We addressed this issue by editing and revising the manuscript for clarity and flow.

R1-10: Similarly, the next sentence "These results support the maintenance..." while clearer, is not well phrased. Though individually minor, issues like this re-occur throughout the manuscript and sometimes make it difficult to follow so the text should be revised to correct them.

R1-A10: We thank the reviewer for highlighting this point. We have edited the overall text to improve clarity and flow.

[abstract] 

These results suggest that maintaining MD neural firing at a wakeful level is sufficient to induce resistance to ethanol-induced hypnosis in WT mice.

R1-11: There are also some problems with labels such as the labels of A1/A2 in Figure 4, which appear to be incorrect.

R1-A11: We noted this issue and have rectified the figure for clarity.

R1-12: Also, S7 has no label on the B panels.

R1-A12: We thank the reviewer for pointing out this lack. We have added the y-label on the panel for clarity.

R1-13: Finally, some references are not included (only a label of [ref]).

R1-A13: We have completed the missing reference and thank the reviewer for pointing that out.

Additional comments

R1-14: Aside from the additional quantification and clarification of the analysis discussed in the weakness section, in general, the experiments included in the manuscript seem reasonable. However, I would suggest one additional experiment as well as one control, both of which are relatively straightforward optogenetic experiments, that I feel would be helpful to further improve the study. First, as the authors note, the optogenetic interventions used do not directly address the relevance of the changes in bursting patterns observed in the knockout (KO), which are by far the most robust effect, with the changes in alcohol sensitivity. One approach that could help address this would be to use patterned suppression via inhibitory opsins (e.g. halorhodopsin) to "rescue" the periods of inhibition associated with bursting in the KO.

R1-A14: Here the reviewer proposes an interesting experiment which we have attempted to perform, however, poses several technical challenges. First, the KO do not have burst firing as they are depleted from Cav3.1 low-threshold calcium channel. Therefore, under ethanol, even if there might exist a rhythmic inhibition that activates Cav3.1 channels and causes a rebound burst, the KO are unable to have it. Therefore, an optogenetic inhibition would only accentuate the total inhibition and could potentially induce an overall decrease in MD firing, resulting in an increase in LOM features. Alternatively, we showed that in a WT with low ethanol dose (where LOM induction is harder), the increased rhythmic inhibition does indeed increase significantly LOM duration and marginally decreases latency to LOM (Fig. S12), indicating that increased inhibition could indeed explain the hypothesis: “ the stronger the decrease in MD firing, the faster and longer the LOM.” The only caveat of using WT here is that optogenetic inhibition might also include rebound burst post-inhibition. Injecting bursts only did not alter the response to ethanol (Fig. S10). These results point to the role of loss of firing in MD as a main factor for LOM, and potentially the contribution of burst necessitating a concurrent inhibition/loss of firing.

We agree that inhibition in KO would further validate this hypothesis, controlling for the role of burst. We regret that we are not in the capacity to perform additional experiments involving the KO mice.

R1-15: For the control, tonic activation of the ventrobasal nucleus, as the authors did for the mediodorsal nucleus, would be beneficial to rule out the possibility that the observed effect would occur with any thalamic nucleus.

R1-A15: We agree with the reviewer that we could have added an additional region control to the gain/loss of function experiments. We would even go further as to suggest that a better control nucleus would be a high order nucleus such as PO or an unrelated sensory relay nucleus such as LGN. VB being a motor relay nucleus, could also mediate movement initiation, which could be hard to interpret. Since the complete control study for all thalamic nuclei Cav3.1 KD is outside the scope of this study, we opted not to redo these experiments and keep the focus of the manuscript on the manipulation of MD activity rather than the various available thalamic nuclei. We also do not claim that MD is the sole center able to initiate a switch in the loss of consciousness, and a more in-depth study on that matter would be clearly needed.

R1-16: In addition to these experiments, I did not note the strategy for sharing data obtained through this study so this should be added.

R1-A16: We have uploaded data and code for most figures at the following repository and provided a clearer statement regarding data sharing. We thank the reviewer for pointing out this missing element.

The link for the repository is the following:

It contains:

- Excel spreadsheet file of all behavior values, including the newly quantified Cv3.1 expression in MD/CL/SMT

- Excel spreadsheet follow-up of all MD cells (single unit; tetrode) analyzed

- Folders for all groups studied with representative figures showing EEG power over time and normalized activity (WT vs KO for 2, 3 and 4 g/kg; MDshKD vs shCTR, VBshKD vs shCTR; CHR2 NOSTIM vs STIM; ESTIM Groups and ARCH NOSTIM vs STIM)

- A1G LORRvsLOM and OPEN FIELD Matlab data

- Matlab and ImageJ Codes: single unit analysis, characterization, brain state characterization, sleep stages, LOM, open field analysis and statistical analysis.

We have added the data sharing subsection in the acknowledgements:

“Part of the analyzed data and codes are available on the open access platform, mendeley:

Latchoumane, Charles-francois (2024), “Mediodorsal thalamic nucleus mediates resistance to ethanol through Cav3.1 T-type Ca2+ regulation of neural activity”, Mendeley Data, V1, doi: 10.17632/7fr427426m.1

Additional data (large size recording and images) can be provided upon reasonable requests.”

Reviewer #2 (Recommendations For The Authors):

R2-1. Consciousness is a contentious subject. Even in humans, there is still intense research on the topic, not to mention animals, about which we still know very little. Moreover, consciousness is not quantified in this study, as there is no standard metric to do so. Accordingly, talking about 'modulation', 'transition', ́level ', or 'reduction' of consciousness can be misleading. Hence, it is probably safer to strictly refer to brain-states and/or stages of the sleep-wake cycle in this study and reframe it entirely around these concepts.

R2-A1. The reviewer points to an important point and we appreciate this highlight. Agreeing that the definition of consciousness is rather loose and arguably difficult to pinpoint. Here, we settle on a definition that relies on the loss of motion and loss of righting reflex. This definition is widely accepted as the “verified” state in which the absence of responsiveness (to continuous stimuli, inducing reflex or discomfort) is observed and uninterrupted by jerks and spurious movements. Additional metrics needed would be the recording of EMG to quantify atonia and EEG to the settling of a dominantly slow-wave frequency (~4 Hz; ethanol-induced sedation at theta rhythm), as shown in Fig S1A. The driver of this 4Hz frequency and its correlation has been investigated previously (e.g. Choi et al, PNAS, 2012), leading to the accepted link between LOM/LORR and loss of consciousness. Our data present the advantage of showing single neuron recordings and that LOM is a state where the lowest firing activity is present (Fig S7AB) and comparable to deep sleep state activity (Fig3D). The first LOM is the most important as it highlights the deepest loss of consciousness before the ethanol starts to be metabolized and cleared, which would be consistent between animals.

As a result, we have edited the manuscript to clarify all mentions related to brain states and states of unconsciousness.

R2-2. It is not clear why the authors focus on the mediodorsal nucleus. This should be better explained in the introduction and developed in the discussion.

R2-A2. This comment converges with the Reviewer 1 comments and we are addressing this lack in the discussion as suggested. We have addressed it with this previous comment and believe it is now clearer.

R2-3. The discussion mentions that 'increased activity in MD might modulate the stability of cortical UP state and synchronization' (pg 21). This point should be either further developed and put into context, or removed. In its current state, it does not seem to contribute much to the discussion of results.

R2-A3. We understand that the working “UP state” might not be clear enough. We have modified this sentences as follows to clarify that UP state could be either a state of where the animal is awake, aroused or attentive:

[Discussion section, MD is a modulator of consciousness, first paragraph]

“The MD is known to innervate limbic region, basal ganglia and medial prefrontal cortex 50 and increased activity in MD might modulate the stability of cortical UP states (e.g. awaken, aroused and attentive states) and synchronization 9,26. Thus, MD might be a major hub involved in cortical state control and brain state stabilization.“

R2-4. The discussion states that 'mutant mice did not exhibit a decreased arousal level (i.e. increased locomotor activity)' (pg 23). This is confusing as decreased arousal should be reflected in decreased locomotor activity.

R2-A4. We understand that the formulation of this sentence may be confusing and we have edited this portion of the text to improve quality in the revised version of the manuscript. To clarify, mutant mice do not exhibit reduced or increased arousal (not quantified, just observational), they do have a phenotypic hyperlocomotion. This comes in contrast with a lower basal firing rate in the MD, which in our interpretation, is not synonymous with lower arousal. We believe that the relative change in MD determines the change in arousal, and that the absolute firing is not indicative of arousal in itself, only in comparison.

[Discussion section, The lower variability in MD Firing reflects Ethanol Resistance in Cav3.1 mutant mice, paragraph 2]

“Mutant RS neurons in MD showed an overall lower excitability and variability of firing in various natural conscious and unconscious states compared to wild type mice. Remarkably, Cav3.1 mutant mice exhibited a clear increased locomotor activity and an increased resistance to ethanol. The general lower firing rate and the high “arousal” observed in mutant mice suggests that the relative change from state to state in tonic firing in MD, and not the absolute value of firing, might be a better correlate of change in brain state in the mice.”

R2-5. The methods (pg 27) state that two genetic backgrounds (129/svjae and C57BL/6J ) were used in the study. Authors should show whether there were significant differences between those backgrounds in the key parameters assessed in the study (particularly resistance to ethanol sedation).

R2-A5. As mentioned in the method section, we only used the F1-background mice, which are the firstgeneration offspring produced by crossing 129/svjae and C57BL/6J strains. To produce F1 KO mice, we kept the heterozygote mice in two strains. We unfortunately did not study the particular difference of the respective KO of these two backgrounds; however, the pure C57BL/6J KO has been used in other studies by our group (Kim et al 2001; Na et al, 2008; Park et al., 2010). The F1 background allows us to work with mice that are less aggressive and can be handled with less inherent stress.

R2-6. It would be convenient to produce a supplementary figure associated with Figure 1C to show the same data with averages per mouse. That is, 9 points for control and 9 points for KO mice. This also applies to all cases where data is not presented per mouse but pooled between animals.

R2-A6. We have added a panel C in Figure S1, to show the scatter values for all the mice corresponding to the figure 1C. We have also generalized this presentation for all behavior graphics showing all the animals in the scatter plot next to the boxplot. We believe that this presentation increases further the transparency of the manuscript. We have then added the scatter plot for all mice in figure Fig1, Fig2, Fig5, Fig.S2, Fig.S3, Fig.S10 and Fig.S12.

R2-7. It would be informative to make a supplementary figure associated with Figure 1D to compare baseline raw activity levels (i.e., baseline walking recording) between control and KO mice. That is, do KO and control mice cover comparable distances and at similar speeds during baseline conditions? Figure 1D and Figure 4A suggest that the variability of locomotor activity is larger in KO mice. Hence, this parameter should be quantified and reported.

R2-A7. We thank the reviewer for this comment. We strived to answer to this question in the manuscript in two ways:

- We first measure the overall hyperlocomotion of the mice using the open field total distance parkoured in our mice cohorts (FigS4C). We did observe that the KO mutant showed hyperlocomotion, but not MD or VB knock-down mice. Which indicates that the hyperlocomotion component is not specific to the two thalamic nuclei studied.

- Using the forced walking task, we impose on the animal to keep a steady pace of roughly 6cm/s. This assay allows to normalize the general walking behavior to a relatively fixed pace making it comparable for all animals.

The reviewer suggested reporting the mean and variance in walking of WT and KO during baseline (prior to the ethanol I.P. injection). We believe that the two points mentioned above are sufficient to describe in a more quantitative way the WT vs KO locomotion differences. Moreover, by construction the normalized locomotion on the forced walking task will return similar means for the baseline, the standard deviation would, however, potentially show differences but would remain inconclusive.

R2-8. The legend in Figure 1 states that 'the loss of consciousness is evaluated using normalized moving index using either video analysis (differential pixel motion), on- head accelerometer-based motion, or neck electromyograms'. Authors should clarify whether these methods are equivalent and support it with data.

R2-A8. We understand the reviewer point and we have made a few modifications to the method description aligning better with what was done. For most mice, video analysis was used to obtain the moving index. When video recording was not available (2 mice), we had an accelerometer attached to the animal’s head stage which helped us derive a moving index that was similar to the video moving index. The neck electromyogram was rather used for animals implanted with the tetrodes to identify sleep stages based on local field potential frequency and muscle tone.  We have then clarified the method for this matter and Figure 1 to avoid this confusion. Since no concurrent recording of both video and accelerometer was performed, we do not have the data to compute the correlation between the two measures, however, no noticeable deviation from loss of motion was observed between the two methods. We realize that this may be a weak argument, however, our observations showed that video and accelerometers returned very similar timings for loss of motion (only a few comparative instances insufficient to present a statistical comparison).

R2-9. How were spike bursts defined? The authors should try different criteria and verify the consistency of results.

R2-A9 For in vivo single unit recording, we opted for a definition that is validated from our works and others as a silencing of at least 100 ms followed by a minimum of 3 spikes with:

- First spike pairs interspike interval less than 4 ms

- Remaining spike pairs interspike interval less than 20 ms

We have performed this analysis using a minimum of 2 spikes, and varied silencing periods between 50 and 100ms, without observing significant deviation of the results. As shown in Figure S6B, with this approach we observed that the burst distribution had a majority with <10 spikes per burst. Figure S6C indicated that with a clear distribution of ISI for first spike within 2-4ms as observed in previous works (Desai and Varela, 2021; Alitto et al, 2019), importantly, not clearly capped at 4 ms, showing that the range for the first ISI might indeed be lower than 4ms for thalamic burst. Within burst spike waveforms can become very variable and the choice of 3 over 2 spikes minimum per burst stems from the aim to reduce false positive detection of ultra-short bursts, which in single unit recording remains controversial (Gray et al. 1995).

Minor:

R2-10: Figure 4A2 'Cav3.1(+/+)' should presumably be Cav3.1(-/-).

R2-A10: this is correct and we have corrected the figure label [This sentence is ambiguous. What is ‘this’ that is correct?]

R2-11: Figure S2C legend states 'Post-hoc group comparison was performed using.' The sentence seems to be incomplete.

R2-A11: We have completed the sentence for clarity.

R2-12: In the methods (pg 29) virus concentration is reported as '107 TU/ul', which probably refers to 10e7.

R2-A12: We have corrected it by superscripting the power 7.

R2-13: Verify Fig 1C1 and correct Y-axis overlap between title and units.

R2-A13: We edited the figure for clarity, thank you.

R2-14: On page 24 there is a '[ref]' that probably stands for (a missing) reference.

R2-A14: the missing reference has been added.

Author response:

We are glad that the reviewers found our work to be interesting and appreciate its contribution to enhancing ecological validity of attention research. We also agree that much more work is needed to solidify this approach, and that some of the results should be considered “exploratory” at this point, but appreciate the recognition of the novelty and scientific potential of the approach introduced here.

We will address the reviewers’ specific comments in a revised version of the paper, and highlight the main points here:

· We agree that the use of multiple different neurophysiological measures is both an advantage and a disadvantage, and that the abundance of results can make it difficult to tell a “simple” story. In our revision, we will make an effort to clarify what (in our opinion) are the most important results and provide readers with a more cohesive narrative.

· Important additional discussion points raised by the reviewers, which will be discussed in a revised version are a) the similarities and differences between virtual and real classrooms; b) the utility of the methods and data to the community and c) the implication of these results for educational neuroscience and ADHD research.

· In the revision, we will also clarify several methodological aspects of the data analysis, as per the reviewers’ requests.

· After final publication, the data will be made available for other researchers to use.

Author response:

Reviewer #1 (Public review):

Weaknesses:

(1) The heatmaps (for example, Figure 3A, B) are challenging to read and interpret due to their size. Is there a way to alter the visualization to improve interpretability? Perhaps coloring the heatmap by general anatomical region could help? We feel that these heatmaps are critical to the utility of the registration strategy, and hence, clear visualization is necessary.

We thank the reviewers for this point on aesthetic improvement, and we agree that clearer visualization of our correlation heatmaps is important. To address this point, we have incorporated the capability of grouping “child” subregions in anatomical order by their more general “parent” region into the package function, plot_correlation_heatmaps(). Parent regions will be visually represented as smaller sub-facets in the heatmaps, and we will be submitting our full revised manuscript with these visual changes.

(2) Additional context in the Introduction on the use of immediate early genes to label ensembles of neurons that are specifically activated during the various behavioral manipulations would enable the manuscript and methodology to be better appreciated by a broad audience.

We thank the reviewers for this suggestion and will be revising parts of our Introduction to reflect the broader use and appeal of immediate early genes (IEGs) for studying neural changes underlying behavior.

(3) The authors mention that their segmentation strategies are optimized for the particular staining pattern exhibited by each reporter and demonstrate that the manually annotated cell counts match the automated analysis. They mention that alternative strategies are compatible, but don't show this data.

We thank the reviewers for this comment. We also appreciate that integration with alternative strategies is a major point of interest to readers, given that others may be interested in compatibility with our analysis and software package, rather than completely revising their own pre-existing workflows.

This specific point on segmentation refers to the import_segmentation_custom()function in the package. As there is currently not a standard cell segmentation export format adopted by the field, this function still requires some data wrangling into an import format saved as a .txt file. However, we chose not to visually demonstrate this capability in the paper for a few reasons.

i. A figure showing the broad testing of many different segmentation algorithms, (e.g., Cellpose, Vaa3d, Trainable Weka Segmentation) would better demonstrate the efficacy of segmentation of these alternative approaches, which have already been well-documented. However, demonstrating importation compatibility is more of a demonstration of API interface, which is better shown in website documentation and tutorial notebooks.

ii. Additionally, showing importation with one well-established segmentation approach is still a demonstration of a single use case. There would be a major burden-of-proof in establishing importation compatibility with all potential alternative platforms, their specific export formats, which may be slightly different depending on post-processing choices, and the needs of the experimenters (e.g., exporting one vs many channels, having different naming conventions, having different export formats). For example, output from Cellpose can take the form of a NumPy file (_seg.npy file), a .png, or Native ImageJ ROI archive output, and users can have chosen up to four channels. Until the field adopts a standardized file format, one flexible enough to account for all the variables of experimental interest, we currently believe it is more efficient to advise external groups on how to transform their specific data to be compatible with our generic import function.

Internally, in collaborative efforts, we have validated the ability to import datasets generated from completely different workflows for segmentation and registration. We intend on releasing this documentation in coming updates on our package website, which we believe will be more demonstrative on how to take advantage of our analysis package, without adopting our entire workflow.

(4) The authors provided highly detailed information for their segmentation strategy, but the same level of detail was not provided for the registration algorithms. Additional details would help users achieve optimal alignment.

We apologize for this lack of detail. The registration strategy depends upon the WholeBrain package for registration to the Allen Mouse Common Coordinate Framework. While this strategy has been published and documented elsewhere, we will be revising our methods to better incorporate details of this approach.

Reviewer #2 (Public review):

Weaknesses:

(1) While I was able to install the SMARTR package, after trying for the better part of one hour, I could not install the "mjin1812/wholebrain" R package as instructed in OSF. I also could not find a function to load an example dataset to easily test SMARTR. So, unfortunately, I was unable to test out any of the packages for myself. Along with the currently broken "tractatus/wholebrain" package, this is a good example of why I would strongly encourage the authors to publish SMARTR on either Bioconductor or CRAN in the future. The high standards set by Bioc/CRAN will ensure that SMARTR is able to be easily installed and used across major operating systems for the long term.

We thank reviewers for pointing out this weakness; long-term maintenance of this package is certainly a mutual goal. Loading an .RDATA file is accomplished by either double-clicking directly on the file in a directory window, or by using the load() function, (e.g., load("directory/example.RData")). We will explicitly outline these directions in the online documentation and in our full revision.

Moreover, we will submit our package to CRAN. Currently, SMARTR is not dependent on the WholeBrain package, which remains optional for the registration portion of our workflow. Ultimately, this independence will allow us to maintain the analysis and visualization portion of the package independently, and allow for submission to a more centralized software repository such as CRAN.

(2) The package is quite large (several thousand lines include comments and space). While impressive, this does inherently make the package more difficult to maintain - and the authors currently have not included any unit tests. The authors should add unit tests to cover a large percentage of the package to ensure code stability.

We appreciate this feedback and will add unit testing to improve the reliability of our package in the full revision.

(3) Why do the authors choose to perform image segmentation outside of the SMARTR package using ImageJ macros? Leading segmentation algorithms such as CellPose and StarMap have well-documented APIs that would be easy to wrap in R. They would likely be faster as well. As noted in the discussion, making SMARTR a one-stop shop for multi-ensemble analyses would be more appealing to a user.

We appreciate this feedback. We believe parts of our response to Reviewer 1, comment 3, are relevant to this point. Interfaces for CellPose and ClusterMap (which processes in situ transcriptomic approaches like STARmap) are both in python, and currently there are ways to call python from within R (https://rstudio.github.io/reticulate/index.html). We will certainly explore incorporating these APIs from R. However, we would anticipate this capability is more similar to “translation” between programming languages, but would not currently preclude users from the issue of still needing some familiarity with the capabilities of these python packages, and thus with python syntax.

(4) Given the small number of observations for correlation analyses (n=6 per group), Pearson correlations would be highly susceptible to outliers. The authors chose to deal with potential outliers by dropping any subject per region that was> 2 SDs from the group mean. Another way to get at this would be using Spearman correlation. How do these analyses change if you use Spearman correlation instead of Pearson? It would be a valuable addition for the author to include Spearman correlations as an option in SMARTR.

We thank reviewers for this suggestion and will provide a supplementary analysis of our results using Spearman correlations.

(5) I see the authors have incorporated the ability to adjust p-values in many of the analysis functions (and recommend the BH procedure) but did not use adjusted p-values for any of the analyses in the manuscript. Why is this? This is particularly relevant for the differential correlation analyses between groups (Figures 3P and 4P). Based on the un-adjusted p-values, I assume few if any data points will still be significant after adjusting. While it's logical to highlight the regional correlations that strongly change between groups, the authors should caution which correlations are "significant" without adjusting for multiple comparisons. As this package now makes this analysis easily usable for all researchers, the authors should also provide better explanations for when and why to use adjusted p-values in the online documentation for new users.

We appreciate the feedback and will more explicitly outline that in our paper, our dataset is presented as a more demonstrative and exploratory resource for readers and, as such, we accept a high tolerance for false positives, while decreasing risk of missing possible interesting findings. As noted by Reviewer #2, it is still “logical to highlight the regional correlations that strongly change between groups.” We will further clarify in our methods that we chose to present uncorrected p-values when speaking of significance. We will also include more statistical detail on our online documentation regarding FDR correction. Ultimately, the decision to correct for multiple comparisons and FDR choice of threshold, should still be informed by standard statistical theory and user-defined tolerance for inclusion of false-positives and missing of false-negatives. This will be influenced by factors, such as the nature and purpose of the study, and quality of the dataset.  

(6) The package was developed in R3.6.3. This is several years and one major version behind the current R version (4.4.3). Have the authors tested if this package runs on modern R versions? If not, this could be a significant hurdle for potential users.

We thank reviewers for pointing out concerns regarding versioning. Analysis and visualization capabilities are currently supported using R version 4.1+. The recommendation for R 3.6.3 is primarily for users interested in using the full workflow, which requires installation of the WholeBrain package. We anticipate supporting of visualization and network analysis capabilities with updated packages and R versions, and maintaining a legacy version for the full workflow presented in this paper.

Author response:

The following is the authors’ response to the previous reviews.

Public Reviews:

Reviewer #1 (Public Review):

Summary:

The authors present a new application of the high-content image-based morphological profiling Cell Painting (CP) to single cell type classification in mixed heterogeneous induced pluripotent stem cellderived mixed neural cultures. Machine learning models were trained to classify single cell types according to either "engineered" features derived from the image or from the raw CP multiplexed image. The authors systematically evaluated experimental (e.g., cell density, cell types, fluorescent channels) and computational (e.g., different models, different cell regions) parameters and convincingly demonstrated that focusing on the nucleus and its surroundings contains sufficient information for robust and accurate cell type classification. Models that were trained on mono-cultures (i.e., containing a single cell type) could generalize for cell type prediction in mixed co-cultures, and describe intermediate states of the maturation process of iPSC-derived neural progenitors to differentiation neurons.

Strengths:

Automatically identifying single-cell types in heterogeneous mixed-cell populations holds great promise to characterize mixed-cell populations and to discover new rules of spatial organization and cell-cell communication. Although the current manuscript focuses on the application of quality control of iPSC cultures, the same approach can be extended to a wealth of other applications including an in-depth study of the spatial context. The simple and high-content assay democratizes use and enables adoption by other labs.

The manuscript is supported by comprehensive experimental and computational validations that raise the bar beyond the current state of the art in the field of high-content phenotyping and make this manuscript especially compelling. These include (i) Explicitly assessing replication biases (batch effects); (ii) Direct comparison of feature-based (a la cell profiling) versus deep-learning-based classification (which is not trivial/obvious for the application of cell profiling); (iii) Systematic assessment of the contribution of each fluorescent channel; (iv) Evaluation of cell-density dependency; (v) Explicit examination of mistakes in classification; (vi) Evaluating the performance of different spatial contexts around the cell/nucleus; (vii) Generalization of models trained on cultures containing a single cell type (mono-cultures) to mixed co-cultures; (viii) Application to multiple classification tasks.

I especially liked the generalization of classification from mono- to co-cultures (Figure 4C), and quantitatively following the gradual transition from NPC to Neurons (Figure 5H).

The manuscript is well-written and easy tofollow.

Thank you for the positive appreciation of our work and constructive comments. 

Weaknesses:

I am not certain how useful/important the specific application demonstrated in this study is (quality control of iPSC cultures), this could be better explained in the manuscript. 

To clarify the importance we have added an additional explanation to the introduction (page 3) and also come back to it in the discussion (page 17).

Text from the introduction:

“However, genetic drift, clonal and patient heterogeneity cause variability in reprogramming and differentiation efficiency10,11. The differentiation outcome is further strongly influenced by variations in protocol12. This can significantly impact experimental outcomes, leading to inconsistent and potentially misleading results and consequently, it hinders the use of iPSC-derived cell systems in systematic drug screening or cell therapy pipelines. This is particularly true for iPSC-derived neural cultures, as their composition, purity and maturity directly affect gene expression and functional activity, which is essential for modelling neurological conditions13,14. Thus, from a preclinical perspective, there is the need for a fast and cost-effective QC approach to increase experimental reproducibility and cell type specificity15. From a clinical perspective in turn, robust QC is required for safety and regulatory compliance (e.g., for cell therapeutic solutions). This need for improved standardization and QC is underscored by large-scale collaborative efforts such as the International Stem Cell Banking Initiative16, which focusses on clinical quality attributes and provides recommendations for iPSC validation testing for use as cellular therapeutics, or the CorEuStem network, aiming to harmonize iPSC practices across core facilities in Europe.”

Text from the discussion: 

“Many groups highlight the difficulty of reproducible neural differentiation and attribute this to culture conditions, cultivation time and variation in developmental signalling pathways in the source iPSC material43,44. Spontaneous neural differentiation has previously been shown to require approximately 80 days before mature neurons arise that can fire action potentials and show neural circuit formation. Although these differentiation processes display a stereotypical temporal sequence34, the exact timing and duration might vary. This variation negatively affects the statistical power when testing drug interventions and thus prohibits the application of iPSC-culture derivatives in routine drug screening. Current solutions (e.g., immunocytochemistry, flow cytometry, …) are often cost-ineffective, tedious, and incompatible with longitudinal/multimodal interrogation. CP is a much more cost-effective solution and ideally suited for this purpose. Routine CP-based could add confidence to and save costs for the drug discovery pipeline. We have shown that CP can be leveraged to capture the morphological changes associated with neural differentiation.”

Another issue that I feel should be discussed more explicitly is how far can this application go - how sensitively can the combination of cell painting and machine learning discriminate between cell types that are more subtly morphologically different from one another?

Thank you for this interesting question. The fact that an approach based on a subregion not encompassing the whole cell (the “nucleocentric” approach) can predict cell types equally well, suggests that the cell shape as such is not the defining factor for accurate cell type profiling. And, while clearly neural progenitors, neurons or glia have vastly different cell shapes. We have shown that cells with closer phenotypes such as 1321N1 vs. SH-SY5Y or astrocytes vs. microglia can be distinguished with equal performance. However, triggered by the reviewers’ question, we have now tested additional conditions with more subtle phenotypes, including the classification of 1321N1 vs. two related retinal pigment epithelial cells with much more similar morphology (ARPE and RPE1 cells). We found that the CNN could discriminate these cells equally well and have added the results on page 8 and in Fig. 3D. To address this question from a different angle, we have also performed an experiment in which we changed cell states to assess whether discriminatory power remains high. Concretely, we exposed co-cultures of neurons and microglia to LPS to trigger microglial activation (more subtly visible as cytoskeletal changes and vacuole formation). This revealed that our approach still discriminates both cell types (neurons vs. microglia) with high accuracy, regardless of the microglial state. Furthermore, using a two-step approach, we could also distinguish LPS-treated (assumed to be activated) from unchallenged microglia (assumed to be more homeostatic), albeit with a lower accuracy. This experiment has been added as an extra results section (Cell type identification can be applied to mixed iPSC-derived neuronal cultures regardless of activation state, p12) and Fig. 7c. Finally, we have also added our take on what the possibilities could be for future applications in even more complex contexts such as tissue slice, 3D and live cell applications (page 17-18). 

Regarding evaluations, the use of accuracy, which is a measure that can be biased by class imbalance, is not the most appropriate measurement in my opinion. The confusion matrices are a great help, but I would recommend using a measurement that is less sensitive for class imbalance for cell-type classification performance evaluations.  

Across all CNNs trained in this manuscript, the sample size of the input classes has always been equalized, ruling out any effects of class imbalance. Nevertheless, to follow the reviewers’ recommendation, we have now used the F-score to document performance as it is insensitive to such imbalance. For clarity, we have now also mentioned the input number (ROIs/class) in every figure.

Another issue is that the performance evaluation is calculated on a subset of the full cell population - after exclusion/filtering. Could there be a bias toward specific cell types in the exclusion criteria? How would it affect our ability to measure the cell type composition of the population?

As explained in the M&M section, filtering was performed based on three criteria:

(1) Nuclear size: values below a threshold of 160, objects are considered to represent debris;

(2) DAPI intensity: values below a threshold of 500 represent segmentation errors;

(3) IF staining intensity: gates were set onto the intensity of the fluorescent markers used with posthoc IF to only retain cells that are unequivocally positive for either marker and to avoid inclusion of double positive (or negative) cells in the ground truth training. 

One could argue that the last criterion introduces a certain bias in that it does not consider part of the cell population. However, this is also not the purpose of our pioneering study that aims at identifying unique cell types for which ground truth is as pure and reliable as possible. Not filtering out these cells with a ‘dubious’ IF profile (e.g., cells that might be transitioning or are of a different type) would negatively affect the model by introducing noise. It is correct that the predictions are based only on these inputs and so cells of a subsequent test set will only be classified according to these labels. For example, in the neuronal differentiation experiment (Fig. 6G-H), cells are either characterized as NPC or as neurons, which leaves the transitioning (or undefined) cells in either category. Despite this simplification, the model adequately predicted the increase in neuron/NPC ratio with culture age. In future iterations, one could envision defining more refined cell (sub-)types in a population based on richer post-hoc information (e.g., through cyclic immunofluorescence or spatial single cell transcriptomics) or longitudinal follow-up of cell-state transitions using live imaging. This notion has been added to page 17 of the manuscript.

I am not entirely convinced by the arguments regarding the superiority of the nucleocentric vs. the nuclear representations. Could it be that this improvement is due to not being sensitive/ influenced by nucleus segmentation errors?

The reviewer has a valid point that segmentation errors may occur. However, the algorithm we have used (Stardist classifier), is very robust to nuclear segmentation errors. To verify the performance, we have now quantified segmentation errors in 20 images for 3 different densities and found a consistently low error rate (0.6 -1.6%) without correlation to the culture density. Moreover, these errors include partial imperfections (e.g., a missed protrusion or bleb) as well as over- (one nucleus detected as more) or under- (more nuclei detected as one) segmentations. The latter two will affect both the nuclear and nucleocentric predictions and should thus not affect the prediction performance. In the case of imperfect segmentations, there may be a specific impact on the nucleus-based predictions (which rely on blanking the non-nuclear part), but this alone cannot explain the significantly higher gain in accuracy for nucleocentric predictions (>5%). Therefore, we conclude that segmentation errors may contribute in part, but not exclusively, to the overall improved performance of nucleocentric input models. We have added this notion in the discussion (pages 14-15 and Suppl. Fig. 1E).

GRADCAM shows cherry-picked examples and is not very convincing.

To help convince the reviewer and illustrate the representativeness of selected images, we have now randomly selected for each condition and density 10 images (using random seeds to avoid cherrypicking) and added these in a Suppl. Fig. 3.

There are many missing details in the figure panels, figure legend, and text that would help the reader to better appreciate some of the technical details, see details in the section on recommendations for the authors.

Please see further for our specific adaptations.

Reviewer #2 (Public Review):

This study uses an AI-based image analysis approach to classify different cell types in cultures of different densities. The authors could demonstrate the superiority of the CNN strategy used with nucleocentric cell profiling approach for a variety of cell types classification. The paper is very clear and well-written. I just have a couple of minor suggestions and clarifications needed for the reader.

The entire prediction model is based on image analysis. Could the authors discuss the minimal spatial resolution of images required to allow a good prediction? Along the same line, it would be interesting to the reader to know which metrics related to image quality (e.g. signal to noise ratio) allow a good accuracy of the prediction.

Thank you for the positive and relevant feedback.

The reviewer has a good point that it is important to portray the imaging conditions that are required for accurate predictions. To investigate this further we have performed additional experiments that give a better view on the operating window in terms of resolution and SNR (manuscript page 7-8 and new figure panels Fig. 3B-C). The initial image resolution was 0.325 µm/pixel. To understand the dependency on resolution we performed training and classifications for image data sets that were progressively binned. We found that a two-fold reduction in resolution did not significantly affect the F-score, but further degradation decreased the performance. At a resolution of 6,0 µm/pixel (20-fold binning), the F-score dropped to 0.79±0.02, comparable to the performance when only the DAPI (nuclear) channel was used as input. The effect of reduced image quality was assessed in a similar manner, by iteratively adding more Gaussian noise to the image. We found that above an SNR of 10 the prediction performance remains consistent but below it starts to degrade. While this exercise provides a first impression of the current confines of our method, we do believe it is plausible that its performance can be extended to even lower-quality images for example by using image restoration algorithms. We have added this notion in the discussion (page 14).

The authors show that nucleocentric-based cell feature extraction is superior to feeding the CNN-based model for cell type prediction. Could they discuss what is the optimal size and shape of this ROI to ensure a good prediction? What if, for example, you increase or decrease the size of the ROI by a certain number of pixels?

To identify the optimal input, we varied the size of the square region around the nuclear centroid from 0.6 to 150 µm for the whole dataset. Within the nuclear-to-cell window (12µm- 30µm) the average Fscore is limited, but an important observation is the increasing error and differences in precision and recall with increasing nucleocentric patch sizes, which will become detrimental in cases of class imbalance. The F-score is maximal for a box of 12-18µm surrounding the nuclear centroid. In this “sweet spot”, the precision and recall are also in balance. Therefore, we have selected this region for the actual density comparison experiment. We have added our results to the manuscript (page 9 and 15).

It would be interesting for the reader to know the number of ROI used to feed each model and know the minimal amount of data necessary to reach a high level of accuracy in the predictions.

The figures have now been adjusted so that the number of ROIs used as input to feed the model are listed. The minimal number of ROIs required to obtain high level accuracy is tested in Figure 2C. By systematically increasing the number of input ROIs for both RF and CNN, we found that a plateau is reached at 5000 input ROIs (per class) for optimal prediction performance. This is also documented in the results section page 6.

From Figure 1 to Figure 4 the author shows that CNN based approach is efficient in distinguishing 1321N1 vs SH-SY5Y cell lines. The last two figures are dedicated to showing 2 different applications of the techniques: identification of different stages of neuronal differentiation (Figure 5) and different cell types (neurons, microglia, and astrocytes) in Figure 6. It would be interesting, for these 2 two cases as well, to assess the superiority of the CNN-based approach compared to the more classical Random Forest classification. This would reinforce the universal value of the method proposed.

To meet the reviewer’s request, we have now also compared CNN to RF for the classification of cells in iPSC-derived models (Figures 6 and 7). As expected, the CNN performed better in both cases. We have now added these results in Fig. 6 D and 7 C and pages 12 and 13 of the manuscript.

Reviewer #3 (Public Review):

Induced pluripotent stem cells, or iPSCs, are cells that scientists can push to become new, more mature cell types like neurons. iPSCs have a high potential to transform how scientists study disease by combining precision medicine gene editing with processes known as high-content imaging and drug screening. However, there are many challenges that must be overcome to realize this overall goal. The authors of this paper solve one of these challenges: predicting cell types that might result from potentially inefficient and unpredictable differentiation protocols. These predictions can then help optimize protocols.

The authors train advanced computational algorithms to predict single-cell types directly from microscopy images. The authors also test their approach in a variety of scenarios that one may encounter in the lab, including when cells divide quickly and crowd each other in a plate. Importantly, the authors suggest that providing their algorithms with just the right amount of information beyond the cells' nuclei is the best approach to overcome issues with cell crowding.

The work provides many well-controlled experiments to support the authors' conclusions. However, there are two primary concerns: (1) The model may be relying too heavily on the background and thus technical artifacts (instead of the cells) for making CNN-based predictions, and (2) the conclusion that their nucleocentric approach (including a small area beyond the nucleus) is not well supported, and may just be better by random chance. If the authors were to address these two concerns (through additional experimentation), then the work may influence how the field performs cell profiling in the future.

Thank you very much for confirming the potential value of our work and raising these relevant items. To better support our claims we have now performed additional validations, which we detail below. 

(1) The model may be relying too heavily on the background and thus technical artifacts (instead of the cells) for making CNN-based predictions 

To address the first point, we have adapted the GradCAM images to show an overlay of the input crop and GradCAM heatmap to give a better view of the structures that are highlighted by the CNN. We further investigated the influence of the background on the prediction performance. Our finding that a CNN trained on a monoculture retains a relatively high performance on cocultures implies that the CNN uses the salient characteristics of a cell to recognize it in more complex heterogeneous environments. Assuming that the background can vary between experiments, the prediction of a pretrained CNN on a new dataset indicates that cellular characteristics are used for robust prediction.  When inspecting GradCAM images obtained from the nucleocentric CNN approaches (now added in Suppl. Fig. 3), we noticed that the nuclear periphery typically contributed the most (but not exclusively) to the prediction performance. When using only the nuclear region as input, GradCAMs were more strongly (but again not exclusively) directed to the background surrounding the nuclei. To train the latter CNN, we had cropped nuclei and set the background to a value of zero. To rule out that this could have introduced a bias, we have now performed the exact same training and classification, but setting the background to random noise instead (Suppl. Fig. 2). While this effectively diverted the attention of the GradCAM output to the nucleus instead of the background, the prediction performance was unaltered. We therefore assume that irrespective of the background, when using nuclear crops as input, the CNN is dominated by features that describe nuclear size. We observe that nuclear size is significantly different in both cell types (although intranuclear features also still contribute) which is also reflected in the feature map gradient in the first UMAP dimension (Suppl. Fig. 2). This notion has been added to the manuscript (page 9) and Suppl. Fig. 2. 

(2) The conclusion that their nucleocentric approach (including a small area beyond the nucleus) is not well supported, and may just be better by random chance. 

To address this second concern, which was also raised by reviewer 2, we have performed a more extensive analysis in which the patch size was varied from 0.6 to 120µm around the nuclear centroid (Fig. 4E and page 9 of the manuscript). We observed that there is little effect of in- or decreasing patch size on the average F-score within the nuclear to cell window, but that the imbalance between the precision and recall increases towards the larger box sizes (>18µm). Under our experimental conditions, the input numbers per class were equal, but this will not be the case in situations where the ground truth is unknown (and needs to be predicted by the CNN). Therefore, a well-balanced CNN is of high importance. This notion has been added to page 15 of the manuscript.

The main advantage of nucleocentric profiling over whole-cell profiling in dense cultures is that it relies on a more robust nuclear segmentation method and is less sensitive to differences in cell density (Suppl. Fig. 1D). In other words, in dense cultures, the segmentation mask will contain similar regional input as the nuclear mask and the nucleocentric crop will contain more perinuclear information which contributes to the prediction accuracy. Therefore, at high densities, the performance of the CNN on whole-cell crops decreases owing to poorer segmentation performance. A CNN that uses nucleocentric crops, will be less sensitive to these errors. This notion has been added to pages 14-15 of the manuscript. 

Additionally, the impact of this work will be limited, given the authors do not provide a specific link to the public source code that they used to process and analyze their data.

The source code is now available on the Github page of the DeVos lab, under the following URL: https://github.com/DeVosLab/Nucleocentric-Profiling

Recommendations for the authors:  

Reviewing Editor (Recommendations For The Authors):

Evaluation summary

The authors present a new application of the high-content image-based morphological profiling Cell Painting (CP) to single cell type classification in mixed heterogeneous induced pluripotent stem cellderived mixed neural cultures. Machine learning models were trained to classify single cell types according to either "engineered" features derived from the image or from the raw CP multiplexed image. The authors systematically evaluated experimental (e.g., cell density, cell types, fluorescent channels, replication biases) and computational (e.g., different models, different cell regions) parameters and argue that focusing on the nucleus and its surroundings contains sufficient information for robust and accurate cell type classification. Models that were trained on mono-cultures (i.e., containing a single cell type) could generalize for cell type prediction in mixed co-cultures, and describe intermediate states of the maturation process of iPSC-derived neural progenitors to differentiation neurons.

Strengths:

Automatically identifying single-cell types in heterogeneous mixed-cell populations is an important application and holds great promise. The simple and high-content assay democratizes use and enables adoption by other labs. The manuscript is supported by comprehensive experimental and computational validations. The manuscript is well-written and easy to follow.

Weaknesses:

The conclusion is that the nucleocentric approach (including a small area beyond the nucleus) is not well supported, and may just be better by random chance. If better supported by additional experiments, this may influence how the field performs cell profiling in the future. Model interpretability (GradCAM) analysis is not convincing. The lack of a public source code repository is also limiting the impact of this study. There are missing details in the figure panels, figure legend, and text that would help the reader to better appreciate some of the technical details.

Essential revisions:

To reach a "compelling" strength of evidence the authors are requested to either perform a comprehensive analysis of the effect of ROI size on performance, or tune down statements regarding the superior performance of their "nucleocentric" approach. Further addition of a public and reproducible source code GitHub repository will lead to an "exceptional" strength of evidence.

To answer the main comment, we have performed an experiment in which we varied the size of the nucleocentric patch and quantified CNN performance. We have also evaluated the operational window of our method by varying the resolution and SNR and we have experimented with different background blanking methods. We have expanded our examples of GradCAM images and now also made our source code and an example data set available via GitHub.

Reviewer #1 (Recommendations For The Authors):

I think that an evaluation of how the excluded cells affect our ability to measure the cell type composition of the population would be helpful to better understand the limitations and practical measurement noise introduced by this approach. A similar evaluation of the excluded cells can also help to better understand the benefit of nucleocentric vs. cell representations by more convincingly demonstrating the case for the nucleocentric approach. In any case, I recommend discussing in more depth the arguments for using the nucleocentric representation and why it is superior to the nuclear representation.

The benefits of nucleocentric representation over nuclear and whole-cell representation are discussed more in depth at pages 14-15 of the manuscript. 

“The nucleocentric approach, which is based on more robust nuclear segmentation, minimizes such mistakes whilst still retaining input information from the structures directly surrounding the nucleus. At higher cell density, the whole-cell body segmentation becomes more error-prone, while also loosing morphological information (Suppl. Fig. 1D). The nucleocentric approach is more consistent as it relies on a more robust segmentation and does not blank the surrounding region. This way it also buffers for occasional nuclear segmentation errors (e.g., where blebs or parts of the nucleus are left undetected).”

It is not entirely clear to me why Figure 5 moves back to "engineered" features after previous figures showed the superiority of the deep learning approach. Especially, where Figure 6 goes again to DL. Dimensionality reduction can be also applied to DL-based classifications (e.g., using the last layer).

Following up on the reviewers’ interesting comment, we extracted the embeddings from the trained CNN and performed UMAP dimensionality reduction. The results are shown in Fig. 3D, 6F and supplementary figure 1B and added to the manuscript on pages 6, 8 and 12. 

We concluded that unsupervised dimensionality reduction using the feature embeddings could separate cell type clusters, where the distance between the clusters reflected the morphological similarity between the cell lines. 

I would recommend including more comprehensive GRADCAM panels in the SI to reduce the concern of cherry-picking examples. What is the interpretation of the nucleocentric area?

A more extensive set of GradCAM images have now been included in supplementary material (Supplementary figure 3) using the same random seeds for all conditions, thus avoiding any cherry picking. We interpret the GradCAM maps on the nucleocentric crops as highlighting the structures surrounding the nucleus (reflecting ER, mitochondria, Golgi) indicating their importance in correct cell classification. This was added to the manuscript on pages 9 and 15.

Missing/lacking details and suggestions in the figure panels and figure legend:

- Scale bars missing in some of the images shown (e.g., Figure 2F, Figure 3D, Figure 4, Supplementary Figure 4), what are the "composite" channels (e.g., Figure 2F), missing x-label in Figure 3B. 

These have now been added.

- Terms that are not clear in the figure and not explained in the legend, such as FITC and cy3 energy (Figure 1C). 

The figure has been adapted to better show the region, channel and feature. We have now added a Table (Table 5), detailing the definition of each morphological feature that is extracted. On page 27, information on feature extraction is noted.

- Details that are missing or not sufficiently explained in the figure legends such as what each data point represents and what is Gini importance (Figure 1D) 

We have added these explanations to the figure legends. The Gini importance or mean decrease in impurity reflects how often this feature is used in decision tree splits across all random forest trees.

Is it the std shown in Figure 2C?

Yes, this has now been added to the legend.  

It is not fully clear what is single/mixed (Figure 2D)

Clarification is added to the legend and in the manuscript on page 6.

explain what is DIV 13-90 in the legend (Figure 5).

DIV stands for days in vitro, here it refers to the days in culture since the start of the neural induction process. This has been added in the legend.

and state what are img1-5 (Supplementary Figures 1B-C) Clarification has been added to the legend.

- Supplementary Figure 1. What is the y-axis in panel C and how do the results align with the cell mask in panel B?

The y-axis represents the intersection over union (IoU). The IoU quantifies the overlap between ground truth (manually segmented ROI) and the ROI detected by the segmentation algorithm. It is defined as the area of the overlapping region over the total area. This clarification has been added to the legend.

- Supplementary Figure 1 and Methods. Please explain when CellPose and when StarDist were applied.

Added to supplementary figure and methods at page 24. In the case of nuclear segmentation (nucleus and nucleocentric crops), Stardist was used. For whole-cell crops, cell segmentation using Cellpose was used.

- Supplementary Figure 4C - the color code is different between nuclear and nucleocentric - this is confusing.

We have changed to color code to correspond in both conditions in Fig. 1A.

- Figure 3B - better to have a normalized measure in the x-axis (number of cells per area in um^2)

We agree and have changed this.

Suggestions and missing/lacking details in the text:

• Line #38: "we then applied this" because it is the first time that this term is presented.

This has been rephrased.

• Line #88: a few words on what were the features extracted would be helpful.

Short description added to page 26-27 and detailed definition of all features added in table 5.

-  Line #91: PCA analysis - the authors can highlight what (known) features were important to PC1 using the linear transformation that defined it.

The 5 most important features of PC1 were (in order of decreasing importance): channel 1 dissimilarity, channel 1 homogeneity, nuclear perimeter, channel 4 dissimilarity and nuclear area.  

- Line #92: Order of referencing Supplementary Figure 4 before referencing Supplementary Figure 13.

The order of the Supplementary images was changed to follow the chronology. 

• Line #96: Can the authors show the data supporting this claim?

The unsupervised UMAP shown in fig. 1B is either color coded by cell type (left) or replicate (right). Based on this feature map, we observe clustering along the UMAP1 axis to be associated with the cell type. Variations in cellular morphology associated with the biological replicate are more visible along the UMAP2 axis. When looking at fig. 1C, the feature map reflecting the cellular area shows a gradient along the UMAP1 direction, supporting the assumption that cell area contributes to the cell type separation. On the other hand, the average intensity (Channel 2 intensity) has a gradient within the feature map along the UMAP2 direction. This corresponds to the pattern associated with the inter-replicate variability in panel B.

- Line #108: what is "nuclear Cy3 energy"?

This represents the local change of pixel intensities within the ROI in the nucleus in the 3rd channel dimension. This parameter reflects the texture within the nuclear region for the phalloidin and WGA staining. The definitions of all handcrafted features are added in table 5 of the manuscript.

- Line #110-112: Can the authors show the data supporting this claim?

The figure has been changed to include the results from a filtered and unfiltered dataframe (exclusion and inclusion of redundant features). Features could be filtered out if the correlation was above a threshold of 0.95. This has been added to page 6 of the manuscript and fig. 1D.  

- Line #115-116: please state the size of the mask.

Added to the text (page 6). We used isotropic image crops of 60µm centred on individual cell centroids.

- Lines 120-122: more details will make this more clear (single vs. mixed).

This has been changed on page 6 of the manuscript.

• Line #142: "(mimics)" - is it a typo?

Tissue mimics refers to organoids/models that are meant to replicate the physiological behaviour.

• Line #159: the bounding box for nucleocentric analysis is 15x15um (and not 60), as stated in the Methods.

Thank you for pointing out this mistake. We have adapted this.

- Line #165: what is the interpretation of what was important for the nucleocentric classification?

The colour code in GradCAM images is indicative of the attention of the CNN (the more to the red, the more attention). In fig. 4D and Suppl. Fig. 3 the structures directly surrounding the nucleus receive high attention from the CNN trained on nucleocentric crops. This has been added to the manuscript page 9 and 15.

• Section starting in line #172: not explicitly stated what model was used (nucleocentric?).

Added in the legend of fig. 5. For these experiments, the full cell segmentation was still used. 

- Section starting in line #199: why use a feature-based model rather than nucleocentric? A short sentence would be helpful.

For CNN training, nucleocentric profiling was used. In response to a legitimate question of one of the reviewers, the feature-based UMAP analysis was replaced with the feature embeddings from the CNN. 

- Line #213: Fig. 5B does not show transitioning cells.

Thank you for pointing this out, this was a mistake and has been changed.

Lines #218-220: not fully clear to some readers (culture condition as a weak label), more details can be helpful.

We changed this at page 11 of the manuscript for clarity. 

“This gating strategy resulted in a fractional abundance of neurons vs. total (neurons + NPC) of 36,4 % in the primed condition and 80,0% in the differentiated condition (Fig. 6C). We therefore refer to the culture condition as a weak label as it does not take into account the heterogeneity within each condition (well).”

-  Line #230: "increasing dendritic outgrowth" - what does it mean? Can you explicitly highlight this phenotype in Figure 5G?

When the cells become more mature during differentiation, the cell body becomes smaller and the neurons form long, thin ramifications. This explanation has been added to page 12 of the manuscript.

• Line #243: is it the nucleocentric CNN?

Yes.

• Lines #304-313, the authors might want to discuss other papers dealing with continuous (non-neural) differentiation state transitions (eg PMID: 38238594).  

A discussion of the use of morphological profiling for longitudinal follow-up of continuous differentiation states has been added to the manuscript at page 18. 

- Line #444: cellpose or stardist? How did the authors use both?

Clarification has been added to supplementary figure 1 and methods at page 24. Stardist was used for nuclear segmentation, whereas Cellpose was used for whole-cell segmentation. 

• Line #470-474: I would appreciate seeing the performance on the full dataset without exclusions.

Cells have been excluded based on 3 arguments: the absence of DAPI intensity, too small nuclear size and absence of ground truth staining. The first two arguments are based on the assumption that ROIs that contain no DAPI signal or are too small are errors in cell segmentation and therefore should not be taken along in the analysis. The third filtering step was based on the ground-truth IF signal. Not filtering out these cells with a ‘dubious’ IF profile (e.g., cells that might be transitioning or are of a different type) would negatively affect the model by introducing noise. It is correct that the predictions are based only on these inputs and so cells of a subsequent test set will only be classified according to these labels which might introduce bias. However, the model could predict increase in neuron/NPC ratio with culture age in absence of ground-truth staining (and thus IF-based filtering).

Reviewer #2 (Recommendations For The Authors):

Figure 1A: it would be interesting to the reader to see the SH-SY5Y data as well.

This has been added in fig. 1A.

Figure 3A: 95-100% image: showing images with the same magnification as the others would help to appreciate the cell density.

Now fig. 4A. The figure has been changed to make sure all images have the same magnification. 

Figure Supp 4 (line 132) is referred to before Figure Supp1 (line 152).

The image order and numbering has been changed to solve this issue.

Figure Supp 2 & 3 are not referred to in the text.

This has been adjusted.

Line 225: a statistical test would help to convince of the accuracy of these results (Figure 5C vs Figure 5F)?

These figures represent the total ROI counts and thus represent a single number.

Line 227: Could you explain to the reader, in a few words, what a dual SMAD inhibition is?

This has been added to the manuscript at page 20. 

“This dual blockade of SMAD signalling in iPSCs is induces neural differentiation by synergistically causing the loss of pluripotency and push towards neuroectodermal lineage.”

Reviewer #3 (Recommendations For The Authors):

I have a few concerns and several comments that, if addressed, may strengthen conclusions, and increase clarity of an already technically sound paper.

Concerns

• The results presented in Figure 3 panel D, may indicate a critical error in data processing and interpretation that the authors must address. The GradCAM method highlights the background as having the highest importance. While it can be argued in the nucleocentric profiling method that GradCAM focuses on the nuclear membrane, the background is highly important even for the nuclear profiling method, which should provide little information. What procedure did the authors use for mask subtraction prior to CNN training? Could the segmentation algorithm be performing differently between cell lines? The authors interpret the GradCAM results to indicate a proxy for nuclear size, but then why did the CNN perform so much better than random forest using hand-crafted features that include this variable? The authors should also present size distributions between cell lines (and across seeding densities, in case one of the cell lines has different compaction properties with increasing density).

Perhaps clarifying this sentence (lines 166-168) would help as well: "As nuclear area dropped with culture density, the dynamic range decreased, which could explain the increased error rate of the CNN for high densities unrelated to segmentation errors (Suppl. Fig. 4B)." What do the authors mean by "dynamic range" and it is not clear how Supplementary Figure 4B provides evidence for this? 

The dynamic range refers to the difference between the minimum and maximum nuclear area. We expect the difference to decrease at highe rdensity owing to the crowding that forces all nuclei to take on a more similar (smaller) size.

More clarification on this has been added to page 9 of the manuscript.

I certainly understand that extrapolating the GradCAM concern to the remaining single-cell images using only four (out of tens of thousands of options) is also dangerous, but so is "cherry-picking" these cells to visualize. Finally, I also recommend that the authors quantitatively diagnose the extent of the background influence according to GradCAM by systematically measuring background influence in all cells and displaying the results per cell line per density.

To avoid cherry picking of GradCAM images, we have now randomly selected for each condition and density 10 images (using random seeds to avoid cherry-picking) and added these in a Suppl. Fig. 3.

In answer to this concern, we refer to the response above: 

“To address the first point, we have adapted the GradCAM images to show an overlay of the input crop and GradCAM heatmap to give a better view of the structures that are highlighted by the CNN. We further investigated the influence of the background on the prediction performance. Our finding that a CNN trained on a monoculture retains a relatively high performance on cocultures implies that the CNN uses the salient characteristics of a cell to recognize it in more complex heterogeneous environments. Assuming that the background can vary between experiments, the prediction of a pretrained CNN on a new dataset indicates that cellular characteristics are used for robust prediction.  When inspecting GradCAM images obtained from the nucleocentric CNN approaches (now added in Suppl. Fig. 3), we noticed that the nuclear periphery typically contributed the most (but not exclusively) to the prediction performance. When using only the nuclear region as input, GradCAMs were more strongly (but again not exclusively) directed to the background surrounding the nuclei. To train the latter CNN, we had cropped nuclei and set the background to a value of zero. To rule out that this could have introduced a bias, we have now performed the exact same training and classification, but setting the background to random noise instead (Suppl. Fig. 2). While this effectively diverted the attention of the GradCAM output to the nucleus instead of the background, the prediction performance was unaltered. We therefore assume that irrespective of the background, when using nuclear crops as input, the CNN is dominated by features that describe nuclear size. We observe that nuclear size is significantly different in both cell types (although intranuclear features also still contribute) which is also reflected in the feature map gradient in the first UMAP dimension (Suppl. Fig. 2). This notion has been added to the manuscript (page 9) and Suppl. Fig. 2.”

• The data supporting the conclusion about nucleocentric profiling outperforming nuclear and full-cell profiling is minimal. I am picking on this conclusion in particular, because I think it is a super cool and elegant result that may change how folks approach issues stemming from cell density disproportionately impacting profiling. Figures 3B and 3C show nucleocentric slightly outperforming full cell, and the result is not significant. The authors state in lines 168-170: "Thus, we conclude that using the nucleocentric region as input for the CNN is a valuable strategy for accurate cell phenotype identification in dense cultures." This is somewhat of a weak conclusion, that, with additional analysis, could be strengthened and add high value to the community. Additionally, the authors describe the nucleocentric approach insufficiently. In the methods, the authors state (lines 501-503): "Cell crops (60μm whole cell - 15μm nucleocentric/nuclear area) were defined based on the segmentation mask for each ROI." This is not sufficient to reproduce the method. What software did the authors use?

Presumably, 60μm refers to a box size around cytoplasm? Much more detail is needed. Additionally, I suggest an analysis to confirm the impact of nucleocentric profiling, which would strengthen the authors' conclusions. I recommend systematically varying the subtraction (-30μm, -20μm, -10μm, 5μm, 0, +5μm, +10μm, etc.) and reporting the density-based analysis in Figure 3B per subtraction. I would expect to see some nucleocentric "sweet spot" where performance spikes, especially in high culture density. If we don't see this difference, then the non-significant result presented in Figures 3B and C is likely due to random chance. The authors mention "iterative data erosion" in the abstract, which might refer to what I am recommending, but do not describe this later.

More detail was added to the methods describing the image crops given as input to the CNN (page 28 of the manuscript). 

“Crops were defined based on the segmentation mask for each ROI. The bounding box was cropped out of the original image with a fixed patch size (60µm for whole cells, 18µm for nucleus and nucleocentric crops) surrounding the centroid of the segmentation mask. For the whole cell and nuclear crops, all pixels outside of the segmentation mask were set to zero. This was not the case for the nucleocentric crops. Each ROI was cropped out of the original morphological image and associated with metadata corresponding to its ground truth label.”

To address this concern, we also refer to the answer above. 

“We have performed a more extensive analysis in which the patch size was varied from 0.6 to 120µm around the nuclear centroid (Fig. 4E and page 9 of the manuscript). We observed that there is little effect of in- or decreasing patch size on the average F-score within the nuclear to cell window, but that the imbalance between the precision and recall increases towards the larger box sizes (>18µm). Under our experimental conditions, the input numbers per class were equal, but this will not be the case in situations where the ground truth is unknown (and needs to be predicted by the CNN). Therefore, a well-balanced CNN is of high importance. This notion has been added to page 12 of the manuscript.

The main advantage of nucleocentric profiling over whole-cell profiling in dense cultures is that it relies on a more robust nuclear segmentation method and is less sensitive to differences in cell density (Suppl. Fig. 1D). In other words, in dense cultures, the segmentation mask will contain similar regional input as the nuclear mask and the nucleocentric crop will contain more perinuclear information which contributes to the prediction accuracy. Therefore, at high densities, the performance of the CNN on whole-cell crops decreases owing to poorer segmentation performance. A CNN that uses nucleocentric crops, will be less sensitive to these errors. This notion has been added to pages 14-15 of the manuscript.“

Comments

• There is a disconnect between the abstract and the introduction. The abstract highlights the nucleocentric model, but then it is not discussed in the introduction, which focuses on quality control. The introduction would benefit from some additional description of the single-cell or whole-image approach to profiling.

We highlight the importance of QC of complex iPSC-derived neural cultures as an application of morphological profiling. We used single-cell profiling to facilitate cell identification in these mixed cultures where the whole-image approach would be unable to deal with the heterogeneity withing the field of view. In the introduction, we added a description of the whole-image vs. single-cell approach to profiling (page 4). In the discussion (page 18), we further highlight the application of this single-cell profiling approach for QC purposes. 

- Comments on Figure 1. It is unclear how panel B shows "without replicate bias". 

In response to this comment, we refer to the answer above: “The unsupervised UMAP shown in fig. 1B is either color coded by cell type (left) or replicate (right). Based on this feature map, we observe clustering along the UMAP1 axis to be associated with the cell type. Variations in cellular morphology associated with the biological replicate are more visible along the UMAP2 axis. When looking at fig. 1C, the feature map reflecting the cellular area shows a gradient along the UMAP1 direction, supporting the assumption that cell area contributes to the cell type separation. On the other hand, the average intensity (Channel 2 intensity) has a gradient within the feature map along the UMAP2 direction. This corresponds to the pattern associated with the inter-replicate variability in panel B.” We added this notion to page 5 of the manuscript.

The paper would benefit from a description of how features were extracted sooner.

Information on the feature extraction was added to the manuscript at page 27. An additional table (table 5) has been added with the definition of each feature.  

- Comments on Supplementary Figure 4. The clustering with PCA is only showing 2 dimensions, so it is not surprising UMAP shows more distinct clustering.

We used two components for UMAP dimensionality reduction, so the data was also visualized in two dimensions. However, we agree that UMAP can show more distinct clustering as this method is non-linear.

Why is Figure S4 the first referenced Supplementary Figure?

This has been changed. 

• Comments on Figure 2. Need discussion of the validation set - how was it determined? Panel E might have the answer I am looking for, but it is difficult to decipher exactly what is being done. The terminology needs to be defined somewhere, or maybe it is inconsistent. It is tough to tell. For example, what exactly are the two categories of model validation (cross-validation and independent testing)?

Additional clarification has been added to the manuscript at pages 6-7 and figure 2.

The metric being reported is accuracy for the independent replicate if the other two are used to train?

Yes. 

Panel C is a very cool analysis. Panel F needs a description of how those images were selected, randomly?

Added in the methods section (page 29). GradCAM analysis was used to visualize the regions used by the CNN for classification. This map is specific to each cell. Images are selected randomly out the full dataset for visualization.  

They also need scale bars.

Added to the figures. 

Panel G would benefit from explicit channel labels (at least a legend would be good!).

Explanation has been added to the legend. All color code and channel numbering are consistent with fig. 1A. 

What do the dots and boxplots represent? The legend says, "independent replicates", but independent replicates of, I assume, different model initializations?

Clarification has been added to the figure legends. For plots showing the performance of a CNN or RF classifier, each dot represents a different model initialization. Each classifier has been initialized at least 3 times. When indicated, the model training was performed with different random seeds for data splitting.

• Comments on Figure 3. Panel A needs scale bar. See comment on Panel D in concern #1 described above. 

This has been added.

• Comments on Supplementary Figure 1. A reader will need a more detailed description in panel C. I assume that the grey bar is the average of the points, and the points represent different single cells?

How many cells? How were these cells selected? 

This information on the figure (now Suppl. Fig. 1D), has been added to the legend.

“Left: Representative images of 1321N1 cells with increasing density alongside their cell and nuclear mask produced using resp. Cellpose and Stardist. Images are numbered from 1-5 with increasing density. Upper right: The number of ROIs detected in comparison to the ground truth (manual segmentation). A ROI was considered undetected when the intersection over union (IoU) was below 0,15. Each bar refers to the image number on the left. The IoU quantifies the overlap between ground truth (manually segmented ROI) and the ROI detected by the segmentation algorithm. It is defined as the area of the overlapping region over the total area. IoU for increasing cell density for cell and nuclear masks is given in the bottom right. Each point represents an individual ROI. Each bar refers to the image number on the left.”

• Comments on Figure 4. More details on quenching are needed for a general audience. The markers chosen (EdU and BrdU) are generally not specific to cell type but to biological processes (proliferation), so it is confusing how they are being used as cell-type markers. 

The base analogues were incorporated into each cell line prior to mixing them, i.e.  when they were still growing in monoculture so they could be labelled and identified after co-seeding and morphological profiling. Additional clarification has been added to the manuscript (page 26) 

It is also unclear why reducing CV is an important side-effect of finetuning. CV of what? The legend says, "model iterations", but what does this mean? 

The dots in the violinplot are different CNN initializations. A lower variability between model initializations is an indicator of certainty of the results. Prior to finetuning, the results of the CNN were highly variable leading to a high CoV between the different CNNs. This means the outcome after finetuning is more robust.

• Comments on Figure 5. This is a very convincing and well-described result, kudos! This provides another opportunity to again compare other approaches (not just nucleocentric). Additionally, since the UMAP space uses hand-crafted features. The authors could consider interpreting the specific morphology features impacted by the striking gradual shift to neuron population by fitting a series of linear models per individual feature. This might confirm (or discover) how exactly the cells are shifting morphology.

The supervised UMAP on the handcrafted features did not highlight any features contributing to the separation. Using the supervised UMAP, the clustering is dominated by the known cell type. Unsupervised UMAP on the handcrafted features does not show any clustering. In response to a previous comment, we adapted the figure to show UMAP dimensionality reduction using the feature embeddings from the cell-based CNN. This unsupervised UMAP does show good cell type separation, but it does not use any directly interpretable shape descriptors.

• General comments on Methods. The section on "ground truth alignment" needs more details. Why was this performed? 

Following sequential staining and imaging rounds, multiple images were captured representing the same cell with different markers. Lifting the plate of the microscope stage and imaging in sequential rounds after several days results in small linear translations in the exact location of each image. These linear translations need to be corrected to align (or register) morphological with ground truth image data within the same ROI. This notion has been added to the manuscript at page 26. 

Handcrafted features extracted using what software? 

The complete analysis was performed in python. All packages used are listed in table 4. Handcrafted features were extracted using the scikit-image package (regionprops and GLCM functions). This has been added to the manuscript at page 27.

Software should be cited more often throughout the manuscript. 

Lastly, the GitHub URL points to the DeVosLab organization, but should point to a specific repository. Therefore, I was unable to review the provided code. A well-documented and reproducible analysis pipeline should be included.

A test dataset and source code are available on GitHub:  https://github.com/DeVosLab/Nucleocentric-Profiling

Author response:

The following is the authors’ response to the original reviews.

Reviewer 1:

Comment 1. In Figure 1, the MafB antibody (Sigma) was used to identify Renshaw cells at P5. However, according to the supplementary Figure 3D, the specificity of the MafB antibody (Sigma) is relatively low. The image of MafB-GFP, V1-INs, and MafB-IR at P5 should be added to the supplementary figure. The specificity of MaFB-IR-Sigma in V1 neurons at P5 should be shown. This image also might support the description of the genetically labeled MafB-V1 distribution at P5 (page 8, lines 28-32). 

We followed the reviewer’s suggestion and moved analyses of the MafB-GFP mouse to a supplemental figure (Fig S3). The characterization of MafB immunoreactivities is now in supplemental Figure S2 and the related text in results was also moved to supplemental to reduce technicalities in the main text. We added confocal images of MafB-GFP V1 interneurons at P5 showing immunoreactivities for both MafB antibodies, as suggested by the reviewer (Fig S2A,B). We agree with the reviewer that this strengthens our comparisons on the sensitivity and specificity of the two MafB antibodies used in this study. 

As explained in the preliminary response we cannot show lack of immunoreactivity for MafB antibodies in MafB GFP/GFP knockout mice at P5 because MafB global KOs die at birth. This is why we used tissues from late embryos to check MafB immunoreactivities (Figure S2C and S2D). We made this point clearer in the text and supplemental figure legends.

Comment 2. The proportion of genetically labeled FoxP2-V1 in all V1 is more than 60%, although immunolabeled FoxP2-V1 is approximately 30% at P5. Genetically labeled Otp-V1 included other nonFoxP2 V1 clades (Fig. 8L-M). I wonder whether genetically labeled FoxP2-V1 might include the other three clades. The authors should show whether genetically labeled FoxP2-V1 expresses other clade markers, such as pou6f2, sp8, and calbindin, at P5. 

We included the requested data in Figure 3E-G. Lineage-labeled Foxp2-V1 neurons in our genetic intersection do not include cells from other V1-clades.

Reviewer 2:

Comment 1. The current version of the paper is VERY hard to read. It is often extremely difficult to "see the forest for the trees" and the reader is often drowned in methodological details that provide only minor additions to the scientific message. Non-specialists in developmental biology, but still interested in the spinal cord organization, especially students, might find this article challenging to digest and there is a high risk that they will be inclined to abandon reading it. The diversity of developmental stages studied (with possible mistakes between text and figures) adds a substantial complexity in the reading. It is also not clear at all why authors choose to focus on the Foxp2 V1 from page 9. Naively, the Pou6f2 might have been equally interesting. Finally, numerous discrepancies in the referencing of figures must also be fixed. I strongly recommend an in-depth streamlining and proofreading, and possibly moving some material to supplement (e.g. page 8, and elsewhere).

The whole text was re-written and streamlined with most methodological discussion (including the section referred to by the reviewer) transferred to supplemental data. Nevertheless, enough details on samples, stats and methods were retained to maintain the rigor of the manuscript. 

The reasons justifying a focus on Foxp2-V1 interneurons were fully explained in our preliminary response. Briefly, we are trying to elucidate V1 heterogeneity, and prior data showed that this is the most heterogeneous V1 clade (Bikoff et al., 2016), so it makes sense it was studied further. We agree that the Pou6f2 clade is equally interesting and is in fact the subject of several ongoing studies.

Comment 2. … although the different V1 populations have been investigated in detail regarding their development and positioning, their functional ambition is not directly investigated through gain or loss of function experiments. For the Foxp2-V1, the developmental and anatomical mapping is complemented by a connectivity mapping (Fig 6s, 8), but the latter is fairly superficial compared to the former. Synapses (Fig 6) are counted on a relatively small number of motoneurons per animal, that may, or may not, be representative of the population. Likewise, putative synaptic inputs are only counted on neuronal somata. Motoneurons that lack of axo-somatic contacts may still be contacted distally. Hence, while this data is still suggestive of differences between V1 pools, it is only little predictive of function.

We fully answered the question on functional studies in the preliminary response. Briefly, we are currently conducting these studies using various mouse models that include chronic synaptic silencing using tetanus toxin, acute partial silencing using DREADDs, and acute cell deletion using diphtheria toxin. Each intervention reveals different features of Foxp2-V1 interneuron functions, and each model requires independent validation. Moreover, these studies are being carried out at three developmental stages: embryos, early postnatal period of locomotor maturation and mature animals. Obviously, this is all beyond the goals and scope of the present study. The present study is however the basis for better informed interpretations of results obtained in functional studies.

Regarding the question on synapse counts, we explained in the preliminary results fully why we believe our experimental designs for synapse counting at the confocal level are among the most thorough that can be found in the literature. We counted a very large number of motoneurons per animal when adding all motor column and segments analyzed in each animal. Statistical power was also enough to detect fundamental variation in synaptic density among motor columns.

We focus our analyses on motoneuron cells bodies because analysis of full dendritic arbors on all motor columns present throughout all lumbosacral segments is not feasible. Please see Rotterman et al., 2014 (J. of Neuroscience; doi: 10.1523/JNEUROSCI.4768-13.2014) for evaluation of what this entails for a single motoneuron. We agree with the reviewer that analyses of V1 synapses over full dendrite arbors in specific motoneurons will be very relevant in further studies. These should be carried out now that we know which motor columns are of high interest. Nevertheless, inhibitory synapses exert the most efficient modulation of neuronal firing when they are on cell bodies, and our analyses clearly suggest a difference in in cell body inhibitory synapses targeting between different V1 interneuron types that we find very relevant.

Comment 3. I suggest taking with caution the rabies labelling (Figure 8). It is known that this type of Rabies vectors, when delivered from the periphery, might also label sensory afferents and their postsynaptic targets in the cord through anterograde transport and transneuronal spread (e.g., Pimpinella et al., 2022). Yet I am not sure authors have made all controls to exclude that labelled neurons, presumed here to be premotoneurons, could rather be anterogradely labelled from sensory afferents. 

Over the years, we performed many extensive controls and validation of rabies virus transsynaptic tracing methods. These were presented at two SfN meetings (Gomez-Perez et al., 2015 and 2016; Program Nos. 242.08 and 366.06). Our validation of this technique was fully explained in our preliminary response. We also pointed out that the methods used by Pimpinella et al. have a very different design and therefore their results are not comparable to ours. In this study we injected the virus at P15 into leg muscles, and not directly into the spinal cord. In our hands, and as cited in Pimpinella et al., the rabies virus loses tropism for primary afferents with age when injected in muscle. The lack of primary afferent labeling in key lumbosacral segments (L4 and L5) is now illustrated in a new supplemental figure (Figure S6). This figure also shows some starter motoneurons. As explained in the text and in our previous response, these are few in number because of the reduced infection rate when using this method in mature animals (after P10).  

Comment 4. The ambition to differentiate neuronal birthdate at a half-day resolution (e.g., E10 vs E10.5) is interesting but must be considered with caution. As the author explains in their methods, animals are caged at 7pm, and the plug is checked the next morning at 7 am. There is hence a potential error of 12h. 

We agree with the reviewer, and we previously explicitly discussed these temporal resolution caveats. We have now further expanded on this in new text (see middle paragraph in page 5). Nevertheless, the method did reveal the temporal sequence of neurogenesis of V1 clades with close to 12-hour resolution.

As explained in text and preliminary response this is because we analyzed a sufficient number of animals from enough litters and utilized very stringent criteria to count EdU positives. 

Moreover, our results fit very well with current literature. The data agree with previous conclusions from Andreas Sagner group (Institut für Biochemie, Friedrich-Alexander-Universität Erlangen-Nürnberg), on spinal interneurons (including V1s) birthdates based on a different methodology (Delile J et al.

Development. 2019 146(12):dev173807. doi: 10.1242/dev.173807. PMID: 30846445; PMCID: PMC6602353). In the discussion we compared in detail both the data and methods between Delile article and our results. We also cite Sagner 2024 review as requested later in the reviewer’s detailed comments. Our results also confirmed our previous report on the birthdates of V1-derived Renshaw cells and Ia inhibitory interneurons (Benito-Gonzalez A, Alvarez FJ J Neurosci. 2012 32(4):1156-70. doi: 10.1523/JNEUROSCI.3630-12.2012. PMID: 22279202; PMCID: PMC3276112). Finally, we recently received a communication notifying us that our neurogenesis sequence of V1s has been replicated in a different vertebrate species by Lora Sweeney’s group (Institute of Science and Technology Austria; direct email from this lab) and we shared our data with them for comparison. This manuscript is currently close to submission. Therefore, we are confident that despite the limitations of EdU birthdating we discussed, the conclusions we offered are strong and are being validated by other groups using different methods and species. We also want to acknowledge the positive comments of reviewer 3 regarding our birthdating study, indicating it is one the most rigorous he or she has ever seen.

Reviewer 3:

Comment 1. My only criticism is that some of the main messages of the paper are buried in technical details. Better separation of the main conclusions of the paper, which should be kept in the main figures and text, and technical details/experimental nuances, which are essential but should be moved to the supplement, is critical. This will also correct the other issue with the text at present, which is that it is too long.

Similar to our response to comment 1 from Reviewer 2 we followed the reviewers’ recommendations and greatly summarized, simplified and removed technical details from the main text, trying not to decrease rigor.  

Reviewer #1 (Recommendations For The Authors):

In Figure 1, the definition of the area to analyze MafB ventral and MafB dorsal is unclear. It should be described.

This has been clarified in both text and supplemental figure S3.

“We focused the analyses on the brighter dorsal and ventral MafB-V1 populations defined by boxes of 100 µm dorsoventral width at the level of the central canal (dorsal) or the ventral edge of the gray matter (ventral) (Supplemental Figure S3B).”

Problems with figure citation.

We apologize for the mistakes. All have been corrected. 

Reviewer #2 (Recommendations For The Authors):

As indicated in the public review, I'd recommend to substantially revise the writing, for clarity. As such, the paper is extremely hard to read. I would also recommend justifying the focus on Foxp2 neurons.

Also, the scope of the present paper is not clearly stated in the introduction (page 4).

Done. We also modified the introduction such that the exact goals are more clearly stated.

I would also recommend toning down the interpretation that V1 clades constitute "unique functional subsets" (discussion and elsewhere). Functional investigation is not performed, and connectomic data is partial and only very suggestive.

We include the following sentence at the end of the 1st paragraph in the discussion:

“This result strengthens the conclusion that these V1 clades defined by their genetic make-up might represent distinct functional subtypes, although further validation is necessary in more functionally focused studies.”

Different post-natal stages are used for different sections of the manuscript. This is often confusing, please justify each stage. From the beginning even, why is the initial birthdating (Figure 1) done here at p5, while the previous characterization of clades was done at p0? I am not sure to understand the justification that this was chosen "to preserve expression of V1 defining TFs". Isn't the sooner the better?

The birthdating study was carried out at P5. P5 is a good time point because there is little variation in TF expression compared to P0, as demonstrated in the results. Furthermore, later tissue harvesting allows higher replicability since it is difficult to consistently harvest tissue the day a litter is born (P0). Also technically, it is easier to handle P5 tissue compared to P0. The analysis of VGUT1 synapses was also done at P5 rather than later ages. This has two advantages: TFs immunoreactivities are preserved at this age, and also corticospinal projections have not yet reached the lumbar cord reducing interpretation caveats on the origins of VGUT1 synapses in the ventral horn (although VGLUT1 synapses are still maturing at this age, see below).

Other parts of the study focus on different ages selected to be most adequate for each purpose. To best study synaptic connectivity, it is best to study mature spinal cords after synaptic plasticity of the first week. For the tracing study we thoroughly explain in the text the reasons for the experimental design (see also below in detailed comments). For counting Foxp2-V1 interneurons and comparing them to motor columns we analyze mature animals. For testing our lineage labeling we use animals of all ages to confirm the consistency of the genetic targeting strategy throughout postnatal development and into adulthood.

Figure 5: wouldn't it be worth quantifying and illustrating cellular densities, in addition to the average number of Foxp2 neurons, across lumbar segments (panel D & E)? Indeed, the size of - and hence total number of cells within - each lumbar segment might not be the same, with a significant "enlargement" from L2 to L4 (this is actually visible on the transverse sections). Hence, if the total number of cells is in the higher in these enlarged segments, but the total number of Foxp2-V1 is not, it may mean that this class is proportionally less abundant.

We believe the critical parameter is the ratio of Foxp2-V1s to motoneurons. This informs how Foxp2-V1 interneurons vary according to the size of the motor columns and the number of motoneurons overall.

The question asked by the reviewer would best be answered by estimating the proportion of Foxp2-V1 neurons to all NeuN labeled interneurons. This is because interneuron density in the spinal cord varies in different segments. We are not sure what this additional analysis will contribute to the paper.

Why, in the Rabies tracing scheme (Fig 8), the Rabies injection is performed at p15? As the authors explain in the text, rabies uptake at the neuromuscular junction is weak after p10. It is not clear to me why such experiments weren't done all at early postnatal stages, with a "classical" co-injection of TVA and Rabies.

First, we do not need TVA in this experiment because we are using B19-G coated virus and injecting it into muscles, not into the spinal cord directly.

Second, enhanced tracing occurs when the AAV is injected a few days before rabies virus. This is because AAV transgene expression is delayed with respect to rabies virus infection and replication. We have performed full time courses and presented these data in one abstract to SfN: Gomez-Perez et al., 2015 Program Nos. 242. We believe full description of these technical details is beyond the scope of this manuscript that has already been considered too technical.

Third, the justification of P15 timing of injections for anterograde primary afferent labeling and retrograde monosynaptic labeling of interneurons is fully explained in the text. 

“To obtain transcomplementation of RVDG-mCherry with glycoprotein in LG motoneurons, we first injected the LG muscle with an AAV1 expressing B19-G at P4. We then performed RVDG and CTB injections at P15 to optimize muscle targeting and avoid cross-contamination of nearby muscles. Muscle specificity was confirmed post-hoc by dissection of all muscles below the knee. Analyses were done at P22, a timepoint after developmental critical windows through which Ia (VGLUT1+) synaptic numbers increase and mature on V1-IaINs (Siembab et al., 2010)” 

Furthermore, CTB starts to decrease in intensity 7 days after injection because intracellular degradation and rabies virus labeling disappears because cell death. Both limit the time of postinjection for analyses.

Likewise, I am surprised not to see a single motoneuron in the rabies tracing (Fig 8, neither on histology nor on graphs (Fig 8). How can authors be certain that there was indeed rabies uptake from the muscle at this age, and that all labelled cells, presumed to be preMN, are not actually sensory neurons? It is known that Rabies vectors, when delivered from the periphery, might also label sensory afferents and their post-synaptic targets through anterograde transport and transneuronal spread (e.g., Pimpinella et al., 2022). This potential bias must be considered.

This is fully explained in our previous response to the second reviewer’s general comments. We have also added a confocal image showing starter motoneurons as requested (Figure S6A).

Please carefully inspect the references to figures and figure panels, which I suspect are not always correct.

Thank you. We carefully revised the manuscript to correct these deficiencies and we apologize for them.

Reviewer #3 (Recommendations For The Authors):

Figure 1: Data here is absolutely beautiful and provides one of the most thorough studies, in terms of timepoints, number of animals analyzed, and precision of analysis, of edU-based birth timing that has been published for neuron subtypes in the spinal cord so far. My only suggestion is to color code the early and late born populations (in for example, different shades of green for early; and blue for late, to better emphasize the differences between them). It is very difficult to differentiate between the purple, red and black colors in G-I, which this would also fix. The antibody staining for Pou6f2 (F) is also difficult to see; gain could be increased on these images or insets added for clarity.

The choice of colors is adapted for optimal visualization by people with different degrees of color blindness. Shades of individual colors are always more difficult to discriminate. This is personally verified by the senior corresponding author of this paper who has some color discrimination deficits. Moreover, each line has a different symbol for the same purpose of easing differentiation.

Figure 2: This is also a picture-perfect figure showing further diversity by birth time even within a clade. One small aesthetic comment is that the arrows are quite unclear and block the data. Perhaps the contours themselves could be subdivided by region and color coded by birth time-such that for example the dorsal contours that emerge in the MafB clade at E11 are highlighted in their own color. Some quantification of the shift in distribution as well as the relative number of neurons within each spatially localized group would also be useful. For MafB, for example, it looks as though the ventral cells (likely Renshaw) are generated at all times in the contour plots; in the dot plots however, it looks like the most ventral cells are present at e10.5. This is likely because the contours are measuring fractional representations, not absolute number. An independent measure of absolute number of ventral and dorsal, by for example, subdividing the spinal cord into dorsoventral bins, would be very useful to address this ambiguity.

We believe density plots already convey the message of the shift in positions with birthdate. We are not sure how we can quantify this more accurately than showing the differences in cellular density plots. We used dorsoventral and mediolateral binning in our first paper decades ago (Avarez et al., 2005). This has now been replaced by more rigorous density profiles that describe better cell distributions. Unfortunately, to obtain the most accurate density profiles we need to pool all cells from all animals precluding statistical comparisons. This is because for some groups there have very few cells per animal (for example early born Sp8 or Foxp2 cells).

Figure 3 and Figure 4: These, and all figures that compare the lineage trace and antibody staining, should be moved to the supplement in my opinion-as they are not for generalist readers but rather specialists that are interested in these exact tools. In addition, the majority of the text that relates to these figures should be transferred to the supplement as well. Figure 5: Another great figure that sets the stage for the analysis of FoxP2V1-to-MN synaptic connectivity, and provides basic information about the rostrocaudal distribution of this clade, by analyzing settling position by level. I have only minor comments. The grid in B obscures the view of the cells and should be removed. The motor neuron cell bodies in C would be better visible if they were red.

We moved some of the images to supplemental (see new supplemental Fig S4). However, we also added new data to the figure as requested by reviewers (Fig 3E-G). We preserved our analyses of Foxp2 and non-Foxp2 V1s across ages and spinal segments because we think this information is critical to the paper. Finally, we want to prevent misleading readers into believing that Foxp2 is a marker that is unique to V1s. Therefore, we also preserved Figures 3H to 3J showing the non-V1 Foxp2 population in the ventral horn. 

Figure 6: Very careful and quantitative analysis of V1 synaptic input to motor neurons is presented here.  For the reader, a summary figure (similar to B but with V1s too) that schematizes V1 FoxP2 versus Renshaw cell connectivity with LMC, MMC, and PGC motor neurons are one level would be useful.

Thanks for the suggestion. A summary figure has now been included (Figure 5G). 

Figure 7: The goal of this figure is to highlight intra-clade diversity at the level of transcription factor expression (or maintenance of expression), birth timing and cell body position culminating in the clear and concise diagram presented in G. In panels A-F however, it takes extra effort to link the data shown to these I-IV subtypes. The figure should be restructured to better highlight these links. One option might be to separate the figure into four parts (one for each type): with the individual spatial, birth timing and TF data for each population extracted and presented in each individual part.

We agree with the reviewer that this is a very busy figure. We tried to re-structure the figure following the suggestions of the reviewer and also several alternative options. All resulted in designs that were more difficult to follow than the original figure. We apologize for its complexity, but we believe this is the best organization to describe all the data in the simplest form.

Figure 8: in A-D, the main point of the figure - that V1FoxP2Otp preferentially receive proprioceptive synapses is buried in a bunch of technical details. To make it easier for the reader, please:

(1) add a summary as in B of the %FoxP2-V1 Otp+ cells (82%) with Vglut1 synapses to make the point stronger that the majority of these cells have synapses.

We added this graph by extending the previous graph to include lineage labeled Foxp2-V1s with OTP or Foxp2 immunoreactivity. It is now Figure 7B.

(2) Additionally, add a representative example that shows large numbers of proximal synapses on an FoxP2-V1 Otp+.

The image we presented before as Figure 8A was already immunostained for OTP, so we just added the OTP channel to the images. Now all this information is in panels that are subparts of Figure 7A.

(3) Move the comparison between FoxP2-V1 and FoxP2AB+V1s to the supplement.

We preserved the quantitative data on Foxp2-V1 lineage cells with Foxp2-immunoreactivity but made this a standalone figure, so it is not as busy.

(4) Move J-M description of antibody versus lineage trace of Otp to supplement as ending with this confuses the main message of the paper (see comment above).

All results for the Otp-V1 mouse model have now been placed in a supplemental figure (Figure 5S).

Discussion: A more nuanced and detailed discussion of how the temporal pattern of subtype generation presented here aligns with the established temporal transcription factor code (nicely summarized in Sagner 2024) would be helpful to place their work in the broader context of the field.

This aspect of the discussion was expanded on pages 20 and 21. We replaced the earlier cited review (Sagner and Briscoe, 2019, Development) with the updated Sagner 2024 review and further discussed the data in the context of the field and neurogenesis waves throughout the neural tube, not only the spinal cord. We previously carefully compared our data with the spinal cord data from Sagner’s group (Delile et, 2019, Development). We have now further expanded this comparison in the discussion.

Author response:

The following is the authors’ response to the original reviews.

Reviewer #1:

In this manuscript by Napoli et al, the authors study the intracellular function of Cytosolic S100A8/A9 a myeloid cell soluble protein that operates extracellularly as an alarmin, whose intracellular function is not well characterized. Here, the authors utilize state-of-the-art intravital microscopy to demonstrate that adhesion defects observed in cells lacking S100A8/A9 (Mrp14-/-) are not rescued by exogenous S100A8/A9, thus highlighting an intrinsic defect. Based on this result subsequent efforts were employed to characterize the nature of those adhesion defects.

The authors thank reviewer #1 for his/her insightful comments and suggestions. Please find our point to point responses below.

(1) Ex vivo characterization of the function of S100A8/A9 in adhesion, spreading, and calcium signaling requires at least one rescue experiment to support the direct role of these proteins in the biological processes under study.

We thank the reviewer for this comment. We agree that rescue experiments would be helpful to confirm the direct role of intracellular S100A8/A9 in adhesion, spreading, and Ca2+ signaling. Although transfection of primary cells, especially neutrophils, poses challenges due to their short half-life, we now have undertaken additional in vitro rescue experiments. Specifically, we used extracellular S100A8/A9 and coated Ibidi flow chambers with E-selectin, ICAM-1 and CXCL1 alone or alongside S100A8/A9, and measured rolling and adhesion of blood neutrophils. Our data reveal that extracellular S100A8/A9 can induce increased adhesion in WT neutrophils but fails to rescue the adhesion defect in Mrp14-/- neutrophils (Author response image 1). This result corroborates our in vivo findings, emphasizing that the observed adhesion defect is due to the lack of intracellular S100A8/A9.

Author response image 1.

Extracellular S100A8/A9 does not rescue the adhesion defect in Mrp14/- neutrophils. Analysis of number of adherent leukocytes FOV-1 normalized to the WBC of WT and Mrp14-/- mice. Whole blood was harvested through a carotid artery catheter and perfused with a high precision pump at constant shear rate using flow cambers coated with either E-selectin, ICAM-1 and CXCL1 or E-selectin, ICMA-1, CXCL1 and S100A8/A9. [mean+SEM, n=5 mice per group, 12 (WT) and 14 (Mrp14-/-) flow chambers, 2way ANOVA, Sidak’s multiple comparison]. ns, not significant; *p≤0.05, **p≤0.01, ***p≤0.001.

(2) There is room for improvement in the analysis of signaling pathways presented in Figures 3 H and I. Western blots and analyses are not convincing, in particular for p-Pax.

We acknowledge the reviewer's concern regarding the clarity of the signaling pathway analysis, particularly the western blots for p-Paxillin. To address this, we have repeated the western blot experiments using murine neutrophils. Our new data confirm the defective paxillin phosphorylation upon CXCL1 stimulation and ICAM-1 binding in the absence of cytosolic S100A8/A9. We have now integrated these new findings with the original data and included the updated results in the manuscript (Figure 3I revised). These enhanced analyses provide a more robust and convincing demonstration of the signaling defects in Mrp14-/- neutrophils.

(3) At least one western blot showing a knockdown of S100A8/A9 should be included towards the beginning of the result section.

We appreciate the reviewer's suggestion to include a western blot demonstrating the knockout of S100A8/A9 early in the results section. In a recent publication by our group, we have already demonstrated the absence of S100A8/A9 at the protein level in Mrp14-/- neutrophils via western blotting ([1], please refer to Extended Data Fig. 1h). We agree that visual confirmation of the absence of S100A8/A9 protein is crucial for establishing the validity of our study.

(4) The Ca2+ measurements at LFA-1 nanoclusters using the Mrp14-/- Lyz2xGCamP5 are interesting; It is understood that the authors are correcting calcium levels by normalizing by LFA-1 cluster areas and that seems fine to me. The issue is that the total calcium signal seems decreased in Mrp14-/- cells compared to WT cells (Fig. 4E)...why is totalCa2+ low? Please discuss.

We thank the reviewer for this insightful comment. Indeed, our observations reveal reduced overall Ca2+ levels in Mrp14-/- neutrophils compared to WT neutrophils. Initially, we noticed a general decrease in Ca2+ intensity (Author response image 2A-B) and lifetime in Mrp14-/- neutrophils (Author response image 2C-D). Further analysis indicated that these differences in Ca2+ levels are localized specifically to the LFA-1 nanocluster sites. In contrast, the cytosolic Ca2+ levels outside of the LFA-1 nanocluster areas were comparable between Mrp14-/- and WT neutrophils (Figure 4H-J). This suggests that the reduced total Ca2+ levels observed in Mrp14-/- neutrophils are primarily due to the impaired Ca2+ supply at the LFA-1 nanocluster areas. Our data support the notion that cytosolic S100A8/A9 plays a crucial role in actively supplying Ca2+ to LFA-1 nanoclusters during neutrophil crawling. In the absence of S100A8/A9, the increase in overall Ca2+ levels (summing both inside and outside LFA-1 nanocluster areas) is minimal, further highlighting the specific role of S100A8/A9 in maintaining localized Ca2+ concentrations at these crucial sites.

Author response image 2.

Overall Ca2+ levels in WT and Mrp14-/- neutrophils (A) Representative confocal images of neutrophils from WT Lyz2xGCaMP5 and Mrp14-/- Lyz2xGCaMP5 mice, labeled with Lyz2 td Tomato marker. The images illustrate overall cytosolic Ca2+ levels during neutrophil crawling flow chambers coated with E-selectin, ICAM-1, and CXCL1 (scale bar=10μm). (B) Quantitative analysis of total cytosolic Ca2+ intensity in single cells from WT Lyz2xGCaMP5 and Mrp14-/- Lyz2xGCaMP5 neutrophils measured over three time intervals: min 0-1, 5-6 and 9-10 [mean+SEM, n=5 mice per group, 56 (WT) and 54 (Mrp14-/-) neutrophils, 2way ANOVA, Sidak’s multiple comparison]. (C) Representative traces and (D) single cell analysis of total Ca2+ lifetime over the first 5 minutes in WT Lyz2xGCaMP5 and Mrp14-/- Lyz2xGCaMP5 neutrophils crawling on Eselectin, ICAM-1, and CXCL1 coated flow chambers recorded with FLIM microscopy [mean+SEM, n=3 mice per group, 111 (WT) and 95 (Mrp14-/-) neutrophils, 2way ANOVA, Sidak’s multiple comparison]. ns, not significant; *p≤0.05, **p≤0.01, ***p≤0.001.

(5) Even if the calcium level outside LFA-1 nanoclusters is not significant (Figure 4J), the data at min 9-10 in Figure 4J seems to be affected by a single event that may be an outlier. Additional data may be needed here.

We appreciate the reviewer’s attention to this detail. To address the concern regarding a potential outlier in the Ca2+ level measurements at 9-10 minutes in Figure 4J, we rigorously tested the dataset using the GraphPad outlier calculator. The analysis revealed that no data point was statistically identified as an outlier. Given that the current dataset is robust and the statistical analysis confirms the integrity of the data, we believe that the results accurately reflect the biological variability observed in our experiments. Therefore, we have not added additional data points at this stage but remain open to discussing this further.

(6) Finally, even though there is less calcium at LFA-1 clusters, that does not necessarily mean that "cytosolic S100A8/A9 plays an important role in Ca2+ "supply" at LFA-1 adhesion spots" as proposed. S100A8/A9 may play an indirect role in calcium availability. The analysis of the subcellular localization of S100A8/A9 at LFA-1 clusters together with calcium dynamics in stimulated WT cells would help support the authors' interpretation, which although possibly correct, seems speculative at this point.

We thank the reviewer for this insightful comment and fully agree that additional evidence regarding the subcellular localization of S100A8/A9 would strengthen our conclusions. Although live cell imaging of intracellular S100A8/A9 was initially challenging due to technical limitations, we have now performed additional experiments to address this issue. We conducted end-point measurements where we allowed WT neutrophils to crawl on E-selectin, ICAM-1, and CXCL1 coated flow chambers for 10 minutes. Following this, we fixed and permeabilized the cells to stain intracellular S100A9, along with LFA-1 and a cell tracker for segmentation. Confocal microscopy and subsequent single-cell analysis revealed a significant enrichment of S100A8/A9 at LFA-1 positive nanocluster areas compared to the surrounding cytosol (Figure 4K and 4L, new). This finding supports our hypothesis that S100A8/A9 plays a direct role in the localized supply of Ca2+ at LFA-1 adhesion spots, thus facilitating efficient neutrophil crawling under shear stress. These new data have been included in the revised manuscript, providing stronger evidence for our proposed mechanism.

Reviewer #2:

Napoli et al. provide a compelling study showing the importance of cytosolic S100A8/9 in maintaining calcium levels at LFA-1 nanoclusters at the cell membrane, thus allowing the successful crawling and adherence of neutrophils under shear stress. The authors show that cytosolic S100A8/9 is responsible for retaining stable and high concentrations of calcium specifically at LFA-1 nanoclusters upon binding to ICAM-1, and imply that this process aids in facilitating actin polymerisation involved in cell shape and adherence. The authors show early on that S100A8/9 deficient neutrophils fail to extravasate successfully into the tissue, thus suggesting that targeting cytosolic S100A8/9 could be useful in settings of autoimmunity/acute inflammation where neutrophil-induced collateral damage is unwanted.

The authors appreciate reviewer #2's insightful comments and suggestions. Below are our detailed responses:

(1) Extravasation is shown to be a major defect of Mrp14-/- neutrophils, but the Giemsa staining in Figure 1H seems to be quite unspecific to me, as neutrophils were determined by nuclear shape and granularity. It would have perhaps been more clear to use immunofluorescence staining for neutrophils instead as seen in Supplementary Figure 1A (staining for Ly6G or other markers instead of S100A9).

We acknowledge the reviewer's concern. However, Giemsa staining is a well-established method in hematology, histology, cytology, and bacteriology, widely recognized for its ability to distinguish leukocyte subsets based on nuclear shape and cytoplasmic characteristics. This method is extensively documented in the literature [2-5]. Its advantages are the easy morphological discrimination of leukocytes based on nuclear and cytoplasmic shape and conformation (Author response image 3).

Author response image 3.

Giemsa staining of extravasated leukocyte subsets. (A) Representative image of Giemsa-stained cremaster muscle tissue post-TNF stimulation. The image clearly differentiates leukocyte subsets (white arrow = neutrophils, yellow arrow = eosinophils, red arrow = monocytes). Scale bar = 50µm.

(2) The representative image for Mrp14-/- neutrophils used in Figure 4K to demonstrate Ripley's K function seems to be very different from that shown above in Figures 4C and 4F.

The reviewer correctly observed that the cell in Figure 4K is different from those in Figures 4C and 4F. This is intentional, as Figure 4K is meant to show a representative image that accurately reflects the overall results of the experiments. We assure the reviewer that all cells analyzed in Figures 4C and 4F were also included in the analysis for Figure 4K.

(3) Although the authors have done well to draw a path linking cytosolic S100A8/9 to actin polymerisation and subsequently the arrest and adherence of neutrophils in vitro, the authors can be more explicit with the analysis - for example, is the F-actin co-localized with the LFA-1 nanoclusters? Does S100A8/9 localise to the membrane with LFA-1 upon stimulation? Lastly, I think it would have been very useful to close the loop on the extravasation observation with some in vitro evidence to show that neutrophils fail to extravasate under shear stress.

We thank the reviewer for this comment and questions. 

Concerning the co-localization of F-actin with LFA-1 nanoclusters and S100A8/9 localization: We appreciate the reviewer's interest in the co-localization between F-actin and LFA-1. Unfortunately, due to the limitations of our GCaMP5 mouse model (with neutrophils labeled with td-Tomato and eGFP for LyzM and Ca2+), we could only stain for either LFA-1 or F-actin at a time. However, in our F-actin movies, we observed that F-actin predominantly localizes at the rear of the cell, while LFA-1 is more uniformly distributed at the plasma membrane.

Regarding S100A8/A9 localization, as mentioned in response to Reviewer 1's sixth point, we now conducted endpoint measurements. We stained neutrophils with cell tracker green CMFDA and LFA-1, allowed them to crawl on E-selectin, ICAM-1, and CXCL1-coated flow chambers, and then performed intracellular S100A9 staining after fixation and permeabilization. Our analysis shows higher S100A9 intensity at LFA-1 positive areas compared to LFA-1 negative areas (Figure 4K and 4L, new). This indicates that S100A8/A9 indeed concentrates Ca2+ at LFA-1 nanoclusters, supporting adhesion and post-arrest modification events under flow.

Regarding the extravasation defect under shear stress: To address the reviewer's suggestion, we performed transwell migration assays under static conditions. Our results show no significant difference in transmigration between WT and Mrp14-/- neutrophils without flow, indicating that the extravasation defect in Mrp14-/- neutrophils is shear-dependent. This supports our hypothesis that S100A8/A9-mediated Ca2+ supply at LFA-1 nanoclusters is critical under flow conditions (Author response image 4).

Author response image 4.

Static Transmigration assay. (a) Transmigration of WT and Mrp14-/- neutrophils in static transwell assays (3um pore size, 45min migration time) showing spontaneously migration (PBS) or migration towards CXCL1. [mean+SEM, n=3 mice per group, 2way ANOVA, Sidak’s multiple comparison]. ns, not significant; *p≤0.05, **p≤0.01, ***p≤0.001.

Additional References

(1) Pruenster, M., et al., E-selectin-mediated rapid NLRP3 inflammasome activation regulates S100A8/S100A9 release from neutrophils via transient gasdermin D pore formation. Nature Immunology, 2023. 24(12): p. 2021-2031.

(2) Kuwano, Y., et al., Rolling on E- or P-selectin induces the extended but not high-affinity conformation of LFA-1 in neutrophils. Blood, 2010. 116(4): p. 617-24.

(3) Porse, B., Mouse Hematology – A Laboratory Manual. European Journal of Haematology, 2010. 84(6): p. 554-554.

(4) Frommhold, D., et al., Protein C concentrate controls leukocyte recruitment during inflammation and improves survival during endotoxemia after efficient in vivo activation. Am J Pathol, 2011. 179(5): p. 2637-50.

(5) Braach, N., et al., RAGE Controls Activation and Anti-Inflammatory Signalling of Protein C. PLOS ONE, 2014. 9(2): p. e89422.

Author response:

The following is the authors’ response to the original reviews.

Public Reviews:

Reviewer #1 (Public Review):

Summary:

The study characterized the cellular and molecular mechanisms of spike timing-dependent long-term depression (t-LTD) at the synapses between excitatory afferents from lateral (LPP) and medial (MPP) perforant pathways to granule cells (GC) of the dentate gyrus (DG) in mice.

Strengths:

The electrophysiological experiments are thorough. The experiments are systematically reported and support the conclusions drawn.

This study extends current knowledge by elucidating additional plasticity mechanisms at PP-GC synapses, complementing existing literature.

We thank the reviewer for the positive assessment of our work and the constructive suggestions to improve the manuscript.

Weaknesses:

To more conclusively define the pivotal role of astrocytes in modulating t-LTD at MPP and LPP GC synapses through SNARE protein-dependent glutamate release, as posited in this study, the authors could adopt additional methods, such as alternative mouse models designed to regulate SNARE-dependent exocytosis, as well as optogenetic or chemogenetic strategies for precise astrocyte manipulation during t-LTD induction. This would provide more direct evidence of the influence of astrocytic activity on synaptic plasticity.

We thank the reviewer for the suggestion. As stated in the manuscript and in figure 4, we already used two different approaches (aBAPTA to interfere with astrocyte calcium signalling and dnSNARE mice (that have vesicular release impaired) to determine the involvement of astrocytes in the discovered forms of LTD, and both approaches clearly indicated the requirement of astrocytes for t-LTD. In BAPTA-treated astrocytes and in dnSNARE mice, t-LTD was prevented. Notwithstanding this, and as suggested by the reviewer, we used two additional approaches to confirm astrocyte participation. We loaded astrocytes with the light chain of the tetanus toxin (TeTxLC), which is known to block exocytosis by cleaving the vesicle-associated membrane protein, an important part of the SNARE complex (Schiavo et al., 1992, Nature 359, 832-835). In this experimental condition, we observed a clear lack of t-LTD at both (lateral and medial) pathways, thus confirming the requirement of astrocytes and the SNARE complex and vesicular release for both types of t-LTD. In addition, to gain more insight into the fact that glutamate is released by astrocytes, we blocked glutamate release from astrocytes by loading the astrocytes with Evans blue, known to interfere with glutamate uptake into vesicles as it inhibits the vesicular glutamate transporter (VGLUT). In this experimental condition, again t-LTD was prevented, indicating that t-LTD requires Ca2+dependent exocytosis of glutamate from astrocytes.

Reviewer #2 (Public Review):

Summary:

This work reports the existence of spike timing-dependent long-term depression (t-LTD) of excitatory synaptic strength at two synapses of the dentate gyrus granule cell, which are differently connected to the entorhinal cortex via either the lateral or medial perforant pathways (LPP or MPP, respectively). Using patch-clamp electrophysiological recording of tLTD in combination with either pharmacology or a genetically modified mouse model, they provide information on the differences in the molecular mechanism underlying this t-LTD at the two synapses.

Strengths:

The two synapses analyzed in this study have been understudied. This new data thus provides interesting new information on a plasticity process at these synapses, and the authors demonstrate subtle differences in the underlying molecular mechanisms at play. Experiments are in general well controlled and provide robust data that are properly interpreted.

We thank the reviewer for the positive assessment of our work and the constructive suggestions to improve the manuscript.

Weaknesses:

• Caution should be taken in the interpretation of the results to extrapolate to adult brain as the data were obtained in P13-21 days old mice, a period during which synapses are still maturing and highly plastic.

We thank the reviewer for noticing this. In fact, our experiments were intentionally performed in young animals (P13-21), just knowing that this is a critical period of plasticity. We indicate that in the methods, results, and discussion (where we discuss that in some detail) sections.

• In experiments where the drug FK506 or thapsigargin are loaded intracellularly, the concentrations used are as high as for extracellular application. Could there be an error of interpretation when stating that the targeted actors are necessarily in the post-synaptic neuron? Is it not possible for the drug to diffuse out of the cell as it is evident that it can enter the cell when applied extracellularly?

We thank the reviewer for rising this point. While it would be possible that these compounds cross the cell membranes, to do it and to pass to other cells, this would, in principle, require a relatively long time to occur. Additionally, to have any effect, the same concentration or a relatively high concentration of that we put into the pipette has to reach other cells. Furthermore, even if a compound is able to cross a cell membrane during the duration of an experiment, after this, it may be exposed to the extracellular fluid where will be diluted and most probably washed out. For all these reasons, we do not see this very plausible. Notwithstanding this, and as suggested, we have repeated the experiments using lower concentrations of thapsigargin (1 uM) and FK506 (1 uM), and have obtained the same results. These data are now included in the figure 3 and in the text.

• The experiments implicating glutamate release from astrocytes in t-LTD would require additional controls to better support the conclusions made by the authors. As the data stand, it is not clear, how the authors identified astrocytes to load BAPTA and if dnSNARE expression in astrocytes does not indirectly perturb glutamate release in neurons.

We thank the reviewer for rising this point. We now indicate how astrocytes have been identified to load BAPTA. We reply to this in detail in the “Recommendations for the authors” from reviewer 2.

Significance:

While this is the first report of t-LTD at these synapses, this plasticity process has been mechanistically well investigated at other synapses in the hippocampus and in the cortex. Nevertheless, this new data suggests that mechanistic differences in the induction of t-LTD at these two DG synapses could contribute to the differences in the physiological influence of the LPP and MPP pathways.

Reviewer #3 (Public Review):

Coatl et al. investigated the mechanisms of synaptic plasticity of two important hippocampal synapses, the excitatory afferents from lateral and medial perforant pathways (LPP and MPP, respectively) of the entorhinal cortex (EC) connecting to granule cells of the hippocampal dentate gyrus (DG). They find that these two different EC-DG synaptic connections in mice show a presynaptically expressed form of long-term depression (LTD) requiring postsynaptic calcium, eCB synthesis, CB1R activation, astrocyte activity, and metabotropic glutamate receptor activation. Interestingly, LTD at MPP-GC synapses requires ionotropic NMDAR activation whereas LTD at LPP-GC synapse is NMDAR independent. Thus, they discovered two novel forms of t-LTD that require astrocytes at EC-GC synapses. Although plasticity of EC-DG granule cell (GC) synapses has been studied using classical protocols, These are the first analysis of the synaptic plasticity induced by spike timing dependent protocols at these synapses. Interestingly, the data also indicate that t-LTD at each type of synapse require different group I mGluRs, with LPP-GC synapses dependent on mGluR5 and MPP-GC t-LTD requiring mGluR1.

The authors performed a detailed analysis of the coefficient of variation of the EPSP slopes, miniature responses and different approaches (failure rate, PPRs, CV, and mEPSP frequency and amplitude analysis) they demonstrate a decrease in the probability of neurotransmitter release and a presynaptic locus for these two forms of LTD at both types of synapses. By using elegant electrophysiological experiments and taking advantage of the conditional dominant-negative (dn) SNARE mice in which doxycycline administration blocks exocytosis and impairs vesicle release by astrocytes, they demonstrate that both LTD forms require the release of gliotransmitters from astrocytes. These data add in an interesting way to the ongoing discussion on whether LTD induced by STDP participates in refining synapses potentially weakening excitatory synapses under the control of different astrocytic networks. The conclusions of this paper are mostly well supported by data, but some aspects the results must be clarified and extended.

We thank the reviewer for the positive assessment of our work and the constructive suggestions to improve the manuscript.

(1) It should be clarified whether present results are obtained with or without the functional inhibitory synapse activation. It is not clear if GABAergic synapses are blocked or not. If GABAergic synapses are not blocked authors must discuss whether the LTD of the EPSPs is due to a decrease in glutamatergic receptor activation or an increase in GABAergic receptor activation. Moreover, it should be recommended to analyze not only the EPSPs but also the EPSCs to address whether the decrease in synaptic transmission is caused by a decrease in the input resistance or by a decrease in the space constant (lambda).

We thank the reviewer for rising these points. GABAergic inhibition was not blocked in our experiments. The observed forms of t-LTD seem to be due to a decrease in glutamate release probability as indicated in the manuscript, mediated by the mechanism we uncover and describe here. To determine and clarify whether GABA receptors have any role in these forms of t-LTD, we repeated the experiments in the presence of the GABAA and GABAB receptors antagonists bicuculline and SCH50911, respectively. Blocking GABA receptors do not prevent or affect t-LTD at LPP- or MPP-GC synapses, that is still present and with a similar magnitude that controls. These results indicating that these receptors are not involved in these forms of t-LTD. These results are now included in the text in the results section (page 8) and as a new figure S1. In our experiments, no changes in input resistance or space constant were observed, and importantly, no changes were observed in the amplitude/slopes of EPSP in the control pathway that does not undergo plasticity protocol that we routinely use in our experiments.

(2) Authors show that Thapsigargin loaded in the postsynaptic neuron prevents the induction of LTD at both synapses. Analyzing the effects of blocking postsynaptic IP3Rs (Heparin in the patch pipette) and Ryanodine receptors (Ruthenium red in the patch pipette) is recommended for a deeper analysis of the mechanism implicated in the induction of this novel forms of LTD in the hippocampus.

We thank the reviewer for this suggestion. We repeated the experiments loading the postsynaptic cell with heparin and ruthenium red using the path pipette. In these experimental conditions, we observed that t-LTD was not affected by the heparin treatment (discharging a role of IP3Rs), but that it was prevented by the ruthenium red treatment (indicating the requirement of ryanodine receptors). We include now this data in the text (page 12) and in the Figure 3a, b, e, f.

(3) Authors nicely demonstrate that CB1R activation is required in these forms of LTD by blocking CB1Rs with AM251, however an interesting unanswered question is whether CB1R activation is sufficient to induce this synaptic plasticity. This reviewer suggests studying whether applying puffs of the CB1R agonist, WIN 55,212-2, could induce these forms of LTD.

We thank the reviewer for this suggestion. We repeated the experiments adding WIN55, 212-2 as suggested.  The activation of CB1R by puffs of the agonist WIN 55, 212-2 to the astrocyte, directly induced LTD at both LPP- and MPP-GC synapses. We include now this data in the text (page 14) and in the Figure 3c, d, g, h.

(4) Finally, adding a last figure with a cartoon summarizing the proposed model of action in these novel forms of LTD would add a positive value and would help the reading of the manuscript, especially in those aspects related with the discussion of the results.

We thank the reviewer for the suggestion. We include now a figure showing the proposed mechanisms (Figure 5).

The extension of these results would improve the manuscript, which provides interesting results showing two novel forms of presynaptic t-LTD in the brain synapses with different action mechanisms probably implicated in the different aspects of information processing.

Recommendations for the authors:

Reviewer #1 (Recommendations For The Authors):

There are just a few aspects that could be clarified to bolster the authors' conclusions.

The author centered the conclusion of their study on the role of astrocytic activity in regulating these two forms of plasticity (see title). To strengthen the evidence that astrocytes are key regulators of t-LTD at MPP and LPP GC synapses by regulating SNARE protein-dependent glutamate release, additional complementary approaches should be considered, such as other mouse models enabling the control of SNARE-dependent exocytosis and/or optogenetic/chemogenetic tools to selectively manipulate astrocytes during the induction of t-LTD, thereby directly assessing the impact of astrocytic activity on synaptic plasticity. Implementing calcium imaging or glutamate sensors to visualize the dynamics of astrocytic calcium signaling and glutamate release during t-LTD could be also considered.

We thank the reviewer for the suggestion. As stated in the manuscript and in figure 4, we already used two different approaches (aBAPTA to interfere with astrocyte calcium signalling and dnSNARE mice (that have vesicular release impaired) to determine the involvement of astrocytes in the discovered forms of LTD, and both approaches clearly indicated the requirement of astrocytes for t-LTD. In BAPTA-treated astrocytes and in dnSNARE, t-LTD was prevented. Notwithstanding this, and as suggested by the reviewer, we used two additional approaches to confirm astrocytes participation. We loaded astrocytes with the light chain of the tetanus toxin (TeTxLC), which is known to block exocytosis by cleaving the vesicle-associated membrane protein, an important part of the SNARE complex (Schiavo et al., 1992, Nature 359, 832-835). In this experimental condition, we observed a clear lack of t-LTD at both (lateral and medial) pathways, thus confirming the requirement of astrocytes and the SNARE complex and vesicular release for both types of t-LTD. In addition, to gain more insight into the fact that glutamate is released by astrocytes, we blocked glutamate release from astrocytes by loading the astrocytes with Evans blue, known to interfere with glutamate uptake into vesicles as it inhibits the vesicular glutamate transporter (VGLUT). In this experimental condition, again t-LTD was prevented, indicating that t-LTD requires Ca2+-dependent exocytosis of glutamate from astrocytes. This information is now included in the text, pages 14 and 15 and in figure 4.

• How were astrocytes identified to be loaded with BAPTA? The author should clarify this methodological aspect and provide confocal images of patched astrocytes situated 50-100 um from the recorded neuron.

We thank the reviewer for the comment. We include now this information in the Methods section (page 6) and in figure S3. Astrocytes were identified by their rounded morphology under differential interference contrast microscopy, and were characterized by low membrane potential, low membrane resistance and passive responses (they do not show action potentials) to both negative and positive current injection.

• Please provide confocal images of EGFP expression in the DG astrocytes of dnSNARE mice both on and off Dox, to verify transgene expression in astrocytes

We thank the reviewer for this suggestion. We now include an image of GFP expression in the DG astrocytes of off Dox dnSNARE mice. We did not provide the animals with doxycycline since birth and thus the gene was constantly expressed. We now show this image in Fig. S3. All the pups and mice are not DOX fed, meaning that the transgenes are continuously being expressed and therefore the exocytosis should be blocked in astrocytes.

Minor points:

Lines 250-253: It is mentioned that TTX is added at baseline, washed out for the t-LTD experiment, and then reapplied post t-LTD. I suggest clarifying the timing and rationale for this application for a broad audience.

We thank the reviewer for the suggestion. We now include some information related to the timing and rationale of the experiment phases (page 9).

The discussion is quite detailed and provides a comprehensive overview of the study's findings. To enhance clarity and impact, the authors might consider to,

• add subheadings and bullet points for key findings. This will improve readability.

• this section could benefit from streamlining to avoid redundancy.

• some sentences could be made more concise without losing meaning.

We thank the reviewer for these suggestions. We now include subheadings in the discussion section to improve readability and have made some sentences more concise and simple without losing meaning.

In figure legends, consistency with capitalization should be maintained, for example in the statistical significance notation, ***P < 0.001" or ***p < 0.001")

We now include p<0.001 in the figure legend 4 for consistency.

Reviewer #2 (Recommendations For The Authors):

Major:

• All results were obtained in young still quite immature synapses. To strengthen the significance of the findings, the authors could repeat some of the main experiments in adult mice (8 weeks and beyond). If not, they should state clearly that these mechanisms were only evidenced in early post-natal conditions.

We thank the reviewer for noticing this. In fact, our experiments were intentionally performed in young animals (P13-21), just knowing that this is a critical period of plasticity. As the reviewer suggests, we indicate that in the methods (page 5), results (page 8), and discussion (page 19) (where we discuss that in some detail) sections.

• Lines 246-249 and fig 1f,p: Authors need to perform a statistical test on these two graphs to support their claim that 'A plot of CV-2 versus the change in the mean evoked EPSP 246 slope (M) before and after t-LTD mainly yielded points below the diagonal line at LPP-GC and MPP-GC synapses'.

That could not be clear in the previous version. We observed an error in the points (with some points missing) of one of the graphs that we have corrected. In addition, and as suggested by the reviewer we performed a regression analysis that confirms the conclusions stated. This is now included in the text (page 9). Thus, we have added information about mean values ± SEM in the text and the linear regression of the data for LPP-GC (Mean = 0.607 ± 0.054 vs 1/CV2 = 0.439 ± 0.096, R2 = 0.337; n = 14) and MPP-GC synapses (Mean = 0.596 ± 0.056 vs 1/CV2 = 0.461 ± 0.090, R2 = 0.168; n = 13), respectively. Data yielded on the dotted horizontal line, 1/CV2 = 1, indicates no change in the probability of release, in contrast, data yielded below the dotted diagonal line is suggestive of a change in the probability of release parameters (for review, see Brock et al., 2020, Front Synaptic Neurosci 12, 11).

• We are not sure that the experiment with the MK801 provided in the patch pipet can be interpreted correctly (Figure 2 a,b and e,f). How sure are the authors that, when applying MK801 in the patch pipet, it can reach its binding site within the pore? The concentration of MK801 is also very high (500 microM) and used at the same concentration extracellularly and intracellularly. Why did the authors not use lower concentration when applied intracellularly?

We thank the reviewer for rising this point. MK801 in the pipette is reaching the pore when loaded postsynaptically as when we record NMDA currents from postsynaptic neurons loaded with MK801, these currents are blocked. We include now a control experiment showing the effect of postsynaptic MK801 on NMDA current in the text (page 10). NMDA currents has been recorded at +40 mV, blocking AMPAR and GABAR with NBQX and bicuculline. Related to the concentration, it has been described that the affinity from the internal site is much lower (several orders of magnitude) than from the extracellular side(Sun et al., 2018 Neuropharmacology, 143, 122-129) and the concentrations used have been extensively used in previous studies. It is clear that the concentrations used in the present work blocked NMDAR currents but did not prevent LTD.

• Linked to the point above, for the intracellular application of FK506 and thapsigargin, the concentrations used extracellularly and intracellularly are identical. The authors could have used lower concentrations for the intracellular application. Also, how can they be sure of the correct interpretation of these data as the drug essentially reaching a post-synaptic target when applied intracellularly? If the drug can enter the neuron, why could it not diffuse out of the neuron especially when loaded at a high concentration? Maybe using a lower concentration when applied intracellularly could at least partially address this issue.

It is evident that it can enter the cell when applied extracellularly?

We thank the reviewer for rising this point. While it would be possible that these compound cross the cell membranes, to do it and to pass to other cells, this would, in principle, require a relatively long time to occur. Additionally, to have any effect, the same concentration or a relatively high concentration of that we put into the pipette has to reach other cells. Furthermore, even if a compound is able to cross a cell membrane during the duration of an experiment, after this, it may be exposed to the extracellular fluid where it will be diluted and most probably washed out. For all these reasons, we do not see this very plausible. Notwithstanding this, we have repeated the experiments using lower concentrations of thapsigargin (1 uM) and FK506 (1 uM) and have obtained the same results. These data are now included in the figure 3 and the numbers in the text have been updated (pages 12-13).

• The data supporting the possibility of glutamate release by astrocytes as a main source of glutamate to promote t-LTD needs to be strengthened. In experiment Figure a-h, it is not clear how the authors recognize astrocytes to patch. No details are provided in the methods or in the main text. If we understand correctly, it is only by performing a current steps protocol to ensure that the patched cell did not produce action potentials. If this was the case, the authors need to be more specific and provide details of this protocol. More importantly, the one trace that was provided in Figures 4a and 4f suggests, albeit by a rough estimation that we made with a ruler, that the highest current step only depolarized the cell to about -40 mV. This is not sufficient to ensure that the recorded cell is not a neuron. The authors should increase their steps to high depolarizing currents to ensure that the patched cell is not a neuron. Better yet, they should load the cell with an dye to process the slice after the electrophysiological recording for immunohistochemistry to ensure that it was indeed an astrocyte. Alternatively, they can try to aspirate the cell content at the end of the recording to perform a qPCR for astrocyte markers eg. GFAP.

We thank the reviewer for the comment. We include now information regarding how astrocytes were identified (also raised by reviewer 1) in the Methods section (page 6) and in figure S3. Astrocytes were identified by their rounded morphology under differential interference contrast microscopy, eGFP fluorescence (astrocytes from dnSNARE mice), and were characterized by low membrane potential, low membrane resistance and passive responses (they do not show action potentials) to both negative and positive current injection.

We agree with the reviewer that in figure 4a and 4f, the step protocol might not be completely clear. For this, we revised that and now include in a clearer way that we applied pulses that depolarized astrocytes beyond -20 mV, with no action potentials found at any point. We also include now this in figure S3.

• Related to the point above, the use of the model expressing dnSNARE in astrocytes is elegant. Yet, to really interpret the data obtained in these slices as a lack of vesicle release (and most importantly glutamate) we think that the authors should ensure that glutamate release from nearby neurons is not impacted. They could patch nearby neurons in dnSNARE slices and test PPR or synaptic fatigue when stimulating either the LPP or MPP. The authors should avoid overinterpretation of these results. As it stands, it is not evident that dnSNARE expression does not perturb other mechanisms within the astrocyte that in turn perturb pre-synaptic glutamate release. Adding back glutamate as puffs does not help to disentangle this issue.

To gain more insight into the fact that glutamate is released by astrocytes we blocked glutamate release from astrocytes by loading the astrocytes with Evans blue, known to interfere with glutamate uptake into vesicles as it inhibits the vesicular glutamate transporter (VGLUT). In this experimental condition, as indicated above, t-LTD was prevented, indicating that t-LTD requires Ca2+-dependent exocytosis of glutamate from astrocytes. This is included in the text (page 15) and in figure 4d,e, i, j.

In addition, we loaded astrocytes with the light chain of the tetanus toxin (TeTxLC) which is known to block exocytosis by cleaving the vesicle-associated membrane protein, an important part of the SNARE complex (Schiavo et al., 1992, Nature 359, 832-835). In this experimental condition, we observed a clear lack of t-LTD at both (lateral and medial) pathways, thus confirming the requirement of astrocytes and the SNARE complex and vesicular release for both types of t-LTD. These data indicate that t-LTD requires Ca2+-dependent exocytosis of glutamate from astrocytes. This information is now included in the text, page 14 and in figure 4.

Minor points:

• line 107, did the authors mean t-LTP and t-LTD? we don't understand STDP mentioned here.

We meant to say t-LTP. This is now corrected.

• line 108: should STDP be replaced by t-LTD as the authors only focused on this plasticity mechanism.

We agree, we indicate now t-LTD.

• line 131-132 : it is not clear when the animals were fed with doxycycline. If it was from birth, then the 'not' should be removed. Otherwise the authors should clearly state when the doxycyline was provided.

DOX was not provided and that means that the transgene was continuously expressed and therefore the exocytosis should be blocked in astrocytes. We express that clearer in page 5, methods section.

• line 223 : which hippocampal synapses? needs to be stated

As suggested this is now included in the text as for cortical synapses. Synapses are Schaffer collaterals SC-CA1 for hippocampus and layer L4-L2/3 for cortical synapses (page 8).

• line 273: what do the authors mean when writing 'from'? We don't understand the data provided on this line.

We thank the reviewer for noticing this. That refers to the amplitude of NMDAR-mediated currents average before and after D-AP5 or MK801. We express this now in a clearer way (page 10, from 57±8 pA to 6±5 pA).

• line 286 : why do the authors point out work on GluN2B and GluN3A only here when they first investigate GluN2A contribution to t-LTD? what about previous data on GluN2A?

We have now expressed this in a different way to make it clear. We wanted to indicate that the available data for presynaptic NMDAR at MPP-GC synapses has been indicated to contain GluN2B and GluN3A subunits and to our knowledge, no data indicate that they contain GluN2A subunits.

• line 428 : what do the authors mean by 'not least' ?

This is a typo and we have removed that from the text.

Reviewer #3 (Recommendations For The Authors):

My only suggestion for improving data presentation in the manuscript would be to split some figures of the paper. In my opinion, the figures are too dense and therefore difficult to follow for the broad audience of eLife readers. In addition, a real image of the recorded dentate granule cells in the slice showing also the location of the real stimulation electrodes would significantly improve the presentation of Figure 1.

We thank the reviewer for the suggestion, but we would prefer to let the figures as they are organized, as while we agree in some cases they are a bit big, in this way it is easier to compare lateral and medial pathways. For this, it could be better to let information regarding the two pathways in the same figure. Nevertheless, we try now to make figures clearer to use a columnar organization of the figures for each pathway what we think, would make easier to compare pathways. As the reviewer suggests we include now a real image of the recorded dentate granule cells in the slice showing also the location of the real stimulation electrodes in Figure 1, that we agree will improve the presentation of this figure and thank the reviewer for the suggestion.

Author response:

The following is the authors’ response to the original reviews.

The reviewers found this manuscript to present convincing evidence for associative and non-associative behaviors elicited in male and female mice during a serial compound stimulus Pavlovian fear conditioning task. The work adds to ongoing efforts to identify multifaceted behaviors that reflect learning in classic paradigms and will be valuable to others in the field. The reviewers do note areas that would benefit from additional discussion and some minor gaps in data reporting that could be filled by additional analyses or experiments.

We thank the reviewers and the editors for their thoughtful and constructive critiques of our manuscript. We have updated our manuscript with data from additional experiments as suggested by the reviewers, and we have significantly edited the text and figures to reflect these additions. Our detailed, point-by-point responses are below.

Reviewer #1 (Public Review):

The main goal of the study was to tease apart the associative and non-associative elements of cued fear conditioning that could influence which defensive behaviors are expressed. To do this, the authors compared groups conditioned with paired, unpaired, or shock only procedures followed by extinction of the cue. The cue used in the study was not typical; serial presentation of a tone followed by a white noise was used in order to assess switches in behavior across the transition from tone to white noise. Many defensive behaviors beyond the typical freezing assessments were measured, and both male and female mice were included throughout. The authors found changes in behavioral transitions from freezing to flight during conditioning as the tone transitioned into white noise, and a switch in freezing during extinction such that it became high during the white noise as flight behavior decreased. Overall, this was an interesting analysis of transitions in defensive behaviors to a serially presented cue consisting of two auditory stimuli during conditioning and then extinction.

We thank the Reviewer for their supportive insight.

There are some concerns regarding the possibility that the white noise is more innately aversive than the tone, inducing more escape-like behaviors compared to a tone, especially since the shock only group also showed increased escape-like behaviors during the white noise versus tone. This issue would have been resolved by adding a control group where the order of the auditory stimuli was reversed (white noise->tone).

We appreciate this concern, and we have added two additional groups to address this possibility. We have conducted the same experimental paradigm with 2 reverse-SCS groups (WN—tone), one with paired (new PA-R group), and one with unpaired (new UN-R group), presentations to shock during conditioning. These experiments revealed that during conditioning day 2 in both reverse order groups, WN causes reductions in freezing and increases in locomotor activity (see revised Figure 2D), an effect that is stronger in the UN-R compared to the PA-R group. This locomotor effect is neither darting nor escape jumping in the PA-R group (revised Figure 3G, I; Figure 4G). In the UN-R group, WN induces more activity than the PA-R group (Figure 2D), including some jumping at WN onset (Figure 3H), but no darting (Figure 4G). It is worth noting that WN does not elicit defensive behavior before conditioning at the sound intensity we use (75dB; see Fadok et al. 2017, Borkar et al. 2020, Borkar et al. 2024). Together, these results suggest that WN is an inherently more salient stimulus than tone, and it can elicit defensive behaviors in shock-sensitized mice through non-associative mechanisms. Indeed, stimulus salience is a key factor in this paradigm for inducing activity (see Hersman et al. 2020).

While the more complete assessment of defensive behaviors beyond freezing is welcomed, the main conclusions in the discussion are overly focused on the paired group and the associative elements of conditioning, which would likely not be surprising to the field. If the goal, as indicated in the title, was to tease apart the associative and non-associative elements of conditioning and defensive behaviors, there needs to be a more emphasized discussion and explicit identification of the non-associative findings of their study, as this would be more impactful to the field.

We have rewritten the Discussion to provide a greater emphasis on the findings of the study that are more related to non-associative mechanisms. For example, we argue that cue-salience and changes in stimulus intensity can induce non-associative increases in locomotor behavior and tail rattling in shock-sensitized mice.

Reviewer #2 (Public Review):

Summary:

The authors examined several defensive responses elicited during Pavlovian conditioning using a serial compound stimulus (SCS) as the conditioned stimulus (CS) and a shock unconditioned stimulus (US) in male and female mice. The SCS consisted of tone pips followed by white noise. Their design included 3 treatment groups that were either exposed to the CS and US in a paired fashion, in an unpaired fashion, or only exposed to the shock US. They compared freezing, jumping, darting, and tail rattling across all groups during conditioning and extinction. During conditioning, strong freezing responses to the tone pips followed by strong jumping and darting responses to the white noise were present in the paired group but less robust or not present in the unpaired or shock only groups. During extinction, tone-induced freezing diminished while the jumping was replaced by freezing and darting in the paired group. Together, these findings support the idea that associative pairings are necessary for conditioned defensive responses.

Strengths:

The study has strong control groups including a group that receives the same stimuli in an unpaired fashion and another control group that only receives the shock US and no CS to test the associative value of the SCS to the US. The authors examine a wide variety of defensive behaviors that emerge during conditioning and shift throughout extinction: in addition to the standard freezing response, jumping, darting, and tail rattling were also measured.

We thank the Reviewer for their supportive appraisal of this study’s strengths.

Weaknesses:

This study could have greater impact and significance if additional conditions were added (e.g., using other stimuli of differing salience during the SCS), and determining the neural correlates or brain regions that are differentially recruited during different phases of the task across the different groups.

In the revised manuscript, we have conducted experiments with 2 reverse-SCS groups (WN—tone): one with paired (new PA-R group), and one with unpaired (new UN-R group), presentations to shock during conditioning. These experiments revealed that during conditioning day 2 in both reverse order groups, WN causes reductions in freezing and increases in locomotor activity (see revised Figure 2D), an effect that is stronger in the UN-R compared to the PA-R group. This locomotor effect is neither darting nor escape jumping in the PA-R group (revised Figure 3G, I; Figure 4G). In the UN-R group, WN induces more activity than the PA-R group (Figure 2D), including some jumping at WN onset (Figure 3H), but no darting (Figure 4G). Indeed, stimulus salience is a key factor in this paradigm for inducing activity (see Hersman et al. 2020). Together, these results suggest that WN is an inherently more salient stimulus than tone, and it can elicit defensive behaviors in shock-sensitized mice through non-associative mechanisms. It is worth noting that WN does not elicit defensive behavior before conditioning at the sound intensity we use (75dB; see Fadok et al. 2017, Borkar et al. 2020, Borkar et al. 2024).

We agree that determining the neuronal correlates and brain regions that are involved in defensive ethograms at various stages within this paradigm is of great importance, but we feel that those experiments are beyond the scope of the current study, which is focused on identifying behavioral differences based on associative and non-associative factors.

Reviewer #1 (Recommendations For The Authors):

In LINES 72-73, authors say they used a "truly random procedure" as one of their control groups. Then in LINES 113-116, they describe this group as "unpaired" where the "SCS could not reliably predict footshock". Combined, it is unclear if this group is random or unpaired. The "truly random procedure" is defined, by the cited Rescorla paper, as "the two events are programmed entirely randomly and independently in such a way that some "pairings" of CS and US may occur by chance alone". So, truly random would indicate that the shock may occur during the cue, while unpaired indicates the shock was explicitly unpaired from the cue. If the authors used a random procedure, the groups need to be labeled as random, not unpaired, and the # of cues that happened to coincide with footshock per animal needs to be reported somewhere. If the authors used an unpaired procedure (which appears to be the case based on 40-60s ITI between SCS and footshock being reported), it needs to be clearer and consistent throughout that it was explicitly unpaired, as well as removing the claim in LINE 72-73 that they used a "truly random procedure".

We did indeed use an explicitly unpaired procedure. We have adjusted the text and figures to better reflect this, and we removed any mentions of randomness with regards to the presentations of SCS and footshock.

Despite the lack of significant sex differences, it would still be helpful if data panels with individual data points (e.g. Fig 2E-J), were presented as identifiable by sex (e.g. closed vs open circles for males vs females).

The revised manuscript now compares four or five groups per figure, making data presentation complicated. Providing the individual data points in each panel reduces figure clarity, therefore, we feel it is best to present the data as box-and-whisker plots without them. However, the source data files for each figure are available to the reader and the data are clearly labeled to be identifiable by sex.

Is it not odd that all groups showed similar levels of contextual freezing during the 3min baseline? If shocks are unsignaled in the UN and SO groups, one would expect higher levels of contextual freezing compared to a paired group.

We are not certain why one would expect higher levels of contextual freezing in the UN and SO groups compared to the PA group at the beginning of conditioning day 2. Another study also looked at baseline freezing in a contextual fear group (which is the same as shock only in our study) and in an auditory cued fear conditioning group within the conditioning context, and their data show that freezing during the baseline period is equivalent between groups (Sachella et al., 2022).

During baseline on Extinction Day 1, it does seem that the unpaired and SO groups tend to have higher freezing levels compared to the paired groups. Author response image 1 shows baseline freezing during the first 3 minutes of extinction day 1. After two days of conditioning in the conditioned flight paradigm, contextual freezing either is, or trends to be significantly higher in the UN, UN-R, and SO groups than the PA and PA-R groups.

Author response image 1.

Baseline Freezing levels for all groups during the first extinction session. Baseline period is defined as the first 180 seconds of the session, before any auditory stimulus was presented. PA, Paired; UN, Unpaired; SO, Shock Only; PA-R, Paired Reverse; UN-R, Unpaired Reverse. *p<0.05, **p<0.01, ****p<0.0001.

Do the tone and WN elicit similar levels of defensive behaviors in a naïve mouse? Or have the authors tested WN followed by tone? Is there a potential issue that the WN may be innately aversive which is then amplified with training? i.e. does a tone preferentially induce freezing while WN induces active behaviors, regardless of which sensory stimulus is temporally closer to the shock? If the change in behavior is really due to the pairing and temporal proximity to shock, then there should be increased jumps, etc to the tone if trained with WN->tone.

WN can indeed be used as an aversive stimulus under certain conditions and at sufficiently high decibel levels. In the conditioned flight paradigm, WN is presented at 75dB, which is below the threshold for eliciting an acoustic startle response in a C57BL/6J mouse (Fadok et al. 2009). Also, during pre-exposure, when animals are naïve to the SCS, tone and WN stimuli do not elicit defensive behaviors (see Fadok et al. 2017, Borkar et al. 2020, 2024).

As suggested by the Reviewer, during revision we have included reverse-SCS paired (PA-R) and unpaired (UN-R) groups to test for the role of stimulus salience and stimulus order on defensive ethograms. During conditioning day 2, the PA-R group exhibited little freezing to the WN, with a slightly elevated activity index, and they exhibited robust freezing during tone (revised Figure 2A-H). The activity during the WN in the PA-R group was significantly lower than that of the PA group (Figure 2L). The PA-R group also did not respond to WN with escape jumps or darting (Figure 3I, 4G). The UN-R group displayed greater activity during the WN than the UN and PA-R groups, but less activity than the PA group (Figure 2D, H). The UN-R group did not dart but this group displayed some jumping at WN onset (Figure 3H), like what was observed in the UN group.

These data suggest that WN has inherent, salient properties that can induce some non-associative activity after the mouse has been sensitized by shock (see also Hersman et al. 2020 for more detailed analysis of stimulus salience in the conditioned flight paradigm). However, only in the PA group is robust flight behavior (comprised of high numbers of escape jumps and darting) observed. Therefore, both stimulus salience and temporal order are important for eliciting transitions from freezing to flight.

Fig 3G/4G are hard for me to understand. The figure legends say they're survival graphs but the y-axis labels "Latency to initial jump/dart (% of cohort)" confuses me. What is the purpose of these graphs? Perhaps they are not needed. Or consider presenting them similar to Fig 7C, D as those were more intuitive and faster for me to grasp.

We had intended these plots to show that a greater proportion of the paired group jumps and darts during WN compared to the unpaired group, and that the percentage of the cohort that jumps and darts increases across conditioning trials. Because these graphs were not clear, we have removed them, and we have replaced them with graphs comparing total cohort percentages that jumped (Figure 3I) or darted (Figure 4G) over the whole CD2 session.

For the extinction data, I did not see within group analyses for within or between session fear extinction to the tone. So, for the paired group, were the last 4 trials of Ext 1 significantly lower than the first 4 trials? If not, then they did not show within-session extinction. Also, for the paired group, were the last 4 trials of Ext 1 significantly different than the first 4 trials of Ext 2? This would test for long-term retention and spontaneous recovery.

In the original submission and in the revised manuscript, we calculated a delta change score for freezing during tone in the early versus late blocks of 4 trials, and then we statistically compared these differences across groups (Figure 5C, D). This allowed us to assess between-group differences in changes to tone-evoked freezing during extinction. Freezing to tone did decrease significantly over the first extinction session for the paired group (Early Ext1 vs Late Ext1, paired t-test, t(31) \= 6.23, p<0.0001), and when comparing late Ext1 and early Ext2, we found that tone-evoked freezing did significantly increase (Late Ext1 vs Early Ext2, paired t-test, t(31) \= 5.26, p<0.0001). This increase in cue-induced freezing between days of extinction is characteristic of C57BL/6J mice (Hefner et al., 2008). Our study did not test for more distal timepoints, so we cannot comment on the efficacy of long-term retention or spontaneous recovery.

For the conditioning and extinction data across Figs 2, 5 and 6, what I gather from them is that freezing is high to the tone and low to the WN during conditioning, and then low to the tone, and high to the WN across extinction. Then for activity levels I see they are low to the tone and high to the WN during conditioning, and then low to the WN during extinction. The piece that is missing is what are activity levels like to the tone during extinction. Are they low like in conditioning and remain low in extinction? Or do they increase across extinction as freezing decreases? As I was going through these graphs I drew myself out step function summaries of the freezing and activity levels between tone/WN for conditioning vs extinction; maybe the authors could consider a summary figure.

We thank the Reviewer for their interest. We found that within the paired group, activity to tone remained low throughout both days of extinction (though increased within each session) and did not return to normal activity levels. We present this data in Author response image 2. We thank the Reviewer for the suggestion of a summary figure, but we feel there are too many axes of classification (between-group, within-group, multiple behaviors, tone/WN, conditioning/extinction) to coherently present our findings in a single figure.

Author response image 2.

Trial-by-trial plot of activity index during the tone period of SCS across both extinction sessions for the PA group. SCS, Serial compound stimulus; Ext, extinction; PA, Paired.

In the discussion (LINE 592-3), they discuss that shock sensitization in the SO group may prime a stressed animal to dart more readily to WN upon stimulus transition. Should this not also happen during the transition of silence to tone? What is special about a transition between two auditory stimuli that would result in panic like behavior in an animal that only received shock presentations? This also gets back to an earlier concern above regarding the potentially innately aversiveness of the WN.

After 2 days of shock sensitization, we observe that mice exhibit freezing to the tone during the first three trials of extinction day 1 (Figure 5A). This non-associative freezing response is like that observed in other studies of non-associative fear processing (please see Kamprath and Wotjak, 2004). As trials progress during extinction day 1, mice do become mildly activated during the tone (Author response image 3). The transition to WN in the shock-only group during extinction induces non-associative darting responses, but it does not induce escape jumping behavior (Figure 7).  We hypothesize that the innate salience of the WN is a vital factor contributing to these escalated responses. The importance of stimulus salience in conditioned flight was also demonstrated by Hersman et al., 2020 for SCS conditioning, and by Furuyama et al., 2023 for single tone conditioning.  Just as with conditional freezing responses (Kamprath and Wotjak, 2004), we believe that conditional flight is controlled by summative components, one being associative and the other non-associative.

Author response image 3.

Trial-by-trial plot of activity index during the tone period of SCS across both extinction sessions for the SO group. SCS, Serial compound stimulus; Ext, extinction; SO, Shock Only.

In the discussion (LINE 583), they say that the development of explosive defensive behaviors are "not achievable with traditional single-cue Pavlovian conditioning paradigms". The authors should include a caveat here that the current study did not compare their results to a group of mice that received just WN-shock pairings.

We thank the reviewer for this comment. This statement was meant to highlight that traditional paradigms do not offer an element of signaling the temporal imminence of threat, only its inevitability. It was not our intention to state that defensive escape behaviors were unachievable in single-cue conditioning paradigms, and we regret not making this clear. Indeed, the supplement of Fadok et al. 2017 shows that WN-shock conditioning is capable of inducing flight, Furuyama et al. 2023 shows that tone-shock conditioning is capable of inducing flight under specific parameters, and Gruene et al. 2015 demonstrates that single CS-US pairings induce conditional darting behaviors in female rats. We have adjusted the text to better reflect our intent.  

Minor comment to LINE 613-5: Speaking as someone who has done fear conditioning in both mice and rats, tail rattling may be specific to mice (I have seen this often) and likely not observable in rats (never seen it).

We thank the Reviewer for this information. We have adjusted our text to mainly discuss mouse-specific tail rattling.

Reviewer #2 (Recommendations For The Authors):

The research questions in this study are novel and bring new insight to the field. However, there are some issues that can be addressed to improve the overall quality of the study, namely, the reader is left wanting to know more, especially about how neural circuits contribute to these different defensive behaviors during this task. Below are some recommendations for the authors that would greatly improve the impact and significance of this study.

(1) What are the neural correlates or circuits recruited during these different defensive behaviors across the course of conditioning and extinction? How might they differ between the PA and UN groups? What differences might emerge when an animal is shifting their defensive behavior from freezing to darting, for example? Answering these questions would require intensive additional experiments, therefore more discussion of possible neural mechanisms that might be recruited during this task would be appreciated, given the scope of the subject area.

We agree that understanding the neural circuits recruited during these behaviors and across conditioning and extinction is of vital importance. We are actively working on these questions, and we have published on the role of central amygdala circuits (Fadok et al. 2017) as well as on top-down control of flight by the medial prefrontal cortex (Borkar et al. 2024). Because the current manuscript is focused on learning mechanisms influencing defensive behavior, we would prefer to focus our discussion on that, rather than speculating on possible neural mechanisms. However, we have added a statement in the Discussion (LINES 706-707) emphasizing that future studies should investigate the neuronal mechanisms contributing to threat associations and different defensive behaviors.

(2) Were any vocalizations observed during conditioning or extinction phases? If not, could you speculate how type and occurrence of vocalizations might correlate with the different defensive responses observed?

Audible vocalizations were only observed during footshock presentations (squeaks). Unfortunately, we do not have the proper specialized recording equipment to monitor the full spectrum of mouse vocalizations, especially those in the ultrasonic range. Thus, we cannot speculate on the nuances of vocalizations in mice with respect to this behavioral paradigm. To the best of our knowledge, mice have not been reported to emit specific ultrasonic calls during conditioned threat like those of rats. That said, it would be of interest to determine if mice emit different vocalizations during different defensive behaviors.

(3) The transition from freezing to flight during the SCS is thought to be due to the close proximity of threat imminence between the WN CS and shock US. What if you switched the order of the SCS stimuli to WN followed by tone stimuli? If the salience of the WN stimulus is truly driving the jumping behavior, then it would be observed even if the WN stimulus preceded the pure tone stimulus and that would bring additional evidence that it is the associative value of the stimuli rather than its salience that's driving the defensive behaviors. What do you predict you would observe in rodents that were given a WN-tone SCS paired and unpaired in the same design of this study?

As suggested by the reviewer, we collected data from reverse-SCS paired and unpaired groups and reported our findings within the manuscript. Our detailed findings are also discussed above. Overall, we find that a combination of stimulus salience and temporal proximity, and a summation of non-associative and associative mechanisms, are necessary to elicit explosive flight behavior (escape jumping and darting).

References

Borkar CD, Dorofeikova M, Le QE, Vutukuri R, Vo C, Hereford D, Resendez A, Basavanhalli S, Sifnugel N, Fadok JP (2020) Sex differences in behavioral responses during a conditioned flight paradigm. Behavioural Brain Research 389:112623.

Borkar CD, Stelly CE, Fu X, Dorofeikova M, Le QE, Vutukuri R, Vo C, Walker A, Basavanhalli S, Duong A, Bean E, Resendez A, Parker JG, Tasker JG, Fadok JP (2024) Top-down control of flight by a non-canonical cortico-amygdala pathway. Nature 625: 743-749.

Fadok JP, Krabbe S, Markovic M, Courtin J, Xu C, Massi L, Botta P, Bylund K, Müller C, Kovacevic A, Tovote P, Lüthi A (2017) A competitive inhibitory circuit for selection of active and passive fear response. Nature 542:96-100.

Furuyama T, Imayoshi A, Iyobe T, Ono M, Ishikawa T, Ozaki N, Kato N, Yamamoto R (2023) Multiple factors contribute to flight behaviors during fear conditioning. Scientific Reports 13:10402. 

Gruene TM, Flick K, Stefano A, Shea SD, Shansky RM (2015) Sexually divergent expression of active and passive conditioned fear responses in rats. eLIfe 4:e11352.

Hefner K, Whittle N, Juhasz J, Norcross M, Karlsson RM, Saksida LM, Bussey TJ, Singewald N, Holmes A (2008) Impaired Fear Extinction Learning and Cortico-Amygdala Circuit Abnormalities in a Common Genetic Mouse Strain. Journal of Neuroscience 6:8074-8085.

Hersman S, Allen D, Hashimoto M, Brito SI, Anthony T (2020) Stimulus salience determines defensive behaviors elicited by aversively conditioned serial compound auditory stimuli. elife 9:e53803. 

Kamprath K and Wotjak CT (2004) Nonassociative learning processes determine expression and extinction of conditioned fear in mice. Learning & Memory 11:770-786.

Sachella TE, Ihidoype MR, Proulx CD, Pafundo DE, Medina JH, Mendez P & Piriz J (2022) A novel role for the lateral habenula in fear learning. Neuropsychopharmacology 47:1210-1219.

Author response:

The following is the authors’ response to the current reviews.

We thank the Reviewer for all their effort and suggestions over multiple drafts. Their comments have encouraged us to read and think more deeply about the issue under discussion (BLA spiking in response to CS/US inputs), and to find the papers whose contents we think provide a potential solution. We agree that there is more to understand about the mechanisms underlying associative learning in the BLA. We offer our paper as providing a new way of understanding the role of circuit dynamics (rhythms) in guiding associative learning via STDP. As we pointed out in our response to the previous review, the issue highlighted by the Reviewer is an issue for the entire field of associative learning in BLA: our discussion of the issue suggests why the experimentally observed BLA spiking in response to CS inputs, performed in the absence of US inputs (as done in the papers cited by the Reviewer), may not be what occurs in the presence of the US. Since our explanation involves the role of neuromodulators, such as ACh and dopamine, the suggestion is open to further testing.

The following is the authors’ response to the original reviews.

Reviewer #1:

Public Review’s only objection: “Deficient in this study is the construction of the afferent drive to the network, which does elicit activities that are consistent with those observed to similar stimuli. It still remains to be demonstrated that their mechanism promotes plasticity for training protocols that emulate the kinds of activities observed in the BLA during fear conditioning.”

Recommendations for the Authors: “The authors have successfully addressed most of my concerns. I commend them for their thorough response. The one nagging issue is the unrealistic activation used to drive CS and US activation in their network. While I agree that their stimulus parameters are consistent with a contextual fear task, or one that uses an olfactory CS, this was not the focus of their study as originally conceived. Moreover, the types of activation observed in response to auditory cues, which is the focus of their study, do not follow what is reported experimentally. Thus, I stand by the critique that the proposed mechanism has not been demonstrated to work for the conditioning task which the authors sought to emulate (Krabbe et al. 2019). Frustratingly, addressing this is simple: run the model with ECS neurons driven so that they fire bursts of action potentials every ~1 sec for 30 sec, and with the US activation noncontiguous with that. If the model does not produce plasticity in this case, then it suggests that the mechanisms embedded in the model are not sufficient, and more work is needed to identify them. While 'memory' effects are possible that could extend the temporal contiguity of the CS and US, the authors need to provide experimental evidence for this occurring in the BLA under similar conditions if they want to invoke it in their model. 

(1) Fair response. I accept the authors arguments and changes. 

(2) The authors rightly point out that the simulated afferents need not perfectly match the time courses of the peripheral inputs, since what the amygdala receives them indirectly via the thalamus, cortex, etc. However, it is known how amygdala neurons respond to such stimuli, so it behooves the authors to incorporate that fact into their model. 

Quirk et al. 1997 show that the response to the tone plummets after the first 100 ms in Figs 5A and 6B. The Herry et al. 2007 paper emphasizes the transient response to tone pips, with spiking falling back to a poisson low firing rate baseline outside of the time when the pip is delivered. 

Regarding potential metabotropic glutamate activation, the stimulus in Whittington et al. 1995 was electrical stimulation at 100 Hz that would synchronously activate a large volume of tissue, which is far outside the physiological norm. I appreciate that metabotropic glutamate receptors may play a role here, but ultimately the model depends upon spiking activity for the plastic process to occur, and to the best of my knowledge the spiking activity in BLA in response to a sustained, unconditioned tone, is brief (see also Quirk, Repa, and Ledoux 1995). Perhaps a better justification for the authors would be Bordi and Ledoux 1992, which found that 18% of auditory responsive neurons showed a 'sustained' response, but the sustained response neurons appear to show much weaker responses than those with transient ones (Fig 2).  I am willing to say that their paper IS relevant to contextual fear, but that is not what the authors set out to do. 

(3) Fair response. 

(4) Very good response! 

Minor points: All points were addressed.”

We thank Reviewer 1 (R1) for the positive feedback and also for pointing out that, in R1’s opinion, there is still a nagging issue related to the activation in response to CS we modeled. In (Krabbe et al., 2019), CS is a pulsed input and US is delivered right after the CS offset. The current objection of R1 is that instead, we are modeling CS and US as continuous and overlapping. R1 suggested that we add the actual input and see if they will produce the desired outputs. The answer is simple: it will not work because we need the effects of CS and US on pyramidal cells to overlap. We note that the fear learning community appears to agree with us that such contingency is necessary for synaptic plasticity (Sun et al., 2020; Palchaudhuri et al., 2024). To the best of our understanding, the source of that overlap is not understood in the community, and the gap has been much noticed (Sun et al., 2020). We do note, however, that STDP may not be the only kind of plasticity in fear learning (Li et al., 2009; Kim et al., 2013, 2016).

It is important to emphasize that it is not the aim of our paper to model the origin of the overlap. Rather, our intent is to demonstrate the roles of brain rhythms in producing the appropriate timing for STDP, assuming that ECS and F cells can continue to be active after the offset of CS and US, respectively. This assumption is very close to how the field now treats the plasticity, even for auditory fear conditioning (Sun et al., 2020). Thus, our methodology does not contradict known results. However, the question raised by R1 is indeed very interesting, if not the point of our paper. Hence, below we give details about why our hypothesis is reasonable.

Several papers (Quirk, Repa and LeDoux, 1995; Herry et al, 2007; Bordi and Ledoux 1992) show that the pips in auditory fear conditioning increase the activity of some BLA neurons: after an initial transient, the overall spike rate is still higher than baseline activity. As R1 points out, we did not model the transient increase in BLA spiking activity that occurs in response to each pip in the auditory fear conditioning paradigm. However, we did model the low-level sustained activity that occurs in between pips of the CS in the absence of US (Quirk, Repa and LeDoux, 1995, Fig. 2) and after CS offset (see Fig. 2B, left hand part of our manuscript). We read the data of Quirk et al., 1995 as suggesting that the low-level activity can be sustained for some indefinite time after a pip (cut off of recording was at 500 ms with no noticeable decrease in activity). As such, even if the pips and the US do not overlap in time, as in (Krabbe et al., 2019), the spiking of the ECS can be sustained after CS offset and thus overlap with US, a condition necessary in our model for plasticity through STDP. In Herry et al., 2007 Fig. 3 shows that BLA neurons respond to a pip at the population level with a transient increase in spiking and return to a baseline Poisson firing rate. However, a subset of cells continues to fire at an increased-over-baseline rate after the transient effect wears off (Fig. 3C, top few neurons) and this increased rate extends to the end of the recording time (here ~ 300 ms). These are the cells we consider to be ECS in our model. In Quirk et al., 1997, Fig. 5A also shows sustained low level activity of neurons in BLA in response to a pip. The low-level activity is shown to increase after fear learning, as is also the case in our model since ECS now entrains F so that there are more pyramidal cells spiking in response to CS. The question remains as to whether the spiking is sustained long enough and at a high enough rate for STDP to take place when US is presented sometime after the stop of the CS. 

Experimental recordings cannot speak to the rate of spiking of BLA neurons during US due to recording interference from the shock. However, evidence seems to suggest that ECS activity should increase during the US due to the release of acetylcholine (ACh) from neurons in the basal forebrain (BF) (Rajebhosale et al., 2024). Pyramidal cells of the BLA robustly express M1 muscarinic ACh receptors (Muller et al., 2013; McDonald and Mott, 2021). Thus, ACh from BF should elicit a depolarization in pyramidal cells. Indeed, the pairing of ACh with even low levels of spiking of BLA neurons results in a membrane depolarization that can last 7 – 10 s (Unal et al., 2015). This should induce higher spiking rates and more sustained activity in the ECS and F neurons during and after the presentation of US, thus ensuring a concomitant activation of ECS and fear (F) neurons necessary for STDP to take place. Other modulators, including dopamine, may also play a role in producing the sustained activity. Activation of US leads to increased dopamine release in the BLA (Harmer and Phillips, 1999; Suzuki et al., 2002). D1 receptors are known to increase the membrane excitability of BLA projection neurons by lowering their spiking threshold (Kröner et al., 2005). Thus, the activation of the US can lead to continued and higher firing rates of ECS and F. The effect of dopamine can last up to 20 minutes (Kröner et al., 2005). For CS-positive neurons, the ACh modulation coming from the firing of US may lead to a temporary extension of firing that is then amplified and continued by dopaminergic effects.

Hence, we suggest that a solution to the problem raised by R1 may be solved by considering the roles of ACh and dopamine in the BLA. The involvement of neuromodulators is consistent with the suggestion of (Sun et al., 2020). The model we have may be considered a “minimal” model that puts in by hand the overlap in activity due to the neuromodulation without explicitly modeling it. As R1 says, it is important for us to give the motivation of our hypotheses. We have used the simplest way to model overlap without assumptions about timing specificity in the overlap.

To account for these points in the manuscript, we first specified that we consider the effects of the US and CS inputs on the neuronal network as overlapping, while the actual inputs may not overlap. To do that, we added the following text:

(1) In the introduction: 

“In this paper, we aim to show 1) How a variety of BLA interneurons (PV, SOM and VIP) lead to the creation of these rhythms and 2) How the interaction of the interneurons and the rhythms leads to the appropriate timing of the cells responding to the US and those responding to the CS to promote fear association through spike-timing-dependent plasticity (STDP). Since STDP requires overlap of the effects of the CS and US, and some conditioning paradigms do not have overlapping US and CS, we include as a hypothesis that the effects of the CS and US overlap even if the CS and US stimuli do not. In the Discussion, we suggest how neuromodulation by ACh and/or dopamine can provide such overlap. We create a biophysically detailed model of the BLA circuit involving all three types of interneurons and show how each may participate in producing the experimentally observed rhythms and interacting to produce the necessary timing for the fear learning.”

(2) In the Result section “With the depression-dominated plasticity rule, all interneuron types are needed to provide potentiation during fear learning”:

“The 40-second interval we consider has both ECS and F, as well as VIP and PV interneurons, active during the entire period: an initial bout of US is known to produce a long-lasting fear response beyond the offset of the US (Hole and Lorens, 1975) and to induce the release of neuromodulators. The latter, in particular acetylcholine and dopamine that are known to be released upon US presentation (Harmer and Phillips, 1999; Suzuki et al., 2002; Rajebhosale et al., 2024), may induce more sustained activity in the ECS, F, VIP, and PV neurons during and after the presentation of US, thus ensuring a concomitant activation of those neurons necessary for STDP to take place (see “Assumptions and predictions of the model” in the Discussion).”

(3) In the Discussion section “Synaptic plasticity in our model”:

“Synaptic plasticity is the mechanism underlying the association between neurons that respond to the neutral stimulus CS (ECS) and those that respond to fear (F), which instantiates the acquisition and expression of fear behavior. One form of experimentally observed long-term synaptic plasticity is spike-timing-dependent plasticity (STDP), which defines the amount of potentiation and depression for each pair of pre- and postsynaptic neuron spikes as a function of their relative timing (Bi and Poo, 2001; Caporale and Dan, 2008). All forms of STDP require that there be an overlap in the firing of the pre- and postsynaptic cells. In some fear learning paradigms, the US and the CS do not overlap. We address this below under “Assumptions and predictions of the model”, showing how the effects of US and CS on the spiking of the relevant neurons can overlap even in the absence of overlap of US and CS.”

To fully present our reasoning about the origin of the overlap of the effects of US and CS, we modified and added to the last paragraph of the Discussion section “Assumptions and predictions of the model”, which now reads as follows:

“Finally, our model requires the effect of the CS and US inputs on the BLA neuron activity to overlap in time in order to instantiate fear learning through STDP. Such a hypothesis, that learning uses spike-timing-dependent plasticity, is common in the modeling literature (Bi and Poo, 2001; Caporale and Dan, 2008; Markram et al., 2011). Current paradigms of fear conditioning include examples in which the CS and US stimuli do not overlap (Krabbe et al., 2019). Such a condition might seem to rule out the mechanisms in our paper. Nevertheless, the argument below suggests that the effects of the CS and US can cause an overlap in neuronal spiking of ECS, F, VIP, and SOM, even when CS and US inputs do not overlap.

Experimental recordings cannot speak to the rate of spiking of BLA neurons during US due to recording interference from the shock. However, evidence suggests that ECS activity should increase during the US due to the release of acetylcholine (ACh) from neurons in the basal forebrain (BF) (Rajebhosale et al., 2024). Pyramidal cells of the BLA robustly express M1 muscarinic ACh receptors (McDonald and Mott, 2021). Thus, ACh from BF should elicit a depolarization in pyramidal cells. Indeed, the pairing of ACh with even low levels of spiking of BLA neurons results in a membrane depolarization that can last 7 – 10 s (Unal et al., 2015).   Other modulators, including dopamine, may also play a role in producing the sustained activity. Activation of US leads to increased dopamine release in the BLA (Harmer and Phillips, 1999; Suzuki et al., 2002). D1 receptors are known to increase the membrane excitability of BLA projection neurons by lowering their spiking threshold (Kröner et al., 2005). Thus, neuromodulator release should induce higher spiking rates and more sustained activity in the ECS and F neurons during and after the presentation of US, thus ensuring a concomitant activation of ECS and fear (F) neurons necessary for STDP to take place. Thus, the activation of the US can lead to continued and higher firing rates of ECS and F. The effect of dopamine can last up to 20 minutes (Kröner et al., 2005). For CS-positive neurons, the ACh modulation coming from the firing of US may lead to a temporary extension of firing that is then amplified and continued by dopaminergic effects.

Hence, we suggest that a solution to the problem apparently posed by the non-overlap US and CS in some paradigms of auditory fear conditioning (Krabbe et al., 2019) may be solved by considering the roles of ACh and dopamine in the BLA. The model we have may be considered a “minimal” model that puts in by hand the overlap in activity due to the neuromodulation without explicitly modeling it. We have used the simplest way to model overlap without assumptions about timing specificity in the overlap. We note that, even though ECS and F neurons have the ability to fire continuously when ACh and dopamine are involved, the participation of the interneurons enforces periodic silence needed for the depression-dominated STDP.”

In the Discussion (in section “Involvement of other brain structures”), we also acknowledged that the overlap between the effects of US and CS in the BLA may be provided by other brain structures by writing the following:

“In our model, the excitatory projection neurons and VIP and PV interneurons show sustained activity during and after the US presentation, thus allowing potentiation through STDP to take place. The medial prefrontal cortex and/or the hippocampus may provide the substrates for the continued firing of the BLA neurons after the 2-second US stimulation. We also discuss below that this network sustained activity may originate from neuromodulator release induced by US (see section “Assumptions and predictions of the model” in the Discussion).”

We also improved our discussion about the (Grewe et al., 2017) paper, which questions Hebbian plasticity in the context of fear conditioning based on several critiques. We included a new section in the Discussion entitled “Is STDP needed in fear conditioning?” to discuss those critiques and how our model may address them, which reads as follows:

“Is STDP needed in fear conditioning? The study in (Grewe et al., 2017) questions the validity of the Hebbian model in establishing associative learning during fear conditioning. There are several critiques we discuss here. The first critique is that Hebbian plasticity does not explain the experimental finding showing that both upregulation and downregulation of stimulus-evoked responses are present between coactive neurons. The upregulation is provided by our model, so the issue is the downregulation, which is not addressed by our model. However, our model highlights that coactivity alone does not create potentiation; the fine timing of the pre- and postsynaptic spikes determines whether there is potentiation or depression. Here, we find that PING networks are instrumental in setting up the fine timing for potentiation. We suggest that networks not connected to produce the PING may undergo depression when coactive.

The second critique raised by (Grewe et al., 2017) is that Hebbian plasticity alone does not explain why most of the cells exhibiting enhanced responses to the CS did not react to the US before fear conditioning. They suggest that neuromodulators may provide a third condition (besides the activity of the pre- and postsynaptic neurons) that changes the plasticity rule. Our model also does not explicitly address this experimental finding since it requires F to be initially activated by US in order for the fear association to be established. We agree that the fear cells described in (Grewe et al. 2017) may be depolarized by the US without reaching the spiking threshold; however, with neuromodulation provided during the fear training, the same input can lead to spiking, enabling the conditions for Hebbian plasticity. Our discussions above about how neuromodulators affect excitability are relevant to this point. We do not exclude that other forms of plasticity may play a role during fear conditioning in cells not initially activated by the US, but this is not the topic of our modeling study.

The third critique raised by (Grewe et al., 2017) is that Hebbian plasticity cannot explain why the majority of cells that were US- and CS-responsive before training have a reduced CS-evoked response afterward. The reduced response happens over multiple exposures of CS without US; this can involve processes similar to those present in fear extinction, which require plasticity in further networks, especially involving the infralimbic cortex (Milad and Quirk, 2002; Burgos-Robles et al., 2007). An extension of our model could investigate such mechanisms. In the fourth critique, (Grewe et al., 2017) suggests that the Hebbian plasticity rule cannot easily account for the reduction of the responses of many CS+-responsive cells, but not of the CS−-responsive cells. We suggest that the circuits involving paradigms similar to fear extinction do not involve the CS- cells.

Overall, we agree with (Grewe et al., 2017) that neuromodulators play a crucial role in fear conditioning, especially in prolonging the US- and CS-encoding activity as discussed in (see section “Assumptions and predictions of the model” in the Discussion), or even participating in changing the details of the plasticity rule. A possible follow-up of our work involves investigating how fear ensembles form and modify through fear conditioning and later stages. This follow-up work may involve using a tri-conditional rule, as suggested in (Grewe et al., 2017), in which the potential role of neuromodulators is taken into account in the plasticity rule in addition to the pre- and postsynaptic neuron activity. Another direction is to investigate a possible relationship between neuromodulation and a depression-dominated Hebbian rule.”

Finally, we made additional minor changes to the manuscript:

(1) In the Result section “Interneurons interact to modulate fear neuron output”, we specified the following:

“The US input on the pyramidal cell and VIP interneuron is modeled as a Poisson spike train at ~ 50 Hz and an applied current, respectively. In the rest of the paper, we will use the words “US” as shorthand for “the effects of US”.” 

(2) In the Result section “Interneuron rhythms provide the fine timing needed for depression dominated STDP to make the association between CS and fear”, we also reported the following:

“Similarly to the US, in the rest of the paper, we will use the words “CS” as shorthand for “the effects of CS”. In our simulations, CS is modeled as a Poisson spike train at ~ 50 Hz, independent of the US input. Thus, we hypothesize that the time structure of the inputs sometimes used for the training (e.g., a series of auditory pips) is not central to the formation of the plasticity in the network.”  

Reviewer #2 (Public Reviews):

The authors of this study have investigated how oscillations may promote fear learning using a network model. They distinguished three types of rhythmic activities and implemented an STDP rule to the network aiming to understand the mechanisms underlying fear learning in the BLA. 

After the revision, the fundamental question, namely, whether the BLA networks can or cannot intrinsically generate any theta rhythms, is still unanswered. The author added this sentence to the revised version: "A recent experimental paper, (Antonoudiou et al., 2022), suggests that the BLA can intrinsically generate theta oscillations (3-12 Hz) detectable by LFP recordings under certain conditions, such as reduced inhibitory tone." In the cited paper, the authors studied gamma oscillations, and when they applied 10 uM Gabazine to the BLA slices observed rhythmic oscillations at theta frequencies. 10 uM Gabazine does not reduce the GABA-A receptor-mediated inhibition but eliminates it, resulting in rhythmic populations burst driven solely by excitatory cells. Thus, the results by Antonoudiou et al., 2022 contrast with, and do not support, the present study, which claims that rhythmic oscillations in the BLA depend on the function of interneurons. Thus, there is still no convincing evidence that BLA circuits can intrinsically generate theta oscillations in intact brain or acute slices. If one extrapolates from the hippocampal studies, then this is not surprising, as the hippocampal theta depends on extrahippocampal inputs, including, but not limited to the entorhinal afferents and medial septal projections (see Buzsaki, 2002). Similarly, respiratory related 4 Hz oscillations are also driven by extrinsic inputs. Therefore, at present, it is unclear which kind of physiologically relevant theta rhythm in the BLA networks has been modelled. 

In our public reply to the Reviewer’s point, we reported the following:

(1) We kindly disagree that (Antonoudiou et al., 2022) contrasts with our study. (Antonoudiou et al., 2022) is a slice study showing that the BLA theta power (3-12 Hz) increases with gabazine compared to baseline. With all GABAergic currents omitted due to gabazine, the LFP is composed of excitatory currents and intrinsic currents. In our model, the high theta (6-12 Hz) comes from the spiking activity of the SOM cells, which increase their activity if the inhibition from VIP cells is removed. Thus, the model produces high theta in the presence of gabazine (see Fig. 1 in our replies to the Reviewers’ public comments). The model also shows that a PING rhythm is produced without gabazine, and that this rhythm goes away with gabazine because PING requires feedback inhibition from PV to fear cells. Thus, the high theta increase and gamma reduction with gabazine in the (Antonoudiou et al., 2022) paper can be reproduced in our model.

(2) We agree that (Antonoudiou et al., 2022) alone is not sufficient evidence that the BLA can produce low theta (3-6 Hz); we discussed a new paper (Bratsch-Prince et al., 2024) that provides further evidence of BLA ability to produce low theta and under what circumstances. The authors reported that intrinsic BLA theta is produced in slices with ACh stimulation (without needing external glutamate input) which, in vivo, would be provided by the basal forebrain (Rajebhosale et al., eLife, 2024) in response to salient stimuli. The low theta depends on muscarinic activation of CCK interneurons, a group of interneurons that overlaps with the VIP neurons in our model (Krabbe 2017; Mascagni and McDonald, 2003). We suspect that the low theta produced in (Bratsch-Prince et al., 2024) is the same as the low theta in our model. In future work, we will aim to show that ACh activates the BLA VIP cells, which are essential to the low theta generation in the network.

In the manuscript, we added to and modified the Discussion section “Where the rhythms originate, and by what mechanisms”. This text aims to better discuss (Antonoudiou et al. 2022) and introduce (Bratsch-Prince et al., 2024) with its connection to our hypothesis that the theta oscillations can be produced within the BLA. The new version is:

“Where the rhythms originate, and by what mechanisms. A recent experimental paper (Antonoudiou et al., 2022) suggests that the BLA can intrinsically generate theta oscillations (312 Hz) detectable by LFP recordings when inhibition is totally removed due to gabazine application. They draw this conclusion in mice by removing the hippocampus, which can volume conduct to BLA, and noticing that other nearby brain structures did not display any oscillatory activity. In our model, we note that when inhibition is removed, both AMPA and intrinsic currents contribute to the network dynamics and the LFP. Thus, interneurons with their specific intrinsic currents (i.e., D-current in the VIP interneurons, and NaP- and H- currents in SOM interneurons) can indeed affect the model LFP and support the generation of theta and gamma rhythms (Fig. 6G). 

Another slice study, (Bratsch-Prince et al., 2024), shows that BLA is intrinsically capable of producing a low theta rhythm with ACh stimulation and without needing external glutamate input. ACh is produced in vivo by the basal forebrain in response to US (Rajebhosale et al., 2024). Although we did not explicitly include the BF and ACh modulation of BLA in our model, we implicitly include the effect of ACh in BLA by increasing the activity of the VIP cells, which then produce the low theta rhythm. Indeed, low theta in the BLA is known to depend on the muscarinic activation of CCK interneurons, a group of interneurons that overlaps with the class of VIP neurons in our model (Mascagni and McDonald, 2003; Krabbe et al., 2018). 

Although the BLA can produce these rhythms, this does not rule out that other brain structures also produce the same rhythms through different mechanisms, and these can be transmitted to the BLA. Specifically, it is known that the olfactory bulb produces and transmits the respiratoryrelated low theta (4 Hz) oscillations to the dorsomedial prefrontal cortex, where it organizes neural activity (Bagur et al., 2021). Thus, the respiratory-related low theta may be captured by BLA LFP because of volume conduction or through BLA extensive communications with the prefrontal cortex. Furthermore, high theta oscillations are known to be produced by the hippocampus during various brain functions and behavioral states, including during spatial exploration (Vanderwolf, 1969) and memory formation/retrieval (Raghavachari et al., 2001), which are both involved in fear conditioning. Similarly to the low theta rhythm, the hippocampal high theta can manifest in the BLA. It remains to understand how these other rhythms may interact with the ones described in our paper. However, we emphasize that there is also evidence (as discussed above) that these rhythms arise within the BLA.”

Reviewer #2 (Recommendations for the Authors):

(1) Three different types of VIP interneurons with distinct firing patterns have been revealed in the BLA (Rhomberg et al., 2018). Does the generation of rhythmic activities depend on the firing features of VIP interneurons? Does it matter whether VIP interneurons fire burst of action potentials or they discharge more regularly?  

(2) The authors used data for modeling SST interneurons obtained e.g., in the hippocampus. However, there are studies in the BLA where the intrinsic characteristics of SST interneurons have been reported (Unal et al., 2020; Guthman et al., 2020; Vereczki et al., 2021). Have the authors considered using results of studies that were conducted in the BLA? 

We thank the Reviewer for their questions, which have helped us further improve our manuscript in response to similar queries from Reviewer 3 in the previous review round. More in detail:

(1) Although other electrophysiological types exist (Sosulina et al., 2010), we hypothesized that the electrophysiological type of VIP neurons that display intrinsic stuttering is the type that would be involved in mediating low theta oscillations during fear conditioning. This is because VIP intrinsic stuttering in cortical neurons is thought to involve the D-current, which helps create low theta bursting oscillations in the neuronal spiking patterns (Chartove et al., 2020). We think that the other subtypes of VIP interneurons are not essential for the low theta oscillatory dynamics observed during fear conditioning and, thus, did not provide an essential constraint for the phenomena we are trying to capture. VIP interneurons in our network must fire bursts at low theta to be effective in creating the pauses in ECS and F spiking needed for potentiation; single spikes at theta are not sufficient to create these pauses.

(2) In our model, we used the results conducted in a BLA study (Sosulina et al., 2010). SOM cells in the BLA display several physiologic types. We chose to include in our model the type showing early adaptation in response to a depolarizing current and inward (outward) rectification upon the initiation (release) of a hyperpolarizing current. We hypothesize that this type can produce high theta oscillations, a prominently observed rhythm in the BLA. Unal et al., 2020 (Unal et al., 2020) found two populations of SOM cells in the BLA, which have been previously recorded in (Sosulina et al., 2010), including the one type we chose to model. This SOM cell type shows a low threshold spiking profile characterized by spike frequency adaptation and voltage sag indicative of an H-current used in our model. Guthman et al., 2020, (Guthman et al., 2020), also found a population of SOM cells with hyperpolarization induced sag.

Our model also uses a NaP-current for which there is no data in the BLA. However, it is known to exist in hippocampal SOM cells and that NaP- and H- currents can produce such a high theta in hippocampal cells. It is a standard practice in modeling to use the best possible replacement for unknown currents. Of course, it is unfortunate to have to do this. We also note that models can be considered proof of principle, that can be proved or disproved by further experimental work. Both (Guthman et al., 2020) and (Vereczki et al., 2021) also uncover further heterogeneity among BLA SOM interneurons involving more than electrophysiology. We hypothesize that such a level of heterogeneity revealed by these three studies is not key to the question we are asking (where crucial ingredients are the rhythms) and, therefore, was not included in our minimal model.

We modified the Discussion section titled “Assumptions and predictions of the model” as follows:

“Our model, which is a first effort towards a biophysically detailed description of the BLA rhythms and their functions, does not include the neuron morphology, many other cell types, conductances, and connections that are known to exist in the BLA; models such as ours are often called “minimal models” and constitute most biologically detailed models. For example, although there is considerable variability in the activity patterns of both VIP cells and SOM cells (Sosulina et al., 2010; Guthman et al., 2020; Ünal et al., 2020; Vereczki et al., 2021), our focus was specifically on those subtypes that generate critical rhythms within the BLA. Such minimal models are used to maximize the insight that can be gained by omitting details whose influence on the answers to the questions addressed in the model are believed not to be qualitatively important. We note that the absence of these omitted features constitutes hypotheses of the model: we hypothesize that the absence of these features does not materially affect the conclusions of the model about the questions we are investigating. Of course, such hypotheses can be refuted by further work showing the importance of some omitted features for these questions and may be critical for other questions. Our results hold when there is some degree of heterogeneity of cells of the same type, showing that homogeneity is not a necessary condition.”

(3) The authors may double-check the reference list, as e.g., Cuhna-Reis et al., 2020 is not listed. 

We thank the Reviewer for spotting this. We checked the reference list and all the references are now listed.

Finally, we wanted to acknowledge that we made other changes to the manuscript unrelated to the reviewers’ questions with the purpose of gaining clarity. More specifically:

(1) We included a section titled “Significance” after the abstract and keywords, which reads as follows:

“Our paper accounts for the experimental evidence showing that amygdalar rhythms exist, suggests network origins for these rhythms, and points to their central role in the mechanisms of plasticity involved in associative learning. It is one of the few papers to address high-order cognition with biophysically detailed models, which are sometimes thought to be too detailed to be adequately constrained. Our paper provides a template for how to use information about brain rhythms to constrain biophysical models. It shows in detail, for the first time, how multiple interneurons help to provide time scales necessary for some kinds of spike-timing-dependent plasticity (STDP). It spells out the conditions under which such interactions between interneurons are needed for STDP and why. Finally, our work helps to provide a framework by which some of the discrepancies in the fear learning literature might be reevaluated. In particular, we discuss issues about Hebbian plasticity in fear learning; we show in the context of our model how neuromodulation might resolve some of those issues. The model addresses issues more general than that of fear learning since it is based on interactions of interneurons that are prominent in the cortex, as well as the amygdala.”

(2) The Result section “Physiology of the interneuron types is critical to their role in depression-dominated plasticity”, which is now titled “Mechanisms by which interneurons contribute to potentiation in depression-dominated plasticity”, now reads as follows:

“Mechanisms by which interneurons contribute to potentiation during depressiondominated plasticity. The PV cell is necessary to induce the correct pre-post timing between ECS and F needed for long-term potentiation of the ECS to F conductance. In our model, PV has reciprocal connections with F and provides lateral inhibition to ECS. Since the lateral inhibition is weaker than the feedback inhibition, PV tends to bias ECS to fire before F. This creates the fine timing needed for the depression-dominated rule to instantiate plasticity. If we used the classical Hebbian plasticity rule (Bi and Poo, 2001) with gamma frequency inputs, this fine timing would not be needed and ECS to F would potentiate over most of the gamma cycle, and thus we would expect random timing between ECS and F to lead to potentiation (Fig. S4). In this case, no interneurons are needed (See Discussion “Synaptic plasticity in our model” for the potential necessity of the depression-dominated rule). 

In this network configuration, the pre-post timing for ECS and F is repeated robustly over time due to coordinated gamma oscillations (PING, as shown in Fig. 4A, Fig. 1C) arising through the reciprocal interactions between F and PV (Feng et al., 2019). PING can arise only when PV is in a sufficiently low excitation regime such that F can control PV activity (Börgers et al., 2005), as in Fig. 4A. However, although such a low excitation regime establishes the correct fine timing for potentiation, it is not sufficient to lead to potentiation (Fig. 4A, Fig. S2C): the depression-dominated rule leads to depression rather than potentiation unless the PING is periodically interrupted. During the pauses, made possible only in the full network by the presence of VIP and SOM, the history-dependent build-up of depression decays back to baseline, allowing potentiation to occur on the next ECS/F active phase. (The detailed mechanism of how this happens is in the Supplementary Information, including Fig. S2). Thus, a network without the other interneuron types cannot lead to potentiation. Though a low excitation level for a PV cell is necessary to produce a PING, a higher excitation level is necessary to produce a pause in the ECS and F. This higher excitation level is consistent with the experimental literature showing a strong activation of PV after the onset of CS (Wolff et al., 2014). The higher excitation happens when the VIP cell is silent, whereas a low excitation level is achieved when the VIP cell fires and partially inhibits the PV cell (Fig. 4B, Fig. S2D). The interruption in the ECS and F activity requires the participation of another interneuron, the SOM cell (Figs. 2B, S2): the pauses in inhibition from the VIP periodically interrupt ECS and F firing by releasing PV and SOM from inhibition and thus indirectly silencing ECS and F. Without these pauses, depression dominates (see SI section “ECS and F activity patterns determine overall potentiation or depression”).”

We also removed a supplementary figure (Fig. S2).

(3) We wanted to be clear and motivate our choice to extend the low theta range to 2-6 Hz and the high theta range to 6-14 Hz, compared to the 3-6 Hz and 6-12 Hz, respectively in the BLA experimental literature. Our main reason for extending the ranges was because the peaks of low and high theta power in the VIP and SOM cells, respectively, (the cells that generate these oscillations) occurred at the borders of the experimental ranges. Thus, in order to include the peaks of the model LFP, we lowered the low theta range by 1 Hz and increased the high theta range by 2 Hz.

We present a new supplementary figure (Fig. S1) containing the power spectra of VIP, which is the source of low theta in our model, and SOM interneuron, which is the source of high theta:

We mention Fig. S1 in the Result section “Rhythms in the BLA can be produced by interneurons”, where we added the following text: o “In the baseline condition, the condition without any external input from the fear conditioning paradigm (Fig. 1B, top), our VIP neurons exhibit short bursts of gamma activity (~38 Hz) at low theta frequencies (~2-6 Hz) (peaking at ~3.5 Hz) (see Fig. S1A).” o “In our baseline model, SOM cells have a natural frequency of ~12 Hz (Fig. 1B, middle; Fig. S1B), which is at the upper limit of the experimental high theta range; this motivates our choice to extend the high theta range up to 14 Hz in order to include the peak.” 

Knowing the natural frequencies of VIP and SOM interneurons from the Result section “Rhythms in the BLA can be produced by interneurons”, we specified more clearly that we quantify the change of power in the low and high theta range around the power peaks in those ranges. Specifically, we changed some sentences in the first paragraph of the Result section “Increased low-theta frequency is a biomarker of fear learning” as follows:

“We find that fear conditioning leads to an increase in low theta frequency power of the network spiking activity compared to the pre-conditioned level (Fig. 6 A,B); there is no change in the high theta power. We also find that the LFP, modeled as the linear sum of all the AMPA, GABA, NaP-, D-, and H- currents in the network, similarly reveals a low theta power increase when considering the peak of the low theta power, and no significant variation in the high theta power again when considering the peak of the high theta power (Fig. 6 C,D,E).”

Finally, we made a few other small changes:

In the Introduction, we mention the following: “We also note that there is not uniformity on the exact frequencies associated with low and high theta, e.g., ((Lorétan et al., 2004) used 2-6 Hz for low theta). Here, we use 2-6 Hz for the theta range and 6-14 Hz for the high theta range.”

In Fig. 6DE (reported below point 3)), we reran the statistics using a smaller interval for high theta (11.5-13 Hz) to focus around the peak. Our initial result showing significant change in low theta between pre and post fear conditioning and no change in high theta still holds.

In Fig. 6 of the Result section “Increase low-theta frequency is a biomarker of fear learning”, we switched the order of panels F and G. This change allows us to first focus on the AMPA currents, which are the major contributors of the low theta power increase, and to specify what AMPA current drives that increase. After that, we present the power spectrum of the GABA currents, as well.

The corresponding text in the Result section, now reads as follows:

“We find that fear conditioning leads to an increase in low theta frequency power of the network spiking activity compared to the pre-conditioned level (Fig. 6 A,B); there is no change in the high theta power. We also find that the LFP, modeled as the linear sum of all the AMPA, GABA, NaP-, D-, and H- currents in the network, similarly reveals a low theta power increase when considering the peak of the low theta power, and no significant variation in the high theta power again when considering the peak of the high theta power (Fig. 6 C,D,E). These results are consistent with the experimental findings in (Davis et al., 2017). Specifically, the newly potentiated AMPA synapse from ECS to F ensures F is active after fear conditioning, thus generating strong currents in the PV cells to which it has strong connections (Fig. 6F). It is the AMPA currents to the PV interneurons that are directly responsible for the low theta increase; it is the newly potentiated ECS to F synapse that paces the AMPA currents in the PV interneurons to go at low theta. Thus, the low theta increase is due to added excitation provided by the new learned pathway.”

(4) In the Discussion section “Assumptions and predictions of the model”, we specified the following:

“Our model predicts that blockade of D-current in VIP interneurons (or silencing VIP interneurons) will both diminish low theta and prevent fear learning. Finally, the model assumes the absence of significantly strong connections from the excitatory projection cells ECS to PV interneurons, unlike the ones from F to PV. Including those synapses would alter the PING rhythm created by the interactions between F and PV, which is crucial for fine timing between ECS and F needed for LTP.”

(5) Finally, to broaden the potential interest of our study, we added the following sentences:

At the conclusion of the abstract:

“The model makes use of interneurons commonly found in the cortex and, hence, may apply to a wide variety of associative learning situations.” - At the conclusion of the introduction:

“Finally, we note that the ideas in the model may apply very generally to associative learning in the cortex, which contains similar subcircuits of pyramidal cells and interneurons: PV, SOM and VIP cells.” 

Also, changes in the emphasis of the paper led us to remove the following from the abstract: “Finally, we discuss how the peptide released by the VIP cell may alter the dynamics of plasticity to support the necessary fine timing.”

Author response:

The following is the authors’ response to the original reviews.

Recommendations for the authors:

Reviewer #1 (Recommendations For The Authors):

The manuscript could be improved by addressing the following issues.

(1) Fig. 3: The analgesic effects after astrocyte ablation appear to recover after one week. Is this due to repopulation of astrocytes?

Although we did not detect the proliferation of astrocytes, we hypothesized that it was likely related to the microglia phagocytosis of astrocyte debris after astrocyte ablation. Microglia are known to have the function of phagocytosis of cell debris. Diphtheria toxin-mediated cell ablation caused AAV2/5-GfaABC1D-Cre labeled astrocytes death and cell fragmentation. We hypothesized that the microglia could phagocyte the astrocyte fragments and were stimulated to activate type I interferon signal. When microglia phagocyte debris ended, the activation of type I interferon signal was also declined. Reduced activation of type I interferon signal may also be accompanied by recurrence of pain.

(2) Fig. 3: Please justify the large sample size of n=30-36. Is this sample size based on previous studies or statistical estimation?

The number of mice was based on our previous report [1], and the increased number of mice may also ensure that the pain data would also be reliable. Not only did we explore the differences between the sexes, and we also needed to obtain samples at different times for different experiments.

(3) Please try to plot individual data points for some critical time points to demonstrate data distribution. It is also helpful to plot male and female data points separately for some time points.

Individual data have been plotted as your request and added in the supplementary material.

(4) It is unclear if the same number of males and females were used in this study, as females were typically used for SCI studies. I wonder if you can use repeated measures Two-Way ANOVA for statistical analysis.

According to our observations, the number of males and females was not the same, while both of them were sufficient for statistical analysis. In addition, in the process of breeding transgenic mice, we would obtain both male and female mice, and rational use of mice may be better for us. Indeed, previous studies have shown that female mice are more commonly used in pain studies. Although we did not observe a gender difference in this study, it has been reported in the previous studies that gender is one of the factors for pain differences. According to your suggestion, we adopted the Two-Way ANOVA for statistical analysis and updated it in the part of statistical methods, but the statistical results were consistent with the previous results, so we did not modify the statistical results of the pictures.

(5) Fig. 3C, D: The effects of astrocyte ablation on mechanical pain are mild, compared to thermal pain. Electronic von Frey apparatus may be difficult for mice. It works very well for rats and large animals.

Since the animals involved in this study were all mice, we did not know how electronic von Frey was used in rats and large animals. But after the using of electronic von Frey, it seems to us that electronic von Frey is very suitable for mouse experiments. Best of all, our electronic von Frey can achieve accuracy as low as 0.01g. This allows us to detect very sensitive pain data, which may be more accurate and intuitive than before.

(6) Fig. 2B: In the figure legend it states n = 3 biological repeats. There are many more dots in each column. Are these individual animals or spinal cord sections?

As we describe in our method, n = 3 biological repeats represented three biological repeats per group, i.e., three mice/group with three IF per mouse. We take three or more values in each ascending tract (depending on the partition size of the different ascending tracts of lumbar enlargements). So, we would get more data as shown in Figure 2, which could be also more reliable.

(7) Fig. 4C: It appears that GFAP is increased by toxin treatment. Please explain this result.

This figure was calculated for astrocyte activation in the lesion area (T9-10), but not for the lumbar enlargement.

Reviewer #2 (Recommendations For The Authors):

Specific Comments:

RNA-Sequencing Analysis: The strength of the RNA-sequencing data in elucidating the impact of astrocyte elimination is compelling. While the focus on IFN signaling is well-supported, the manuscript overlooks other differentially expressed genes. A deeper analysis or at least a discussion of these genes could enrich the study's conclusions, offering a more holistic view of the underlying mechanisms.

Although we did not focus more on other relevant differential genes, we focused on the most significant differential genes, for these differential genes have a more significant effect on pain.

Q2: Figure Presentation: Consolidating Figures 1-3 could increase the clarity of the result presentation, reducing distractions from the main narrative. Certain aspects, such as the comparison of different tracts in Figure 2B and the body weight data in Figure 3C, seem tangential and might be better suited for supplementary materials.

The comparison of astrocyte activation in different ascending tracts of lumbar enlargements explained the relationships between astrocyte activation and pain, and laid the foundation for the subsequent astrocyte elimination. The weight data is also important, reflecting not only the changes in the overall recovery process after spinal cord injury, but also the effect of astrocyte elimination on the overall effect of mice. Thus, the weight data together with the pain test results will be more intuitive for the reader to understand the change of overall conditions of mice after astrocyte elimination.

Q3: Schematic Clarity: The schematic in Figure 1A is confusing, particularly in distinguishing between transgenic mice and viral constructs. The inconsistent naming of Cre recombinase (alternatively referred to as Cre, CRE, and sometimes DRE) further complicates understanding. Standardizing these elements would greatly enhance clarity for the readers.

As we described in the part of method, Gt(ROSA)26Sorem1(CAG-LSL-RSR-tdTomato-2A-DTR)Smoc mice contain both Loxp-stop-Loxp sequence and Rox-stop-Rox sequence. In the process of reproduction, Gt(ROSA)26Sorem1(CAG-LSL-RSR-tdTomato-2A-DTR)Smoc mice crossed with C57BL/6JSmoc-Tg(CAG-Dre)Smoc mice could remove the Rox-stop-Rox sequence, which could further crossed with mice containing Cre recombinase, or with AAV2/5-GfaABC1D-Cre intervention to remove the Loxp-stop-Loxp sequence and induce the expression of tdTomato and DTR.

Q4: Pathway Analysis: The discussion of the signal pathway analysis in Figure 8 leans heavily on speculation without direct evidence from the study. Distinguishing clearly between findings and literature-derived hypotheses is crucial. A more detailed discussion that properly cites sources for each pathway element would strengthen the manuscript.

According to your question, we have added this figure to the supplementary picture.

Q5: Statistical Analysis: The use of one-way ANOVA, despite presenting data in groups, is misaligned with the data's structure. Employing two-way ANOVA followed by post-hoc comparisons is appropriate for statistical analysis.

According to your suggestions, we adopted the Two-Way ANOVA for statistical analysis and updated it in the part of statistical methods, but the statistical results are consistent with the previous ones. Therefore, we did not modify the statistical results of the pictures.

Author response:

The following is the authors’ response to the original reviews.

Reviewer #1 (Public Review):

We appreciate the valuable and constructive comments of Reviewer #1 on our manuscript. We have addressed the comments from Reviewer #1 in the public review in the response to the recommendations for the authors, as the public review comments largely overlap with that of the recommendations for the authors.

Reviewer #1 (Recommendations For The Authors):

(1.1) Figure 1 did not use a mock-infected control for the development of R-loops but only a time before infection. I think it would have been a good control to have that after the same time of infection non-infected cells did not show increases in R-loops and this is not a product of the cell cycle.

We prepared our DRIPc-seq library using cell extracts harvested at 0, 3, 6, and 12 h post-infection (hpi), all at the same post-seeding time point. Each sample was infected with HIV-1 virus in a time-dependent manner. Therefore, it is unlikely that the host cellular R-loop induction observed in our DRIPc-seq results was due to R-loop formation during the cell cycle. In Lines 93–95 of the Results section of the revised manuscript, we have provided a more detailed description of our DRIPc-seq library experimental scheme. Thank you. 

(1.2) Figure 2 should have included a figure showing the proportion of DRIPc-seq peaks located in different genome features relative to one another instead of whether they were influenced by time post-infection. Figure 2C was performed in HeLa cells, but primary T cell data would have been more relevant as primary CD4+ T cells are more relevant to HIV infection.

We have included a new figure presenting the relative proportion of DRIPc-seq peaks mapped to different genomic features at each hpi (Fig. 2C of the revised manuscript). We found that the proportion of DRIPc-seq peaks mapped to various genomic compartments remained consistent over the hours following the HIV-1 infection. This further supports our original claim that HIV-1 infection does not induce R-loop enrichment at specific genomic features but that the accumulation of R-loops after HIV-1 infection is widely distributed.

We considered HeLa cells as the primary in vitro infection model, therefore, we conducted RNA-seq only on HeLa cells. However, we agree with the reviewer's opinion that data from primary CD4+ T cells may be more physiologically relevant. Nevertheless, as demonstrated in the new figure (Fig. 2C of the revised manuscript), HIV-1 infection did not significantly alter the proportion of R-loop peaks mapped to specific genomic compartments, such as gene body regions, in HeLa, primary CD4+ T, and Jurkat cells. Therefore, we anticipate no clear correlation between changes in gene expression levels and R-loop peak detection upon HIV-1 infection, even in primary T cells. Thank you.   

(1.3) Figure 5G is very hard to see when printed, is there a change in brightness or contrast that could be used? The arrows are helpful but they don't seem to be pointing to much.

We have highlighted the intensity of the PLA foci and magnified the images in Fig. 5G in the revised manuscript. While editing the images according to your suggestion, we found a misannotation regarding the multiplicity of infection in the number of PLA foci per nucleus quantification analysis graph in Fig. 5G of the original manuscript. We have corrected this issue and hope that it is now much clearer. 

(1.4) The introduction provided a good background for those who may not have a comprehensive understanding of DNA-RNA hybrids and R-loops, but the rationale that integration in non-expressed sequence implies that R-loops may be involved is very weak and was not addressed experimentally. A better rationale would have been to point out that, although integration in genes is strongly associated with gene expression, the association is not perfect, particularly in that some highly expressed genes are, nonetheless, poor integration targets.

In accordance with the reviewer's comment, we revised the Introduction. We have deleted the statement and reference in the introduction "... the most favored region of HIV-1 integration is an intergenic locus, ...”, which may overstate the relevance of the R-loop in HIV-1 integration events in non-expressed sequences. Instead, we introduced a more recent finding that high levels of gene expression do not always predict high levels of integration, together with the corresponding citation (Lines 46– 47 of the revised manuscript), according to the reviewer’s suggestion in the reviewer's public review 2)-(a).

(1.5) The discussion was seriously lacking in connecting their conclusions regarding R-loop targeting of integration to how integration works at the structural level, where it is very clear that concerted integration on the two DNA strands ca 5 bp apart is essential to correct, 2-ended integration. It is very difficult to visualize how this would be possible with the triple-stranded R-loop as a target. The manuscript would be greatly strengthened by an experiment showing concerted integration into a triplestranded structure in vitro using PICs or pure integrase.

We believe there has been a misunderstanding of our interpretation regarding the putative role of R-loop structures in the HIV-1 integration site mechanism because of some misleading statements in our original manuscript. Based primarily on our current data, we believe that R-loop structures are bound by HIV-1 integrase proteins and lead to HIV-1 viral genome integration into the vicinity regions of the host genomic R-loops. By carefully revising our manuscript, we found that the title, abstract, and discussion of our original manuscript includes phrases, such as “HIV-1 targets R-loops for integration,” which may overstate our finding on the role of R-loop in HIV-1 integration site selection. We replaced these phrases. For example, we used phrases, such as, “HIV-1 favors vicinity regions of R-loop for the viral genome integration,” in the revised manuscript. We apologize for the inconvenience caused by the unclear and nonspecific details of our findings.  

Using multiple biochemical experiments, we successfully demonstrated the interaction between the cellular R-loop and HIV-1 integrase proteins in cells and in vitro (Fig. 5 of the revised manuscript). However, we could not validate whether the center of the triple-stranded R-loops is the extraction site of HIV-1 integration, where the strand transfer reaction by integrase occurs. This is because an R-loop can be multi-kilobase in size (1, 2); therefore, we displayed a large-scale genomic region (30-kb windows) to present the integration sites surrounding the R-loop centers. Nevertheless, we believe that we validated R-loop-mediated HIV-1 integration in R-loop-forming regions using our pgR-poor and pgR-rich cell line models. When infected with HIV-1, pgR-rich cells, but not pgR-poor cells, showed higher infectivity upon R-loop induction in designated regions following DOX treatment (Fig. 3C and 3D of the revised manuscript). In addition, we quantified site-specific integration events in R-loop regions, and found that a greater number of integration events occurred in designated regions of the pgR-rich cellular genome upon R-loop induction by DOX treatment, but not in pgR-poor cells (Fig. 3E–G of the revised manuscript). 

We agree with the reviewer that an experiment showing the concerted integration of purified PICs into a triple-stranded structure in vitro would greatly strengthen our manuscript. We attempted the purification of viral DNA (vDNA)-bound PICs using either Sso7d-tagged HIV-1 integrase proteins or non-tagged HIV-1 integrase proteins (F185K/C280S) procured from the NIH HIV reagent program (HRP-20203), following the method described by Passos et al., Science, 2017; 355 (89-92) (3). Despite multiple attempts, we could not purify the nucleic acid-bound protein complexes for in vitro integration assays. However, we believe that pgR-poor and pgR-rich cell line models provide a strong advantage in specificity of our primer readouts. Compounded with our in cellulo observation, we believe that our work provides strong evidence for a causative relationship between R-loop formation/R-loop sites and HIV-1 integration.

Additionally, in the Discussion section of the revised manuscript, we have expanded our discussion on the role of genomic R-loops contributing in molding the host genomic environment for HIV-1 integration site selection, and the potential explanation on how R-loops are driving integration over long-range genomic regions. Thank you. 

(1.6) There are serious concerns with the quantitation of integration sites used here, which should be described in detail following line 503 but isn't. In Figure 3, E-G, they are apparently shown as reads per million, while in Figure 4B as "sites (%)" and in 4C as log10 integration frequency." Assuming the authors mean what they say, they are using the worst possible method for quantitation. Counting reads from restriction enzyme-digested, PCR-digested DNA can only mislead. At the numbers provided (MOI 0.6, 10 µg DNA assayed) there would be about 1 million proviruses in the samples assayed, so the probability of any specific site being used more than once is very low, and even less when one considers that a 10% assay efficiency is typical of integration site assays. Although the authors may obtain millions of reads per experiment, the number of reads per site is an irrelevant value, determined only by technical artefacts in the PCR reactions, most significantly the length of the amplicons, a function of the distance from the integration site to the nearest MstII site, further modified by differences in Tm. Better is to collapse identical reads to 1 per site, as may have been done in Figure 4B, however, the efficiency of integration site detection will still be inversely related to the length of the amplicon. Indeed, if the authors were to plot the read frequency against distance to the nearest MstII site, it is likely that they would get plots much like those in Figure 4B.

Detailed methods for integration site sequencing data processing are described in the Materials and Methods section of the revised manuscript (Line 621–631 of the revised manuscript). We primarily followed HIV-1 integration site sequencing data processing methods previously described by Li et al., mBio, 2020; 11(5) (4).  

While it may be correct that the HIV-1 integration event cannot occur more than once at a given site, our Fig. 3E, 4C, and 4D of the revised manuscript present the number of integration-site sequencing read counts expressed in reads-per-million (RPM) units or as log10-normalized values. Based on the number of mapped reads from the integration site sequencing results, we can infer that there was an integration event at this site, whether it was a single or multiple event.

We believe that the original annotation of y-axis, “Integration frequency,” may be misleading as it can be interpreted as a probability of any specific site being used for HIV-1 integration. Therefore, we corrected it as “number of mapped read” for clarity (Fig. 3E–G, 4C and 4D, and the corresponding figure legends of the revised manuscript). We apologize for any confusion. Thank you.

Other points:

(1.7) Overall: There are numerous grammatical and usage errors, especially in agreement of subject and verb, and missing articles, sometimes multiple times in the same sentence. These must be corrected prior to resubmission.

The revised manuscript was edited by a professional editing service. Thank you.

(1.8) Line 126-134: A striking result, but it needs more controls, as discussed above, including a dose-response analysis.

We determined the doses of NVP and RAL inhibitors in HeLa cells by optimizing the minimum dose of drug treatment that provided a sufficient inhibitory effect on HIV1 infection (Author response image 1). The primary objective of this experiment was to determine R-loop formation while reverse transcription or integration of the HIV-1 life cycle was blocked, therefore, we do not think that a dose-dependent analysis of inhibitors is required.

Author response image 1.

(A and B) Representative flow cytometry histograms of VSV-G-pseudotyped HIV-1-EGFP-infected HeLa cells at an MOI of 1, harvested at 48 hpi. The cells were treated with DMSO, the indicated doses of nevirapine (NVP) (A) or indicated doses of raltegravir (RAL) (B) for 24 h before infection. 

(1.9) Line 183: Please tell us what ECFP is and why it was chosen. Is there a reference for its failure to form R-loops?

Ibid: The human AIRN gene is a very poor target for HIV integration in PBMC.

A high GC skew value (> 0) is a predisposing factor for R-loop formation at the transcription site. This is because a high GC skew causes a newly synthesized RNA strand to hybridize to the template DNA strand, and the non-template DNA strand remains looped out in a single-stranded conformation (5) (Ref 36 in the revised manuscript). The ECFP sequence possessed a low GC skew value, as previously used for an R-loop-forming negative sequence (6) (Ref 17 of the revised manuscript). We have added this description and the corresponding references to Lines 188–192 of the revised manuscript.  

The human AIRN gene (RefSeq DNA sequence: NC_000006.12) sequence possesses a GC skew value of -0.04, in a window centered at base 2186, while the mouse AIRN (mAIRN) sequence is characterized by a GC skew value of 0.213. The ECFP sequence gave a GC skew value of -0.086 in our calculation. We anticipated that the human AIRN gene region does not form a stable R-loop, and in fact, it did not harbor R-loop enrichment upon HIV-1 infection in our DRIPc-seq data analysis of multiple cell types (Author response image 2)

Author response image 2.

Genome browser screenshot over the chromosomal regions in 20-kb windows centered on human AIRN showing results from DRIPc-seq in the indicated HIV-1-infected cells (blue, 0 hpi; yellow, 3 hpi; green, 6 hpi; red, 12 hpi)

(1.10) Line 190: You haven't shown dependence. Associated is a better word.

Thank you for the suggestion. We have changed “R-loop-dependent site-specific HIV-1 integration events...” to “R-loop-associated site-specific HIV-1 integration events...” (Line 198 of the revised manuscript) according to the reviewer’s suggestion in the revised manuscript. 

(1.11) Line 239: What happened to P1? What is the relationship of the P and N regions to genes?

We have added superimpositions of the P1 chromatin region on DRIPc-seq and the HIV-1 integration frequency to Figure 4C of the revised manuscript. We observed a relevant integration event within the P1 R-loop region, but to a lesser extent than in the P2 and P3 R-loop regions, perhaps because the P1 region has relatively less R-loop enrichment than the P2 and P3 regions, as examined by DRIP-qPCR in S3A Fig. of the revised manuscript.

Genome browser screenshots with annotations of accommodating genes in the P and N regions are shown in S2A–E Fig. of the revised manuscript, and RNA-seq analysis of the relative gene expression levels of the P1-3 and N1,2 R-loop regions are shown in S4 Table of the revised manuscript. Thank you.

(1.12) Line 261: But the binding affinity of integrase to the R-loop is somewhat weaker than to double-stranded DNA according to Figure 5A.

Nucleic acid substrates were loaded at the same molarity, and the percentage of the unbound fraction was calculated by dividing the intensity of the unbound fraction in each lane by the intensity of the unbound fraction in the lane with 0 nM integrase in the binding reaction. The calculated percentages of the unbound fraction from three independent replicate experiments are shown in Fig. 5A, right of the revised manuscript. In our analysis and measurements, the integrase proteins showed higher binding affinities to the R-loop and R-loop comprising nucleic acid structures than to dsDNA in vitro. We hope that this explanation clarifies this point. 

(1.13) Line 337: "accumulate". This is a not uncommon misinterpretation of the results of studies on the distribution of intact proviruses in elite controllers. The only possible correct interpretation of the finding is that proviruses form everywhere else but cells containing them are eliminated, most likely by the immune system.

Thank you for the suggestion. We have changed the Line 337 of the original manuscript to “... HIV-1 proviruses in heterochromatic regions are not eliminated but selected by immune system,” in Lines 361-363 of the revised manuscript. 

(1.14) Line 371 How many virus particles per cell does this inoculum amount to?

We determined the amount of GFP reporter viruses required to transduce ∼50% of WT Jurkat T cells, corresponding to an approximate MOI of 0.6. We repeatedly obtained 30–50% of VSV-G-pseudotyped HIV-1-EGFP positively infected cells for HIV1 integration site sequencing library construction for Jurkat T cells. 

(1.15) Line 503 and Figures 3 and 4: There must be a clear description of how integration events are quantitated.

Detailed methods for integration site sequencing data processing are described in the Materials and Methods section of the revised manuscript (Line 621–631 of the revised manuscript). We primarily followed HIV-1 integration site sequencing data processing methods previously described in Li et al., mBio, 2020; 11(5) (4).

Reviewer #2 (Public Review):

Retroviral integration in general, and HIV integration in particular, takes place in dsDNA, not in R-loops. Although HIV integration can occur in vitro on naked dsDNA, there is good evidence that, in an infected cell, integration occurs on DNA that is associated with nucleosomes. This review will be presented in two parts. First, a summary will be provided giving some of the reasons to be confident that integration occurs on dsDNA on nucleosomes. The second part will point out some of the obvious problems with the experimental data that are presented in the manuscript.

We appreciate your comments. We have carefully addressed the concerns expressed as follows (your comments are in italics):  

(2.1) 2017 Dos Passos Science paper describes the structure of the HIV intasome. The structure makes it clear that the target for integration is dsDNA, not an R-loop, and there are very good reasons to think that structure is physiologically relevant. For example, there is data from the Cherepanov, Engelman, and Lyumkis labs to show that the HIV intasome is quite similar in its overall structure and organization to the structures of the intasomes of other retroviruses. Importantly, these structures explain the way integration creates a small duplication of the host sequences at the integration site. How do the authors propose that an R-loop can replace the dsDNA that was seen in these intasome structures?

We do appreciate the current understanding of the HIV-1 integration site selection mechanism and the known structure of the dsDNA-bound intasome. Our study proposes an R-loop as another contributor to HIV-1 integration site selection. Recent studies providing new perspectives on HIV-1 integration site targeting motivated our current work. For instance, Ajoge et al., 2022 (7) indicated that a guanine-quadruplex (G4) structure formed in the non-template DNA strand of the R-loop influences HIV-1 integration site targeting. Additionally, I. K. Jozwik et al., 2022 (8) showed retroviral integrase protein structure bound to B-to-A transition in target DNA. R-loop structures are a prevalent class of alternative non-B DNA structures (9). We acknowledge the current understanding of HIV-1 integration site selection and explore how R-loop interactions may contribute to this knowledge in the Discussion section of our manuscript. 

Primarily based on our current data, we believe that R-loop structures are bound by HIV-1 integrase proteins and lead to HIV-1 viral genome integration into the vicinity regions of the host genomic R-loops, but we do not claim that R-loops completely replace dsDNA as the target for HIV-1 integration. An R-loop can be multi-kilobase in size and the R-loop peak length widely varies depending on the immunoprecipitation and library construction methods (1, 2), therefore, we could not validate whether the center of triple-stranded R-loops is the extraction site of HIV-1 integration where the strand transfer reaction by integrase occurs. Therefore, we replaced phrases such as, “HIV-1 targets R-loops for integration,” which may overstate our finding on the role of R-loop in HIV-1 integration site selection, with phrases, such as, “HIV-1 favors vicinity regions of R-loop for the viral genome integration,” in the revised manuscript. We apologize for the inconvenience caused by the unclear and non-specific details of our findings. Nevertheless, we believe that we validated R-loop-mediated HIV-1 integration in R-loop-forming regions using our pgR-poor and pgR-rich cell line models. We quantified site-specific integration events in the R-loop regions, and found that a greater number of integration events occurred in designated regions of the pgR-rich cellular genome upon R-loop induction by DOX treatment, but not in pgR-poor cells (Fig. 3E–G of the revised manuscript). 

dsDNA may have been the sole target of the intasome demonstrated in vitro possibly because dsDNA has only been considered as a substrate for in vitro intasome assembly. We hope that our work will initiate and advance future investigations on target-bound intasome structures by considering R-loops as potential new targets for integrated proteins and intasomes.  

(2.2) As noted above, concerted (two-ended) integration can occur in vitro on a naked dsDNA substrate. However, there is compelling evidence that, in cells, integration preferentially occurs on nucleosomes. Nucleosomes are not found in R loops. In an infected cell, the viral RNA genome of HIV is converted into DNA within the capsid/core which transits the nuclear pore before reverse transcription has been completed. Integration requires the uncoating of the capsid/core, which is linked to the completion of viral DNA synthesis in the nucleus. Two host factors are known to strongly influence integration site selection, CPSF6 and LEDGF. CPSF6 is involved in helping the capsid/core transit the nuclear pore and associate with nuclear speckles. LEDGF is involved in helping the preintegration complex (PIC) find an integration site after it has been released from the capsid/core, most commonly in the bodies of highly expressed genes. In the absence of an interaction of CPSF6 with the core, integration occurs primarily in the lamin-associated domains (LADs). Genes in LADs are usually not expressed or are expressed at low levels. Depending on the cell type, integration in the absence of CPSF6 can be less efficient than normal integration, but that could well be due to a lack of LEDGF (which is associated with expressed genes) in the LADs. In the absence of an interaction of IN with LEDGF (and in cells with low levels of HRP2) integration is less efficient and the obvious preference for integration in highly expressed genes is reduced. Importantly, LEDGF is known to bind histone marks, and will therefore be preferentially associated with nucleosomes, not R-loops. LEDGF fusions, in which the chromatin binding portion of the protein is replaced, can be used to redirect where HIV integrates, and that technique has been used to map the locations of proteins on chromatin. Importantly, LEDGF fusions in which the chromatin binding component of LEDGF is replaced with a module that recognizes specific histone marks direct integration to those marks, confirming integration occurs efficiently on nucleosomes in cells. It is worth noting that it is possible to redirect integration to portions of the host genome that are poorly expressed, which, when taken with the data on integration into LADs (integration in the absence of a CPSF6 interaction) shows that there are circumstances in which there is reasonably efficient integration of HIV DNA in portions of the genome in which there are few if any R-loops.

Although R-loops may not wrap around nucleosomes, long and stable R-loops likely cover stretches of DNA corresponding to multiple nucleosomes (10). For example, R-loops are associated with high levels of histone marks, such as H3K36me3, which LEDGF recognizes (2, 11). R-loops dynamically regulate the chromatin architecture. Possibly by altering nucleosome occupancy, positioning, or turnover, R-loop structures relieve superhelical stress and are often associated with open chromatin marks and active enhancers (2, 10). These features are also distributed over HIV-1 integration sites (12). In the Discussion section of the revised manuscript, we explored the R-loop molding mechanisms in the host genomic environment for HIV-1 integration site selection and its potential collaborative role with LEDGF/p75 and CPSF6 governing HIV-1 integration site selection. 

By carefully revising our original manuscript, with respect to the reviewer's comment, we recognized the need to tone down our statements. We found that the title, abstract, and discussion of our original manuscript includes phrases, such as, “HIV-1 targets Rloops for integration,” which may overstate our finding on the role of R-loop in HIV-1 integration site selection. We replaced these phrases. For example, we used phrases, such as “HIV-1 favors vicinity regions of R-loop for the viral genome integration,” in the revised manuscript. We apologize for the inconvenience caused by the unclear and non-specific details of our findings.

(2.3) Given that HIV DNA is known to preferentially integrate into expressed genes and that R-loops must necessarily involve expressed RNA, it is not surprising that there is a correlation between HIV integration and regions of the genome to which R loops have been mapped. However, it is important to remember that correlation does not necessarily imply causation.

We understand the reviewer's concern regarding the possibility of a coincidental correlation between the R-loop regions and HIV-1 integration sites, particularly when the interpretation of this correlation is primarily based on a global analysis. 

Therefore, we designed pgR-poor and pgR-rich cell lines, which we believe are suitable models for distinguishing between integration events driven by transcription and the presence of R-loops. Although the two cell lines showed comparable levels of transcription at the designated region upon DOX treatment via TRE promoter activation (Fig. 3B of the revised manuscript), only pgR-rich cells formed R-loops at the designated regions (Fig. 3C of the revised manuscript). When infected with HIV1, pgR-rich cells, but not pgR-poor cells, showed higher infectivity after DOX treatment (Fig. 3D of the revised manuscript). Moreover, we quantified site-specific integration events in the R-loop regions, and found that a greater number of integration events occurred in designated regions of the pgR-rich cellular genome upon R-loop induction by DOX treatment, but not in pgR-poor cells (Fig. 3E of the revised manuscript). Therefore, we concluded that transcriptional activation without an R-loop (in pgR-poor cells) may not be sufficient to drive HIV-1 integration. We believe that our work provides strong evidence for a causative relationship between R-loop formation/Rloop sites and HIV-1 integration. We hope that our explanation addresses your concerns. Thank you.

If we consider some of the problems in the experiments that are described in the manuscript:

(2.4) In an infected individual, cells are almost always infected by a single virion and the infecting virion is not accompanied by large numbers of damaged or defective virions. This is a key consideration: the claim that infection by HIV affects R-loop formation in cells was done with a VSVg vector in experiments in which there appears to have been about 6000 virions per cell. Although most of the virions prepared in vitro are defective in some way, that does not mean that a large fraction of the defective virions cannot fuse with cells. In normal in vivo infections, HIV has evolved in ways that avoid signaling infected the cell of its presence. To cite an example, carrying out reverse transcription in the capsid/core prevents the host cell from detecting (free) viral DNA in the cytoplasm. The fact that the large effect on R-loop formation which the authors report still occurs in infections done in the absence of reverse transcription strengthens the probability that the effects are due to the massive amounts of virions present, and perhaps to the presence of VSVg, which is quite toxic. To have physiological relevance, the infections would need to be carried out with virions that contain HIV even under circumstances in which there is at most one virion per cell.

Our virus production and in vitro and ex vivo HIV-1 infection experimental conditions, designed for infecting cell types, such as HeLa cells and primary CD4+ T cells with VSV-G pseudotyped HIV, were based on a comprehensive review of numerous references. At the very beginning of this study, we tested HIV-1-specific host genomic R-loop induction using empty virion particles (virus-like particles, VLP) or other types of viruses (non-retrovirus, SeV; retroviruses, FMLV and FIV), all produced with a VSV G protein donor. We could not include a control omitting the VSV G protein or using natural HIV-1 envelope protein to prevent viral spread in culture. We observed that despite all types of virus stocks being prepared using VSV-G, only cells infected with HIV-1 viruses showed R-loop signal enrichment (Author response image 3). Therefore, we omitted the control for the VSV G protein in subsequent analyses, such as DRIPcseq. We have also revised our manuscript to provide a clearer description of the experimental conditions. In particular, we now clearly stated that we used VSV-G pseudotyped HIV-1 in this study, throughout the abstract, results, and discussion sections of the revised manuscript. Thank you.

Author response image 3.

(A) Dot blot analysis of the R-loop in gDNA extracts from HIV-1 infected U2OS cells with MOI of 0.6 harvested at 6 hpi. The gDNA extracts were incubated with or without RNase H in vitro before membrane loading (anti-S9.6 signal). (B) Dot blot analysis of the R-loop in gDNA extracts from HeLa cells infected with 0.3 MOI of indicated viruses. The infected cells were harvested at 6 hpi. The gDNA extracts were incubated with or without RNase H in vitro before membrane loading (anti-S9.6 signal).

HIV-1 co-infection may also be expected in cell-free HIV-1 infections. However, it was previously suggested that the average number of infection events varies within 1.02 to 1.65 based on a mathematical model that estimates the frequency of multiple infections with the same virus (Figure 4c of Ito et al., Sci. Rep, 2017; 6559) (13). 

(2.5) Using the Sso7d version of HIV IN in the in vitro binding assays raises some questions, but that is not the real question/problem. The real problem is that the important question is not what/how HIV IN protein binds to, but where/how an intasome binds. An intasome is formed from a combination of IN bound to the ends of viral DNA. In the absence of viral DNA ends, IN does not have the same structure/organization as it has in an intasome. Moreover, HIV IN (even Sso7d, which was modified to improve its behavior) is notoriously sticky and hard to work with. If viral DNA had been included in the experiment, intasomes would need to be prepared and purified for a proper binding experiment. To make matters worse, there are multiple forms of multimeric HIV IN and it is not clear how many HIV INs are present in the PICs that actually carry out integration in an infected cell.

As the reviewer has noted, HIV IN, even with Sso7d tagging, is difficult. We attempted the purification of viral DNA (vDNA)-bound PICs using either Sso7d-tagged HIV-1 integrase proteins or non-tagged HIV-1 integrase proteins (F185K/C280S), procured from the NIH HIV reagent program (HRP-20203), following the method described by Passos et al., Science, 2017; 355 (89-92) (3). Despite multiple attempts, we were unable to purify the vDNA-bound IN protein complexes for in vitro assays. However, through multiple biochemical experiments, we believe that we have successfully demonstrated the interaction between cellular R-loops and HIV-1 integrase proteins both in cells and in vitro (Fig. 5A–F of the revised manuscript). We also observed a close association between integrase proteins and host cellular Rloops in HIV-1-infected cells, using a fluorescent recombinant virus (HIV-IN-EGFP) with intact IN-EGFP PICs (Fig. 5G of the revised manuscript). 

(2.6) As an extension of comment 2, the proper association of an HIV intasome/PIC with the host genome requires LEDGF and the appropriate nucleic acid targets need to be chromatinized.

The interaction between cellular R-loops and HIV-1 integrase proteins in HeLa cells endogenously expressing LEDGF/p75 was examined using reciprocal immunoprecipitation assays in Fig. 5C–F, S6B, and S6D Fig. of the revised manuscript. In addition, as discussed in more detail in our response to comment [28], we observed a close association between host cellular R-loops and HIV-1 integrase proteins by PLA assay, in HIV-1-infected HeLa cells. 

(2.7) Expressing any form of IN, by itself, in cells to look for what IN associates with is not a valid experiment. A major factor that helps to determine both where integration takes place and the sites chosen for integration is the transport of the viral DNA and IN into the nucleus in the capsid core. However, even if we ignore that important part of the problem, the IN that the authors expressed in HeLa cells won't be bound to the viral DNA ends (see comment 2), even if the fusion protein would be able to form an intasome. As such, the IN that is expressed free in cells will not form a proper intasome/PIC and cannot be expected to bind where/how an intasome/PIC would bind.

As discussed in more detail in our response to comment [2-8], we believe that our PLA experiment using the pVpr-IN-EGFP virus, which has previously been examined for virion integrity, as well as the IN-EGFP PICs (14), demonstrated a close association between host cellular R-loops and HIV-1 integrase proteins in HIV-1-infected cells. 

(2.8) As in comment 1, for the PLA experiments presented in Figure 5 to work, the number of virions used per cell (which differs from the MOI measured by the number of cells that express a viral marker) must have a high, which is likely to have affected the cells and the results of the experiment. However, there is the additional question of whether the IN-GFP fusion is functional. The fact that the functional intasome is a complex multimer suggests that this could be a problem. There is an additional problem, even if IN-GFP is fully functional. During a normal infection, the capsid core will have delivered copies of IN (and, in the experiments reported here, the IN-GFP fusion) into the nucleus that is not part of the intasome. These "free" copies of IN (here IN-GFP) are not likely to go to the same sites as an intasome, making this experiment problematic (comment 4).

The HIV-IN-EGFP virus stock was produced by polyethylenimine-mediated transfection of HEK293T cells with 6 µg of pVpr-IN-EGFP, 6 µg of HIV-1 NL4-3 noninfectious molecular clone (pD64E; NIH AIDS Reagent Program 10180), and 1 µg of pVSV-G as previously described in (14), and described in the Materials and Methods section of our manuscript. The pVpr-IN-EGFP vector used to produce HIV-1-IN-EGFP virus stock was provided by Anna Cereseto group (Albanese et al., PLOS ONE, 2008; 6(6); Ref 34 of the revised manuscript). It was previously reported that the HIV-1INEGFP virions produced by IN-EGFP trans-incorporation through Vpr are intact and infective viral particles (Figure 1 of Albanese et al., PLOS ONE, 2008; 6(6)). Therefore, we believe that the HIV-IN-EGFP used in our PLA experiments was functional. 

Additionally, Albanese et al. showed that the EGFP signal of HIV-IN-EGFP virions colocalizes with the viral protein matrix (p17MA) and capsid (P24CA) as well as with the newly synthesized cDNA produced by reverse transcriptase by labeling and visualizing the synthesized cDNA (14). In addition, the fluorescent recombinant virus (HIV-INEGFP) was structurally intact at the nuclear level (Figure 6 of Albanese et al., PLOS ONE, 2008; 6(6)). Therefore, we believe that our PLA experimental result is not likely misled as the reviewer concerns due to the integrity of the HIV-IN-EGFP virion as well as IN-EGFP PICs.

Furthermore, the in vitro HIV-1 infection setting of our PLA experiments was carefully determined based on multiple studies that performed image-based assays on HIV-1infected cells. For instance, Albanese et al. infected 4 × 104 cells with viral loads equivalent to 1.5 or 3 µg of HIV-1 p24 for their immunofluorescence analysis, in their previous report (14). We titrated the fluorescent HIV-1 virus stocks by examining both the multiplicity of infection (MOI) and quantifying the HIV-1 p24 antigen content (Author response image 4). In our calculation, we infected 5 × 104 HeLa cells with viral loads equivalent to 1.3 ug of HIV-1 p24, which is indicated as 2 MOI in Fig. 5G of our manuscript, for our PLA experiments. 

Image-Based Assays often require increased and enhanced signal for statistical robustness. For example, Achuthan et al. infected cells with VSV-G-pseudotyped HIV1 at the approximate MOI of 350 for vDNA and PIC visualization (15). Therefore, we believe our experimental condition for PLA experiments, which we carefully designed based on previous study that are frequently referred, are reasonable. We really hope that our discussion sufficiently addressed the reviewer’s concern. 

Author response image 4.

Gating strategy used to determine HIV-1-infectivity in HeLa cells at 48 hpi. Cells were infected with a known p24 antigen content in the stock of the VSV-G-pseudotyped HIV-1-EGFP-virus. The percentages of GFP-positive cell population are indicated.

(2.9) In the Introduction, the authors state that the site of integration affects the probability that the resulting provirus will be expressed. Although this idea is widely believed in the field, the actual data supporting it are, at best, weak. See, for example, the data from the Bushman lab showing that the distribution of integration sites is the same in cells in which the integrated proviruses are, and are not, expressed. However, given what the authors claim in the introduction, they should be more careful in interpreting enzyme expression levels (luciferase) as a measure of integration efficiency in experiments in which they claim proviruses are integrated in different places.

We thank the reviewer for the constructive comment. We have changed the statement in Lines 41–42 in the Introduction section of our original manuscript to “The chromosomal landscape of HIV-1 integration influences proviral gene expression, persistence of integrated proviruses, and prognosis of antiretroviral therapy.” (Lines 39-41 of the revised manuscript). We believe that this change can tone-down the relevance between the site of integration and the provirus expression level.

The piggyBac transposase randomly insert the “cargo (transposon)” into TTAA chromosomal sites of the target genome, generating efficient insertions at different genomic loci (16, 17). We believe that this random insertion of the pgR-poor/rich vector mediated by the piggyBac system allows us not to mislead the R-loop-mediated HIV1 integration site because of the genome locus bias of the vector insertion. Therefore, Figure 3 in our manuscript does not claim any relevance between the site of integration and the resulting provirus expression levels. Instead, as noted in Line 214 of the revised manuscript, using the luciferase reporter HIV-1 virus, we attempted to examine HIV-1 infection in cells with an "extra number of R-loops” in the host cellular genome. We observed that pgR-rich cells showed higher luciferase activity upon DOX treatment than pgR-poor cells (Fig. 3D of the revised manuscript). We believe that this is because a greater number of HIV-1 integration events may occur in pgR-rich cells, where DOX-inducible de novo R-loop regions are introduced. This has been further examined in Fig. 3E–G of the revised manuscript. We hope this explanation clarifies the Figure 3. Thank you. 

(2.10) Using restriction enzymes to create an integration site library introduces biases that derive from the uneven distribution of the recognition sites for the restriction enzymes.

As described in the Materials and Methods section, we adopted a sequencing library construction method using a previously established protocol (18, 19). Although we recognize the advantages of DNA fragmentation by sonication, in in vitro or ex vivo HIV-1 infection settings, where the multiplicity of infection is carefully determined based on multiple references, more copies of integrated viral sequences are expected compared to that in samples from infected patients (18). Therefore, in these settings, restriction enzyme-based DNA fragmentation and ligation-mediated PCR sequencing are well-established methods that provide significant data sources for HIV-1 integration site sequencing (15, 20-22). Furthermore, our data showing the proportion of integration sites over R-loop regions (Fig. 4B of the revised manuscript) are presented alongside the respective random controls (i.e., proportion of integration sites within the 30-kb windows centered on randomized DRIPc-seq peaks, gray dotted lines; control comparisons between randomized integration sites with DRIPc-seq peaks, black dotted lines; and randomized integration sites with randomized DRIPcseq peaks, gray solid lines), which do not show such a correlation between the HIV-1 integration sites and nearby areas of the R-loop regions. Therefore, we believe that our results from the integration site sequencing data analysis are unlikely to be biased. 

Reviewer #3 (Public Review):

In this manuscript, Park and colleagues describe a series of experiments that investigate the role of R-loops in HIV-1 genome integration. The authors show that during HIV-1 infection, R-loops levels on the host genome accumulate. Using a synthetic R-loop prone gene construct, they show that HIV-1 integration sites target sites with high R-loop levels. They further show that integration sites on the endogenous host genome are correlated with sites prone to R-loops. Using biochemical approaches, as well as in vivo co-IP and proximity ligation experiments, the authors show that HIV-1 integrase physically interacts with R-loop structures.

My primary concern with the paper is with the interpretations the authors make about their genome-wide analyses. I think that including some additional analyses of the genome-wide data, as well as some textual changes can help make these interpretations more congruent with what the data demonstrate. Here are a few specific comments and questions:

We are grateful for the time and effort we spent on our behalf and the reviewer’s appreciation for the novelty of our work, in particular, R-loop induction by HIV-1 infection and the correlation between host R-loops and the genomic site of HIV-1 integration. In the following sections, we provide our responses to your comments and suggestions. Your comments are in italics. We have carefully addressed the following issues.

(3.1) I think Figure 1 makes a good case for the conclusion that R-loops are more easily detected HIV-1 infected cells by multiple approaches (all using the S9.6 antibody). The authors show that their signals are RNase H sensitive, which is a critical control. For the DRIPc-Seq, I think including an analysis of biological replicates would greatly strengthen the manuscript. The authors state in the methods that the DRIPc pulldown experiments were done in biological replicates for each condition. Are the increases in DRIPc peaks similar across biological replicates? Are genomic locations of HIV-1-dependent peaks similar across biological replicates? Measuring and reporting the biological variation between replicate experiments is crucial for making conclusions about increases in R-loop peak frequency. This is partially alleviated by the locus-specific data in Figure S3A. However, a better understanding of how the genome-wide data varies across biological replicates will greatly enhance the quality of Figure 1.

DRIPc-seq experiments were conducted with two biological replicates. To define consensus DRIPc-seq peaks using these two replicates, we used two methods applicable to ChIP-seq analysis: the irreproducible discovery rate (IDR) method and sequencing data pooling. We found that the sequencing data pooling method yielded significantly more DRIPc-seq peaks than consensus peak identification through IDR, and we decided to utilize R-loop peaks from pooled sequencing data for our downstream analyses, as described in the figure legends and Materials and Methods of the revised manuscript. 

As noted by the reviewer, it is important to verify whether the increasing trend in the number of R-loop peaks and genomic locations of HIV-1 dependent R-loops were consistently observed across the two biological replicates. Therefore, we independently performed R-loop calling on each replicate of the sequencing data of primary CD4+ T cells from two individual donors to verify that the increase in R-loop numbers was consistent (Author response image 5). Additionally, the overlap of the R-loop peaks between the two replicates was statistically significant across the genome (Author response table 1). Thank you.

Author response image 5.

Bar graph indicating DRIPc-seq peak counts for HIV-1-infected primary CD4+ T cells harvested at the indicated hours post infection (hpi). Pre-immunoprecipitated samples were untreated (−) or treated (+) with RNase H, as indicated. Each dot corresponds to an individual data set from two biologically independent experiments.

Author response table 1.

DRIPc-seq peak length and Chi-square p-value in CD4+ T cells from individual donor 1 and 2 

(3.2) I think that the conclusion that R-loops "accumulate" in infected cells is acceptable, given the data presented. However, in line 134 the authors state that "HIV1 infection induced host genomic R-loop formation". I suggest being very specific about the observation. Accumulation can happen by (a) inducing a higher frequency of the occurrence of individual R-loops and/or (b) stabilizing existing R-loops. I'm not convinced the authors present enough evidence to claim one over the other. It is altogether possible that HIV-1 infection stabilizes R-loops such that they are more persistent (perhaps by interactions with integrase?), and therefore more easily detected. I think rephrasing the conclusions to include this possibility would alleviate my concerns.

We thank the reviewer for the considerable discussion on our manuscript. We have now changed Line 134 to, “HIV-1 infection induces host genomic R-loop enrichment” (Lines 132-133 of the revised manuscript), and added a new conclusion sentence implicating the possible explanation for the R-loop signal enrichment upon HIV-1 infection (Lines 133–135 of the revised manuscript), according to the reviewer's suggestion.    

(3.3) A technical problem with using the S9.6 antibody for the detection of R-loops via microscopy is that it cross-reacts with double-stranded RNA. This has been addressed by the work of Chedin and colleagues (as well as others). It is absolutely essential to treat these samples with an RNA:RNA hybrid-specific RNase, which the authors did not include, as far as their methods section states. Therefore, it is difficult to interpret all of the immunofluorescence experiments that depend on S9.6 binding.

We understand the reviewer's concern regarding the cross-reactivity of the S9.6 antibody with more abundant dsRNA, particularly in imaging applications. We carefully designed the experimental and analytical methods for R-loop detection using microscopy. For example, we pre-extracted the cytoplasmic fraction before staining with the S9.6 antibody and quantified the R-loop signal by subtracting the nucleolar signal. Both of these steps were taken to eliminate the possibility of misdetecting Rloops via microscopy because of the prominent cytoplasmic and nucleolar S9.6 signals, which primarily originate from ribosomal RNA. In addition, we included R-loop negative control samples in our microscopy analysis that were subjected to intensive RNase H treatment (60U/mL RNase H for 36 h) and observed a significant reduction in the S9.6 signal (Figure 1E of the revised manuscript). RNase H-treated samples served as essential and widely accepted negative controls for R-loop detection. 

We would like to point out that recent studies have reported strong intrinsic specificity of S9.6 anybody for DNA:RNA hybrid duplex over dsDNA and dsRNA, along with the structural elucidations of S9.6 antibody recognition of hybrids (23, 24). Therefore, our interpretation of host cellular R-loop enrichment after HIV-1 infection using S9.6 antibodies in multiple biochemical approaches is well supported. Nevertheless, we agree with the reviewer's opinion that additional negative controls for the detection of R-loops via microscopy, such as RNase T1-and RNase III-treated samples, could improve the robustness and accuracy of R-loop imaging data (25).  

(3.4) Given that there is no clear correlation between expression levels and R-loop peak detection, combined with the data that show increased detection of R-loop frequency in non-genic regions, I think it will be important to show that the R-loop forming regions are indeed transcribed above background levels. This will help alleviate possible concerns that there are technical errors in R-loop peak detection.

Figures S5D and S5E in the revised manuscript show the relative gene expression levels of the R-loop-forming positive regions (P1-3) and the referenced Rloop-positive loci (RPL13A and CALM3). The gene expression levels of these R-loopforming regions were significantly higher than those of the ECFP or mAIRN genes without DOX treatment, which can be considered background levels of transcription in cells. Thank you. 

(3.5) In Figures 4C and D the hashed lines are not defined. It is also interesting that the integration sites do not line up with R-loop peaks. This does not necessarily directly refute the conclusions (especially given the scale of the genomic region displayed), but should be addressed in the manuscript. Additionally, it would greatly improve Figure 4 to have some idea about the biological variation across replicates of the data presented 4A.

We thank the reviewer for the considerable comment on our study. First of all, we added an annotation for the dashed lines in the figure legends of Figures 4C and 4D in the revised manuscript.

We agree with the reviewer's interpretation of the relationship between the integration sites and R-loop peaks. Primarily based on our current data, we believe R-loop structures are bound by HIV-1 integrase proteins and lead HIV-1 viral genome integration into the “vicinity” regions of the host genomic R-loops. We displayed a large-scale genomic region (30-kb windows) to present integration sites surrounding R-loop centers because an R-loop can be multi-kilobase in size (1, 2). Depending on the immunoprecipitation and library construction methods, the R-loop peaks varied in size, and the peak length showed a wide distribution (Figure 3B of Malig et al., 2020, Figure 1B of Sanz et al., 2016, and Figure 2A of the revised manuscript). Therefore, presenting integration site events within a wide window of R-loop peaks could be more informative and better reflect the current understanding of R-loop biology.

R-loop formation recruits diverse chromatin-binding protein factors, such as H3K4me1, p300, CTCF, RAD21, and ZNF143 (Figure 6A and 6B of Sanz et al., 2016) (26), which allow R-loops to exhibit enhancer and insulator chromatin states, which can act as distal regulatory elements (26, 27). We have demonstrated physical interactions between host cellular R-loops and HIV-1 integrase proteins (Figure 5 of the revised manuscript), therefore, we believe that this ‘distal regulatory element-like feature’ of the R-loop can be a potential explanation for how R-loops drive integration over longrange genomic regions.

According to your suggestion, we added this explanation to the relevant literature in the Discussion section of the revised manuscript.

Author response image 6 which represents the biological variation across replicates of the data shown in Figure 4A. The integration site sequencing data for Jurkat cells were adopted from SRR12322252 (4), which consists of the integration site sequencing data of HIV-1-infected wild type Jurkat cells with one biological replicate. We hope that our explanations and discussion have successfully addressed your concerns. Thank you. 

Author response image 6.

Bar graphs showing the quantified number of HIV-1 integration sites per Mb pair in total regions of 30-kb windows centered on DRIPc-seq peaks from HIV-1 infected HeLa cells and primary CD4+ T cells (magenta) or non-R-loop region in the cellular genome (gray). Each dot corresponds to an individual data set from two biologically independent experiments.

(3.6) The authors do not adequately describe the Integrase mutant that they use in their biochemical experiments in Figure 5A. Could this impact the activity of the protein in such a way that interferes with the interpretation of the experiment? The mutant is not used in subsequent experiments for Figure 5 and so even though the data are consistent with each other (and the conclusion that Integrase interacts with R-loops) a more thorough explanation of why that mutant was used and how it impacts the biochemical activity of the protein will help the interpretation of the data presented in Figure 5.

We appreciate the reviewer’s suggestions. In our EMSA analysis, we purified and used Sso7d-tagged HIV-1 integrase proteins with an active-site amino acid substitution, E152Q. First, we used the Sso7d-tagged HIV-1 integrase protein, as it has been suggested in previous studies that the fusion of small domains, such as Sso7d (DNA binding domain) can significantly improve the solubility of HIV integrase proteins without affecting their ability to assemble with substrate nucleic acids and their enzymatic activity (Figure 1B of Li et al., PLOS ONE, 2014;9 (8) (28, 29). We used an integrase protein with an active site amino acid substitution, E152Q, in our mobility shift assay, because the primary goal of this experiment was to examine the ability of the protein to bind or form a complex with different nucleic acid substrates. We thought that abolishing the enzymatic activity of the integrase protein, such as 3'-processing that cleaves DNA substrates, would be more appropriate for our experimental objective. This Sso7d tagged- HIV-1 integrase with the E152Q mutation has also been used to elucidate the structural model of the integrase complex with a nucleic acid substrate by cryo-EM (3) and has been shown to not disturb substrate binding.   Based on the reviewer’s comments, we have added a description of the E152Q mutant integrase protein in Lines 268–270 of the revised manuscript. Thank you.

Reviewer #3 (Recommendations For The Authors):

The paper suffers from many grammatical errors, which sometimes interfere with the interpretations of the experiments. In the view of this reviewer, the manuscript must be carefully revised prior to publication. For example, lines 247-248 "Intasomes consist of HIV-1 viral cDNA and HIV-1 coding protein, integrases." It is unclear from this sentence whether there are multiple integrases or multiple proteins that interact with the viral genome to facilitate integration. This makes the subsequent experiments in Figure 5 difficult to interpret. There are many other examples, too numerous to point out individually.

We thoughtfully revised the original manuscript, making the best efforts to provide clearer details of our findings. We believe that we have made substantial changes to the manuscript, including Lines 247–248 of the original manuscript that the reviewer noted. Furthermore, the revised manuscript was edited by a professional editing service. Thank you.     (1) M. Malig, S. R. Hartono, J. M. Giafaglione, L. A. Sanz, F. Chedin, Ultra-deep Coverage Singlemolecule R-loop Footprinting Reveals Principles of R-loop Formation. J Mol Biol 432, 22712288 (2020).

(2) L. A. Sanz et al., Prevalent, Dynamic, and Conserved R-Loop Structures Associate with Specific Epigenomic Signatures in Mammals. Mol Cell 63, 167-178 (2016).

(3) D. O. Passos et al., Cryo-EM structures and atomic model of the HIV-1 strand transfer complex intasome. Science 355, 89-92 (2017).

(4) W. Li et al., CPSF6-Dependent Targeting of Speckle-Associated Domains Distinguishes Primate from Nonprimate Lentiviral Integration. mBio 11,  (2020).

(5) P. A. Ginno, Y. W. Lim, P. L. Lott, I. Korf, F. Chedin, GC skew at the 5' and 3' ends of human genes links R-loop formation to epigenetic regulation and transcription termination. Genome Res 23, 1590-1600 (2013).

(6) S. Hamperl, M. J. Bocek, J. C. Saldivar, T. Swigut, K. A. Cimprich, Transcription-Replication Conflict Orientation Modulates R-Loop Levels and Activates Distinct DNA Damage Responses. Cell 170, 774-786 e719 (2017).

(7) H. O. Ajoge et al., G-Quadruplex DNA and Other Non-Canonical B-Form DNA Motifs Influence Productive and Latent HIV-1 Integration and Reactivation Potential. Viruses 14,  (2022).

(8) I. K. Jozwik et al., B-to-A transition in target DNA during retroviral integration. Nucleic Acids Res 50, 8898-8918 (2022).

(9) F. Chedin, C. J. Benham, Emerging roles for R-loop structures in the management of topological stress. J Biol Chem 295, 4684-4695 (2020).

(10) F. Chedin, Nascent Connections: R-Loops and Chromatin Patterning. Trends Genet 32, 828838 (2016).

(11) P. B. Chen, H. V. Chen, D. Acharya, O. J. Rando, T. G. Fazzio, R loops regulate promoterproximal chromatin architecture and cellular differentiation. Nat Struct Mol Biol 22, 9991007 (2015).

(12) A. R. Schroder et al., HIV-1 integration in the human genome favors active genes and local hotspots. Cell 110, 521-529 (2002).

(13) Y. Ito et al., Number of infection events per cell during HIV-1 cell-free infection. Sci Rep 7, 6559 (2017).

(14) A. Albanese, D. Arosio, M. Terreni, A. Cereseto, HIV-1 pre-integration complexes selectively target decondensed chromatin in the nuclear periphery. PLoS One 3, e2413 (2008).

(15) V. Achuthan et al., Capsid-CPSF6 Interaction Licenses Nuclear HIV-1 Trafficking to Sites of Viral DNA Integration. Cell Host Microbe 24, 392-404 e398 (2018).

(16) X. Li et al., piggyBac transposase tools for genome engineering. Proc Natl Acad Sci U S A 110, E2279-2287 (2013).

(17) Y. Cao et al., Identification of piggyBac-mediated insertions in Plasmodium berghei by next generation sequencing. Malar J 12, 287 (2013).

(18) E. Serrao, P. Cherepanov, A. N. Engelman, Amplification, Next-generation Sequencing, and Genomic DNA Mapping of Retroviral Integration Sites. J Vis Exp,  (2016).

(19) K. A. Matreyek et al., Host and viral determinants for MxB restriction of HIV-1 infection. Retrovirology 11, 90 (2014).

(20) G. A. Sowd et al., A critical role for alternative polyadenylation factor CPSF6 in targeting HIV-1 integration to transcriptionally active chromatin. Proc Natl Acad Sci U S A 113, E10541063 (2016).

(21) B. Lucic et al., Spatially clustered loci with multiple enhancers are frequent targets of HIV-1 integration. Nat Commun 10, 4059 (2019).

(22) P. K. Singh, G. J. Bedwell, A. N. Engelman, Spatial and Genomic Correlates of HIV-1 Integration Site Targeting. Cells 11,  (2022).

(23) C. Bou-Nader, A. Bothra, D. N. Garboczi, S. H. Leppla, J. Zhang, Structural basis of R-loop recognition by the S9.6 monoclonal antibody. Nat Commun 13, 1641 (2022).

(24) Q. Li et al., Cryo-EM structure of R-loop monoclonal antibody S9.6 in recognizing RNA:DNA hybrids. J Genet Genomics 49, 677-680 (2022).

(25) J. A. Smolka, L. A. Sanz, S. R. Hartono, F. Chedin, Recognition of RNA by the S9.6 antibody creates pervasive artifacts when imaging RNA:DNA hybrids. J Cell Biol 220,  (2021).

(26) L. A. Sanz, F. Chedin, High-resolution, strand-specific R-loop mapping via S9.6-based DNARNA immunoprecipitation and high-throughput sequencing. Nat Protoc 14, 1734-1755 (2019).

(27) M. Merkenschlager, D. T. Odom, CTCF and cohesin: linking gene regulatory elements with their targets. Cell 152, 1285-1297 (2013).

(28) M. Li, K. A. Jurado, S. Lin, A. Engelman, R. Craigie, Engineered hyperactive integrase for concerted HIV-1 DNA integration. PLoS One 9, e105078 (2014).

(29) M. Li et al., A Peptide Derived from Lens Epithelium-Derived Growth Factor Stimulates HIV1 DNA Integration and Facilitates Intasome Structural Studies. J Mol Biol 432, 2055-2066 (2020).

Author response:

Public Reviews:

Reviewer #1 (Public review):

[…] Strengths:

The study has several important strengths: (i) the work on GDA stability and competition of GDA with point mutations is a very promising area of research and the authors contribute new aspects to it, (ii) rigorous experimentation, (iii) very clearly written introduction and discussion sections. To me, the best part of the data is that deletion of lon stimulates GDA, which has not been shown with such clarity until now.

Weaknesses:

The minor weaknesses of the manuscript are a lack of clarity in parts of the results section (Point 1) and the methods (Point 2).

We thank the reviewer for their comments and suggestions on our manuscript. We also appreciate the succinct summary of key findings that the Reviewer has taken cognisance of in their assessment, in particular the association of the Lon protease with the propensity for GDAs as well as its impact on their eventual fate. Going ahead, we plan to revise the manuscript for greater clarity as suggested by Reviewer #1.

Reviewer #2 (Public review):

[…] The study does what any bold and ambitious study should: it contains large claims and uses multiple sorts of evidence to test those claims.

Weaknesses:

While the general argument and conclusion are clear, this paper is written for a bacterial genetics audience that is familiar with the manner of bacterial experimental evolution. From the language to the visuals, the paper is written in a boutique fashion. The figures are even difficult for me - someone very familiar with proteostasis - to understand. I don't know if this is the fault of the authors or the modern culture of publishing (where figures are increasingly packed with information and hard to decipher), but I found the figures hard to follow with the captions. But let me also consider that the problem might be mine, and so I do not want to unfairly criticize the authors.

For a generalist journal, more could be done to make this study clear, and in particular, to connect to the greater community of proteostasis researchers. I think this study needs a schematic diagram that outlines exactly what was accomplished here, at the beginning. Diagrams like this are especially important for studies like this one that offer a clear and direct set of findings, but conduct many different sorts of tests to get there. I recommend developing a visual abstract that would orient the readers to the work that has been done.

Next, I will make some more specific suggestions. In general, this study is well done and rigorous, but doesn't adequately address a growing literature that examines how proteostasis machinery influences molecular evolution in bacteria.

While this paper might properly test the authors' claims about protein quality control and evolution, the paper does not engage a growing literature in this arena and is generally not very strong on the use of evolutionary theory. I recognize that this is not the aim of the paper, however, and I do not question the authors' authority on the topic. My thoughts here are less about the invocation of theory in evolution (which can be verbose and not relevant), and more about engagement with a growing literature in this very area.

The authors mention Rodrigues 2016, but there are many other studies that should be engaged when discussing the interaction between protein quality control and evolution.

A 2015 study demonstrated how proteostasis machinery can act as a barrier to the usage of novel genes: Bershtein, S., Serohijos, A. W., Bhattacharyya, S., Manhart, M., Choi, J. M., Mu, W., ... & Shakhnovich, E. I. (2015). Protein homeostasis imposes a barrier to functional integration of horizontally transferred genes in bacteria. PLoS genetics, 11(10), e1005612

A 2019 study examined how Lon deletion influenced resistance mutations in DHFR specifically: Guerrero RF, Scarpino SV, Rodrigues JV, Hartl DL, Ogbunugafor CB. The proteostasis environment shapes higher-order epistasis operating on antibiotic resistance. Genetics. 2019 Jun 1;212(2):565-75.

A 2020 study did something similar: Thompson, Samuel, et al. "Altered expression of a quality control protease in E. coli reshapes the in vivo mutational landscape of a model enzyme." Elife 9 (2020): e53476.

And there's a new review (preprint) on this very topic that speaks directly to the various ways proteostasis shapes molecular evolution:

Arenas, Carolina Diaz, Maristella Alvarez, Robert H. Wilson, Eugene I. Shakhnovich, C. Brandon Ogbunugafor, and C. Brandon Ogbunugafor. "Proteostasis is a master modulator of molecular evolution in bacteria."

I am not simply attempting to list studies that should be cited, but rather, this study needs to be better situated in the contemporary discussion on how protein quality control is shaping evolution. This study adds to this list and is a unique and important contribution. However, the findings can be better summarized within the context of the current state of the field. This should be relatively easy to implement.

We thank the reviewer for their encouraging assessment of our manuscript. We appreciate that the manuscript may not be accessible for a general readership in its present form. We plan to revise the manuscript, in part by modifying figures and adding schematics, to afford greater clarity. We also appreciate the concern regarding situating this study in the context of other published work that relates proteostasis and molecular evolution. Indeed, this was a particularly difficult aspect for us given the different kinds of literature that were needed to make sense of our study. We plan on revising the manuscript by incorporating the references that the Reviewer has pointed out.

Reviewer #3 (Public review):

[…] Strengths:

The major strength of this paper is identifying an example of antibiotic resistance evolution that illustrates the interplay between the proteolytic stability and copy number of an antibiotic target in the setting of antibiotic selection. If the weaknesses are addressed, then this paper will be of interest to microbiologists who study the evolution of antibiotic resistance.

Weaknesses:

Although the proposed mechanism is highly plausible and consistent with the data presented, the analysis of the experiments supporting the claim is incomplete and requires more rigor and reproducibility. The impact of this finding is somewhat limited given that it is a single example that occurred in a lon strain and compensatory mutations for evolved antibiotic resistance mechanisms are described. In this case, it is not clear that there is a functional difference between the evolution of copy number versus any other mechanism that meets a requirement for increased "expression demand" (e.g. promoter mutations that increase expression and protein stabilizing mutations).

We thank the reviewer for their in-depth assessment of our work and appreciate their concerns regarding reproducibility and rigor in analysis of our data. We will incorporate this feedback and provide the necessary clarifications in the revised version of our manuscript.

Author Response:

We would like thank reviewers for your comprehensive and insightful reviews of our manuscript. We highly value your constructive comments and suggestions and are preparing revisions that will enhance both the clarity and robustness of our study. Below is an outline of the changes we will implement in response to the points you raised.

All three reviewers expressed concerns regarding the robustness of our conclusions about the relationship between task-related theta activity and aperiodic changes. We will revise the manuscript to present these conclusions more cautiously, stating that the findings indicate a potential contribution of aperiodic activity to what is traditionally interpreted as theta activity. While our results emphasize the importance of distinguishing between periodic and aperiodic components, further research is necessary to fully understand this relationship. We will conduct additional control analyses, including a comparison of the scalp topographies of theta and aperiodic components, to better understand the relationship between aperiodic and periodic (theta) activity.

In response to Reviewer #1's request for greater transparency in our reporting of methodological details, we will provide key clarifications. We will add a clear statement noting that the primary results are based on data from middle-aged to older adults, some of whom had subjective cognitive complaints (SCC). However, it is important to note that no differences were observed between the SCC group and the control group regarding periodic or aperiodic changes in power. Additionally, the main findings were replicated in a sample of middle-aged adults.

To address potential confounding factors, we will include an analysis contrasting response-related ERPs with the identified aperiodic components. However, we do not entirely agree with the assertion that this will necessarily clarify the results. ERPs are not inherently distinct from aperiodic (or periodic) activity; they may reflect changes in aperiodic (or periodic) power. In our view, examining aperiodic and periodic power, ERPs, or time-frequency decomposition with baseline correction provides different perspectives on the same data. Nonetheless, the combined analyses and their results are intended to guide future researchers toward the most suitable approach for interpreting this data.

Reviewer #3 raised concerns regarding the task's effectiveness in evoking theta power and the ability of spectral parameterization method (specparam) to adequately quantify background activity around theta bursts. To address these concerns, we will include additional visualizations demonstrating that the task reliably elicited theta (and delta) activity. Regarding the reviewer's concerns about specparam and theta bursts, it is important to clarify that specparam, in the form we used, does not incorporate time information; rather, it can be applied to any power spectral density (PSD), independent of how the PSD is derived. Specparam’s performance depends on the methods used to estimate frequency content. For time-frequency decomposition, we employed superlets (https://doi.org/10.1038/s41467-020-20539-9), which have been shown to resolve short bursts of activity more effectively than other methods. To our knowledge, superlets provide the highest resolution in terms of both time and frequency. Moreover, to improve stability, we performed spectral parameterization on trial-averaged power (in contrast to the approach in https://doi.org/10.7554/eLife.77348). Nonetheless, we will conduct a simulation to test whether specparam can reliably resolve low-frequency peaks over the 1/f activity.

Reviewer #2 suggested that the manuscript would benefit from a more detailed account of the effects. In response, we will include more detailed quantifications of the analyzed effects, such as model error and R² values.

We believe that the planned revisions will strengthen the manuscript and address the primary concerns raised by the reviewers. We sincerely appreciate your thoughtful feedback and look forward to submitting an improved version of the manuscript soon.

Once again, thank you for your time and expertise in reviewing our work.

Sincerely,

Andraž Matkovič & Tisa Frelih

Author response:

The following is the authors’ response to the original reviews.

We greatly appreciate reviewer 2 comments with both insightful and clearly evaluated assessments of this study that include, much appreciated reframing and evaluation of the study’s advances in the sleep field. It is a constructive review and provides considerable added value to this study in better defining the biological significance of the findings, including both advances and limitations.  

Reviewer 2 nicely summarized the work as “…highlight(ing) the accumulation and resolution of sleep need centered on the strength of excitatory synapses onto excitatory neurons.”. The reviewer succinctly placed one of the main electrophysiological findings in context of one of the sleep field’s most prevalent views, “that LTP associated with wake, leads to the accumulation of sleep need by increasing neuronal excitability, and by the "saturation" of LTP capacity.” It has been speculated that “This saturation subsequently impairs the capacity for further ongoing learning. This new data provides a satisfying mechanism of this saturation phenomenon (and its restoration by recovery sleep) by introducing the concept of silent synapses.” We want to emphasize that sleep need and its resolution involves more than just homeostasis of excitatory synaptic strength but may also be extended to include homeostasis of excitatory synaptic potential to undergo LTP (a homeostasis of meta-plasticity), with implications for learning and memory.   

Reviewer 2 also identified another advance made by this study, summarized as, “The new snRNAseq dataset indicates the sleep need is primarily seen (at the transcriptional level) in excitatory neurons, consistent with a number of other studies.” References for these studies are nicely provided by the reviewer. Our analysis of this data extends the evidence for transcriptional sleep-need-driven changes, observed by us and others in excitatory neurons to more particularly involve the excitatory neurons in layers 2-5, targeting  intra-telencephalic neurons.  

Reviewer 2, importantly noted, “New snRNAseq analysis indicates that SD drives the expression of synaptic shaping components (SSCs) consistent with the excitatory synapse as a major target for the restorative basis of sleep function”, and that “SD-induced gene expression is also enriched for autism spectrum disorder (ASD) risk genes”. These comments are well appreciated as they emphasize that beyond identification of the major target cell type of sleep function, the major sleep-target, gene-ontological characteristics are starting to be addressed.

Reviewer 2 commented on the molecular sleep model, making a key observation that “SDinduced gene expression in excitatory neurons overlaps with genes regulated by the transcription factor MEF2C and HDAC4/5 (Figure 4),” and accurately discusses the significance with respect to the proposed model.

We are in complete agreement with the observation that the molecular sleep model presented is not “definitively supported by the new data and in this regard should be viewed as a perspective…”. One of the more glaring gaps in supporting evidence is the absence of understanding of the role of HDAC4/5 (part of the SIK3-HDAC4/5 pathway) in sleep need modulation of excitatory synapses. Resolution of this issue might be approached by assessment of the synaptic effects of constitutively nuclear HDAC4/5. The current study provides a first step in the assessment by showing a correlation between HDAC4/5 and MEF2c target genes and a subset of differentially expressed synaptic shaping component (SSC) genes that modulate excitatory synapse strength and phenotype. However, the functional studies have yet to be completed. Complimentary studies on SD-induced SSC-DEGs (identified in this study) are also needed for follow-up characterization of their sleep need induced functional impact (both strength and meta-plasticity modulation) on the most relevant excitatory synapses (as identified in the current study).

We agree with both reviewers 1 and 2 that, “Additional work is also needed to understand the mechanistic links between SIK3-HDAC4/5 signaling and MEF2C activity”. Reviewer 2 clarifies the key unresolved issue as, “cnHDAC4/5 suppresses NREM amount and NREM SWA but had no effect on the NREM-SWA increase following SD (Zhou et al., Nature 2022). Loss of MEF2C in CaMKII neurons had no effect on NREM amount and suppressed the increase in NREM-SWA following SD (Bjorness et al., 2020)”. One may conclude with reviewer 2, “These instances indicate that cnHDAC4/5 and loss of MEF2C do not exactly match suggesting additional factors are relevant in these phenotypes.”

An understanding of the mechanism(s) responsible for the relationship between sleep need and SWA are critical to the evaluation of sleep need’s correlation with sleep DEGs and synaptic transmission, including “additional factors” as suggested by reviewer 2. SWA might result from a decrease of cortical glutamatergic neurotransmission below some threshold, which might occur in response to prolonged waking (possibly in response to waking activity-induced local increases of adenosine?), rather than being a cause of, or, being intimately involved in resolving sleep need.  

An increase of SWA in association with SD can result directly from an acute SD-induced increase in local adenosine concentration. This will elicit an ADORA1-mediated down-regulation of glutamate excitatory neurotransmission in the cortex (Bjorness et al., 2016) and in cholinergic arousal centers (Rainnie et al., 1994; Porkka-Heiskanen et al., 1997; Portas et al., 1997; Li et al., 2023). When MEF2c is derepressed by chronic loss of HDAC4 function, SWA is facilitated (Kim et al., 2022). It is plausible that loss of HDAC4 function contributes to the increased SWA by downscaling glutamate excitatory transmission (independent of sleep need). This is expected to result from derepressed, MEF2c mediated sleep-gene expression.  

Similarly, over-expression of constitutively active HDAC4 (cnHD4) can contribute to chronic upscaling of cortical glutamate synaptic strength to depress SWA (again, independent of sleep need). Thus, facilitation or depression of SWA correlates with up or down scaling effects on cortical glutamate neurotransmission, respectively, even in the absence of  a direct effects on sleep need (Figure 4D). Many reagents that reduce the excitability of glutamate pyramidal cells by various mechanisms, including anesthetics like isoflurane, barbiturates or benzodiazepines in addition to those activating ADORA1, increase SWA. Finally, it is important to acknowledge that direct evidence for this proposed link of SWA to cortical glutamate transmission remains in need of further investigation. Thus, SWA may reflect generalized cortical glutamate synaptic activity whether modulated by sleep function or by other agents.

Still, other factors that can have a role mediating some of the mis-match between cnHD4/5 DEGs and Mef2c-cKO DEGs, include the broader over-expression of AAV-cnHD4 compared to CamKII- driven Cre KO of Mef2c. The cnHD4 overexpression can increase arousal center activity in the hypothalamus and other arousal areas to interfere with SWA, but not to the exclusion of SD-DEG repression resulting from a repression of MEF2c-mediated sleep gene expression.

The critique by reviewer 1 raises a number of important technical issues with this study. A key, potentially critical issue raised by reviewer 1, is that of our method of experimental sleep deprivation (ESD). The reviewer suggests that “…neuronal activity/induction of plasticity”, peculiar to the ESD methodology employed in this study, “…rather than sleep/wake states are responsible for the observed results…”.  

In this study, a slow-moving treadmill (SMTM; 0.1km/hour, as stated in the methods), requiring locomotion to avoid bumping into the backwall of a false bottomed plexiglass cage was used to induce ESD. A mouse, in its home cage, typically moves much faster than 0.1km/hour and the mouse is able to eat and drink freely while in the cage (see file: video 1). Furthermore, our observations using a beam-break cage, indicate that mice spontaneously travel for comparable to longer distances over 6 hours than the treadmill moves (during the ESD of 6 hours). Finally, our EEG recordings of mice on the active treadmill show 100% waking while it is on (Bjorness et al., 2009), whereas prevention of NREM sleep (including transition time) using the “gentle handling”  (GH) technique occurs depending on the diligence of the experimenter.  

The accommodation (one week prior to ESD) included exposure to the treadmill-on for 30minutes ~ZT=2 & ZT= 14 hours (now spelled out in the “Materials & Methods” section). Thus, the likelihood of motor learning seems vanishingly small.  

As with all ESD methods, there must be some associated increase in sensory and motor neuronal activity to drive arousal and prevent transition to sleep. For example, the more widely employed GH method of ESD involves sensory stimulation (tactile and or auditory) of sufficient intensity to induce postural change from that associated with sleep to that associated with wake (often involving some locomotion). Like the SMTM, both sensory and motor systems are likely to be engaged. Unlike the SMTM method, the stimulation used in GH is variably-intermittent from mouse to mouse and from experimenter to experimenter as it is applied only when the experimenter judges the mouse to be falling asleep. . It can even be argued that the varied and unpredictable ways in which these interactions happen cause plastic changes with a higher likelihood than the constant slow motion of a treadmill – the mice know how to walk, after all. In other protocols, novel objects are introduced to the animals – those will certainly trigger plastic processes –something that is avoided using a slow-running treadmill to which the mouse has been accommodated, for sleep deprivation.  

The changes induced by SMTM technique are reproducible and induce arousal by somatic stimulation of sufficient intensity to induce natural motor activity as with GH. All ESD methods induce motor activity and it is reasonable to speculate that induced, motor activity is essential for effective ESD for the prolonged durations (>4 hours in mice) that elicit high sleep need. Electrophysiological assessment of SD-evoked increases in mEPSC amplitude and frequency using GH-ESD (Liu et al., 2010) are similar in all respects to our observations of the response to SMTMESD (Bjorness et al., 2020). Further studies might directly address a comparison of SMTM-ESD to GH-ESD as suggested by reviewer 1 but are regrettably outside the scope and resources of our study.

The model presented in Figure 4C is consistent with the experimental findings with respect to the observed electrophysiological changes (including loss of silent synapses and increased AMPA/NMDA ratio after ESD of 6 hours) and altered gene expression that includes enrichment of SSC genes, many of which (7 candidates are listed) can affect both AMPA/NMDA ratio and silent synapses. No claim of mechanism linking the changed expression to altered AMPAR or NMDAR activity can be made at this point, even as to polarity of gene expression, related to electrophysiological outcome. Furthermore, some transcripts may involve receptor trafficking while others more directly affect activated receptor function. To help illustrate the complexity of interpreting gene up-regulation, consider the following hypothetical scenario. If a gene like upregulated Grin3a acts rapidly, it may facilitate reduction of NMDAR function (decreasing plasticity) during ESD, whereas upregulation of a gene like Kif17, if acting in a more delayed manner, might enhance NMDAR surface expression and activity (increasing silent synapses) in response to ESD, during recovery sleep. Relevant references, consistent with these various outcomes are supplied in the manuscript but further investigation is clearly needed, or as reviewer 2 so aptly commented, this work “…provides a framework to stimulate further research and advances on the molecular basis of sleep function”.  

Several issues are raised by reviewer 1 concerning the electrophysiological methodology and statistical assessment. In regard to the former, we closely followed established protocols employed in the frontal neocortex (Myme et al., 2003). We did not include the details for series resistance monitoring. Series resistance values ranged between 8 and 15 MOhm and experiments with changes larger than 25% not used for further analyses. Thank you for bringing this  oversight on our part, to our attention. This essential information, that is unfailingly gathered for all our whole cell recordings, is now added to the version of record.

The -90 mV holding potential was chosen according to precedent (Myme et al., 2003). It increases driving force and permits lower stimulus strength for the same response size – reducing the likelihood for polysynaptic responses. Experiments with multiple response peaks at -90 mV were not included in the analysis. The -90 mV holding potential also increases NMDA receptor Mg++ block resulting in a minimally contaminated AMPA response. This information is now added to our submitted version of record.

The statistical assessments shown in Table 1 refer to two sets of data measured from 3X2=6 different cohorts for each sleep condition (CS, SD, RS): 1) AMPA & NMDA EPSCs and 2) AMPA/NMDA FR ratios (FRR; now bolded in row 1, second tab, Table S1). As stated in the results section, “A two-way ANOVA analysis showed a significant interaction between AMPA matched to NMDA EPSC response for each neuron, and sleep condition (F (2, 21) = 7.268, p<0.004; Figure 1 A, C, E). When considered independently, neither the effect of sleep condition nor of EPSC subtype reached significance at p<0.05 (Figure 1 C)”.  

As noted by reviewer 1, we inadvertently dropped one of the data points from the RS FR and FR ratio (FRR) statistical analysis (raw data in the third tab of Table S1, statistical data in fourth and fifth tab and illustrated in figure 1 F). Thanks to this appreciated, rigorous review, we can correct the oversight (using raw data unchanged in Table S1, third tab). The Table S1 and figure 1 F are now corrected for the version of record. For better clarity, we now use two tabs, the fourth and fifth tabs, respectively of Table S1, for separate stat analyses of FR and FRR data.

The significance of the AMPA/NMDA FRR across sleep conditions was assessed with the KruskalWallis test, a non-parametric method. The two-stage linear step-up procedure of Benjamini, Krieger, and Yekutieli (BKY) was used to control for the FDR across multiple sleep conditions, in the non-parametric Kruskal-Wallis test but it is usually less powerful than tests presuming normal distributions like the one-way ANOVA and Holm-Sidak’s test. We have now added re-analyzed  FRR across CS, SD and RS conditions using a normal one-way ANOVA (Table S1, tab5). The results now read, “The difference between  sleep conditions and FRR is significant (F (2, 19) = 11.3, Table S1, tab5). Multiple comparisons (Holm-Sidak, Table S1, tab5) indicate the near absence of silent synapses was reversed by either CS or RS (SD/CS; p<0.0011 and SD/RS: p<0.0006; Table S1, tab 5; Figure 1 F).”. These analyses compare well to the non-parametric assessment using the  KruskalWallis test (significant at p= 0.0006) with BYK correction for multiple comparison analysis to give for CS-SD, p<= 0.0262 and for RS-SD, p<= 0.0006 (statistics also shown in Table S1, tab5). [Also shown in tab5 is the “standard approach of correcting for family wise error rate”, namely, Dunn’s test. It is more conservative but less powerful than the BYK correction- in general the tradeoff of greater power/ less conservative is better tolerated when many comparisons are made, however, it can be argued that in the present analysis type 2 errors are also potentially misleading and thus not well tolerated.]  The modifications of our statistical analyses, inspired by reviewer 1,  did not affect the interpretation of the data nor the conclusions.  

Bjorness TE, Kelly CL, Gao T, Poffenberger V, Greene RW (2009) Control and function of the homeostatic sleep response by adenosine A1 receptors. The Journal of neuroscience : the official journal of the Society for Neuroscience 29:1267-1276.

Bjorness TE, Dale N, Mettlach G, Sonneborn A, Sahin B, Fienberg AA, Yanagisawa M, Bibb JA, Greene RW (2016) An Adenosine-Mediated Glial-Neuronal Circuit for

Homeostatic Sleep. The Journal of neuroscience : the official journal of the Society for Neuroscience 36:3709-3721.

Bjorness TE, Kulkarni A, Rybalchenko V, Suzuki A, Bridges C, Harrington AJ, Cowan CW, Takahashi JS, Konopka G, Greene RW (2020) An essential role for MEF2C in the cortical response to loss of sleep in mice. Elife 9.

Kim SJ et al. (2022) Kinase signalling in excitatory neurons regulates sleep quantity and depth. Nature 612:512-518.

Li B, Ma C, Huang YA, Ding X, Silverman D, Chen C, Darmohray D, Lu L, Liu S, Montaldo G, Urban A, Dan Y (2023) Circuit mechanism for suppression of frontal cortical ignition during NREM sleep. Cell 186:5739-5750 e5717.

Liu ZW, Faraguna U, Cirelli C, Tononi G, Gao XB (2010) Direct evidence for wake-related increases and sleep-related decreases in synaptic strength in rodent cortex. The Journal of neuroscience : the official journal of the Society for Neuroscience 30:8671-8675.

Myme CI, Sugino K, Turrigiano GG, Nelson SB (2003) The NMDA-to-AMPA ratio at synapses onto layer 2/3 pyramidal neurons is conserved across prefrontal and visual cortices. Journal of neurophysiology 90:771-779.

Porkka-Heiskanen T, Strecker RE, Thakkar M, Bjorkum AA, Greene RW, McCarley RW (1997) Adenosine: a mediator of the sleep-inducing effects of prolonged wakefulness. Science 276:1265-1268.

Portas CM, Thakkar M, Rainnie DG, Greene RW, McCarley RW (1997) Role of adenosine in behavioral state modulation: a microdialysis study in the freely moving cat. Neuroscience 79:225-235.

Rainnie DG, Grunze HC, McCarley RW, Greene RW (1994) Adenosine inhibition of mesopontine cholinergic neurons: implications for EEG arousal. Science 263:689692.

Author response

We appreciate the positive comments and constructive suggestions from the editors and reviewers, which will help us improve our manuscript. We will implement the changes as requested by the reviewers, focusing primarily on revising and clarifying the following aspects:

First, we will clarify the use of biological and technical replicates in each experiment and provide more details about the statistical analyses conducted. Additionally, we plan to include a schematic representation of the experimental design.

Second, we will explain the experiment conducted to rule out hormonal effects or differences in the oocyte maturation method used. We will also indicate the concentration of OVGP1 in the oviduct and explain why we selected OVGP1 as the probable cause of species specificity.

Third, by addressing all of the reviewers' suggestions, we aim to resolve any concerns, inconsistencies, or minor errors identified by the reviewers.

We are committed to addressing all the issues raised by the reviewers and believe that the manuscript will greatly benefit from the insightful suggestions and invaluable contributions of the editors and reviewers.

Author response:

The following is the authors’ response to the original reviews.

Public Reviews:

Reviewer #1 (Public Review):

Summary:

The paper begins with phenotyping the DGRP for post-diapause fecundity, which is used to map genes and variants associated with fecundity. There are overlaps with genes mapped in other studies and also functional enrichment of pathways including most surprisingly neuronal pathways. This somewhat explains the strong overlap with traits such as olfactory behaviors and circadian rhythm. The authors then go on to test genes by knocking them down effectively at 10 degrees. Two genes, Dip-gamma and sbb, are identified as significantly associated with post-diapause fecundity, and they also find the effects to be specific to neurons. They further show that the neurons in the antenna but not the arista are required for the effects of Dip-gamma and sbb. They show that removing the antenna has a diapause-specific lifespan-extending effect, which is quite interesting. Finally, ionotropic receptor neurons are shown to be required for the diapause-associated effects.

Strengths and Weaknesses:

Overall I find the experiments rigorously done and interpretations sound. I have no further suggestions except an ANOVA to estimate the heritability of the post-diapause fecundity trait, which is routinely done in the DGRP and offers a global parameter regarding how reliable phenotyping is. A minor point is I cannot find how many DGRP lines are used.

Thank you for the suggestions. We screened 193 lines and we will add that information to the methods. Additionally, we will add the heritability estimate of the post-diapause fecundity trait.

Reviewer #2 (Public Review):

Summary

In this study, Easwaran and Montell investigated the molecular, cellular, and genetic basis of adult reproductive diapause in Drosophila using the Drosophila Genetic Reference Panel (DGRP). Their GWAS revealed genes associated with variation in post-diapause fecundity across the DGRP and performed RNAi screens on these candidate genes. They also analyzed the functional implications of these genes, highlighting the role of genes involved in neural and germline development. In addition, in conjunction with other GWAS results, they noted the importance of the olfactory system within the nervous system, which was supported by genetic experiments. Overall, their solid research uncovered new aspects of adult diapause regulation and provided a useful reference for future studies in this field.

Strengths:

The authors used whole-genome sequenced DGRP to identify genes and regulatory mechanisms involved in adult diapause. The first Drosophila GWAS of diapause successfully uncovered many QTL underlying post-diapause fecundity variations across DGRP lines. Gene network analysis and comparative GWAS led them to reveal a key role for the olfactory system in diapause lifespan extension and post-diapause fecundity.

Weaknesses:

(1) I suspect that there may be variation in survivorship after long-term exposure to cold conditions (10ºC, 35 days), which could also be quantified and mapped using genome-wide association studies (GWAS). Since blocking Ir21a neuronal transmission prevented flies from exiting diapause, it is possible that natural genetic variation could have a similar effect, influencing the success rate of exiting diapause and post-diapause mortality. If there is variation in this trait, could it affect post-diapause fecundity? I am concerned that this could be a confounding factor in the analysis of post-diapause fecundity. However, I also believe that understanding phenotypic variation in this trait itself could be significant in regulating adult diapause.

We agree that it is possible that the ability to endure cool temperatures per se may influence post-diapause fecundity. However, cool temperature is the essential diapause-inducing condition in Drosophila, so it is not obvious how to separate those effects experimentally, and we agree that phenotypic variation in the cool-sensitivity trait itself could be significant in regulating diapause.

(2) On p.10, the authors conclude that "Dip-𝛾 and sbb are required in neurons for successful diapause, consistent with the enrichment of this gene class in the diapause GWAS." While I acknowledge that the results support their neuronal functions, I remain unconvinced that these genes are required for "successful diapause". According to the RNAi scheme (Figure 4I), Dip-γ and sbb are downregulated only during the post-diapause period, but still show a significant effect, comparable to that seen in the nSyb Gal4 RNAi lines (Figure 4K).

Our definition of successful diapause is the ability to produce viable adult progeny post-diapause, which requires that the flies enter, maintain, and exit diapause, alive and fertile. We will restate our conclusion to say that Dip-γ and sbb are required for post-diapause fecundity.

In addition, two other RNAi lines (SH330386, 80461) that did not show lethality did not affect post-diapause fecundity.

We interpret those results to mean that those RNAi lines were not effective since Dip-γ and sbb are known to be essential.

Notably, RNAi (27049, KK104056) substantially reduced non-diapause fecundity, suggesting impairment of these genes affects fecundity in general regardless of diapause experience. Therefore, the reduced post-diapause fecundity observed may be a result of this broader effect on fecundity, particularly in a more "sensitized" state during the post-diapause period, rather than a direct regulation of adult diapause by these genes.

Ubiquitous expression of RNAi lines #27049 or #KK104056 was lethal, so we included the tubGAL80ts repressor to prevent RNAi from taking effect during development. Flies had to be shifted to 30 °C to inactivate the repressor and thereby activate the RNAi. At 30 °C, fecundity of the controls (GFP RNAi lines #9331, KK60102) were also lower (average non-diapause fecundity = 12 and 19 respectively) and similar to #27049 or #KK104056. We also assessed the knockdown using Repo GAL4 and nSyb GAL4 and did not find a significant difference/decline in the non diapause fecundity for #27049 and #KK104056 as compared to a nonspecific RNAi control (#54037).

(3) The authors characterized 546 genetic variants and 291 genes associated with phenotypic variation across DGRP lines but did not prioritize them by significance. They did prioritize candidate genes with multiple associated variants (p.9 "Genes with multiple SNPs are good candidates for influencing diapause traits."), but this is not a valid argument, likely due to a misunderstanding of LD among variants in the same gene. A gene with one highly significantly associated variant may be more likely to be the causal gene in a QTL than a gene with many weakly associated variants in LD. I recommend taking significance into account in the analysis.

We agree with the reviewer, and in Supplemental Table S3 we list top-associated SNPs in order from the lowest (most significant) p-value. Most of the top-associated genes from this analysis were uncharacterized CG numbers for which there were insufficient tools available for validation purposes. Nevertheless, there is overlap amongst the highly significant genes by p-value and those with multiple SNPs. Amongst the top 15 genes with multiple associated SNPs- CG18636 & CR15280 ranked 3rd by p-value, CG7759 ranked 4th, CG42732 ranked 10th, and Drip ranked 30th (all above the conservative Bonferroni threshold of 4.8e-8) while three Sbb-associated SNPs also appear in Table 3 above the standard e-5 threshold.

Reviewer #3 (Public Review):

Summary:

Drosophila melanogaster of North America overwinters in a state of reproductive diapause. The authors aimed to measure 'successful' D. melanogaster reproductive diapause and reveal loci that impact this quantitative trait. In practice, the authors quantified the number of eggs produced by a female after she exited 35 days of diapause. The authors claim that genes involved with olfaction in part contribute to some of the variation in this trait.

Strengths:

The work used the power platform of the fly DRGP/GWAS. The work tried to verify some of the candidate loci with targeted gene manipulations.

Weaknesses:

Some context is needed. Previous work from 2001 established that D. melanogaster reproductive diapause in the laboratory suspends adult aging but reduces post-diapause fecundity. The work from 2001 showed the extent fecundity is reduced is proportional to diapause duration. As well, the 2001 data showed short diapause periods used in the current submission reduce fecundity only in the first days following diapause termination; after this time fecundity is greater in the post-diapause females than in the non-diapause controls.

The 2001 paper by Tatar et al. reports the number of eggs laid after 3, 6, or 9 weeks in diapause conditions. Thus the diapause conditions used in this study (35 days or 5 weeks) are neither short nor long, rather intermediate. Does the reviewer have a specific concern?

In this context, the submission fails to offer a meaningful concept for what constitutes 'successful diapause'. There is no biological rationale or relationship to the known patterns of post-diapause fecundity. The phenotype is biologically ambiguous.

We have unambiguously defined successful diapause as the ability to produce viable adult progeny post-diapause. Other groups have measured % of flies that arrest ovarian development or % of post-diapause flies with mature eggs in the ovary, or # eggs laid post-diapause; however we suggest that # of viable adult progeny produced post-diapause is more meaningful than the other measurements from the point of view of perpetuating the species.

I have a serious concern about the antenna-removal design. These flies were placed on cool/short days two weeks after surgery. Adults at this time will not enter diapause, which must be induced soon after eclosion. Two-week-old adults will respond to cool temperatures by 'slowing down', but they will continue to age on a time scale of day-degrees. This is why the control group shows age-dependent mortality, which would not be seen in truly diapaused adults. Loss of antennae increases the age-dependent mortality of these cold adults, but this result does not reflect an impact on diapause.

We carried out the lifespan study under two different conditions. We either removed the antenna and moved the flies directly to 10 °C or we removed the antenna and allowed a “wound healing” period prior to moving the flies to 10 °C (out of concern that the flies might die quickly because wound healing may be impaired at 10 °C). In both cases, antenna removal shortened lifespan. Furthermore the lifespan extension at 10 °C was similar regardless of whether flies had experienced two weeks at 25 °C or not.

• Appraisal of whether the authors achieved their aims, and whether the results support their conclusions.

The work falls well short of its aim because the concept of 'successful diapause' is not biologically established. The paper studies post-diapause fecundity, and we don't know what that means. The loci identified in this analysis segregate for a minimally constructed phenotype. The results and conclusions are orthogonal.

It is unclear to us why the reviewer has such a negative opinion of measuring post-diapause fecundity, specifically the ability to produce viable progeny post-diapause. The value of this measurement seems obvious from the point of view of perpetuating the species.

• The likely impact of the work on the field, and the utility of the methods and data to the community.

The work will have little likely impact. Its phenotype and operational methods are weakly developed. It lacks insight based on the primary literature on post-diapause. The community of insect diapause investigators are not likely to use the data or conclusions to understand beneficial or pest insects, or the impact of a changing climate on how they over-winter.

The reviewer has not explained why his/her opinion is so negative.

Recommendations for the authors:

Reviewer #1 (Recommendations For The Authors):

(1) Perform an ANOVA to estimate heritability.

We will do this.

(2) List the number of DGRP lines tested.

193

Reviewer #2 (Recommendations For The Authors):

[Minor suggestions]

(1) Check Drosophila italics

We will do this.

(2) It would be informative to include the number of DGRP lines used in this study in the Results and Methods section.

We will include the information that we assessed 193 DGRP lines.

(3) Figure 1C - several dots are missing at the top of the line.

We will correct.

(4) Figures 1E, F - Why use a discontinuous histogram for continuous distribution? Consider using a continuous histogram (e.g. Lafuente et al. (2018) Figure 1C).

We will do this.

(5) Figure 1F - Why have fewer bins than panel E?

Figure 1F is normalized post-diapause fecundity. Individual post-diapause fecundity was normalized to the mean non-diapause fecundity. Then the normalized individual post-diapause fecundity was averaged to get the mean normalized post-diapause fecundity for the DGRP line. So the bins are different in panel E. Please refer to Supplemental Table S1.

(6) Figure 2D - It would be informative to have fold enrichment stats.

The following will be added in the methods section: The Gene Ontology (GO) categories and Q-values from the false discovery rate (FDR)-corrected hypergeometric test for enrichment are reported. Additionally, coverage ratios for the number of annotated genes in the displayed network versus the number of genes with that annotation in the genome are provided. GeneMANIA estimates Q-values using the Benjamini-Hochberg procedure.

(7) Supplementary table (Table S5) or supplemental table (other supplementary tables)? Need consistency (to Supplementary?)

We will change ‘Supplementary Table S5’ to ‘Supplemental Table S5’.

(8) Figure 5D,E - unused ticks on the x-axis.

The unused ticks on the x-axis will be removed from Figures 5D and E.

Reviewer #3 (Recommendations For The Authors):

• Suggestions for improved or additional experiments, data or analyses.

The authors cannot redo the GWAS with an alternative trait that might better reflect 'successful diapause', and I am not even sure what such a trait would involve or mean. Given this limitation, the authors should consider how they can conduct additional experiments to better define, justify, and elaborate how post-diapause reproduction relates to the mechanisms, processes, depth, and 'success' of diapause.

We agree that it is entirely unclear what trait would be a better measure of successful diapause. Other investigators might have chosen to measure something different but there is no reason why a different choice would be a better choice. We do not believe that this is a “limitation.” We believe that we have unambiguously defined and justified  post-diapause reproduction as a measurement of successful diapause with respect to perpetuating the species through a stressful period.

• Recommendations for improving the writing and presentation.

The mechanics of the writing are fine, aside from some typos/grammar issues. But, the paper is conceptually superficial and tautological. It claims to provide a 'stringent criterion' for 'successful diapause', then measures an unjustified trait, then claims this demonstrates variation for 'successful diapause'.

We respectfully disagree with this opinion.

This story is conducted without reference to prior, primary literature or on the mechanisms of reproductive diapause. The presentation may be improved by considering the literature and precedence for what and how reproductive diapause is induced, maintained, and terminated ... in many insects as well as Drosophila

We will revisit our citations of the literature and apologize for any inadvertent omissions.

Author response:

The following is the authors’ response to the original reviews.

In our initial submission, reviewers highlighted that the major limitations of our study were related to both the number of minibinders tested as well as the number of optimizations we evaluated for improving minibinder function. In this revision, we have focused on expanding the minibinders tested. To do so, we selected two previously published minibinders against the epidermal growth factor receptor (EGFR). Selection of EGFR as a target enabled us to evaluate two minibinders that bind at different sites, unlike the previously evaluated binders LCB1 and LCB3 which both bind the same interface on SARS-CoV-2 Spike. Further, using EGFR as a target enabled us to qualitatively compare the efficacy of minibinder-coupled chimeric antigen receptors against an existing anti-EGFR CAR. We believe the results here demonstrate broader generalizability of our approach across binding sites, targets, and minibinders. We hope this addition is sufficient to convince future would-be users of these tools to attempt synthetic receptor engineering using minibinders against their protein of choice.

Reviewers made comments about the presentation of flow data and the use of statistics throughout the manuscript. We did not modify how flow data are presented as the density plots we used are common throughout the field. We have opted to not include statistics – we believe that in the case of most of the experiments we show, our findings are obvious. In cases where statistics would be helpful for discerning whether subtle effects are real – for example, comparing the linker-based optimizations or comparing the anti-EGFR CARs – we believe that other experimental factors like construct expression are sufficient confounds that even in the presence of statistically significant effects we would be leading readers astray to make such claims about our data. As such, we have sought to limit the claims we make and hope that reviewers and audience agree we do not over interpret our data without statistical support.

On more minor points, both reviewers addressed the differences in Figure 5A and 5C, which we addressed in our figure legend and in the previous response to reviews is the result of these data originating from different time points of the same assay. Reviewer #2 believed we should be more staid in our comments about linker optimality, which we have addressed by changing the referenced line in the discussion. Otherwise, we have made no modifications to figures or text beyond the addition of new data.

Author response:

The following is the authors’ response to the previous reviews.

We addressed the issue of “tolerability” in our answers to Reviewer 2 and in the revised manuscript where we had added data concerning tolerability, see the paragraph in the Results Section, page 11:

"Finally, tolerability studies were performed with the administration of up to 20 and 40 mg/kg eq. NT (i.e. 25.8 and 51.6 mg/kg of VH-N412) with n=3 for these doses. The rectal temperature of the animals did not fall below 32.5 to 33.2°C, similar to the temperature induced with the 4 mg/kg eq. NT dose. We observed no mortality or notable clinical signs other than those associated with the rapid HT effect such as a decrease in locomotor activity. We thus report a very interesting therapeutic index since the maximal tolerated dose (MTD) was > 40 mg/kg eq. NT, while the maximum effect is observed at a 10x lower dose of 4 mg/kg eq. NT and an ED50 established at 0.69 mg/kg as shown in Figure 1G.”

We have slightly modified the paragraph above to emphasize that the tolerability studies were performed in “naïve mice”. 

"Finally, tolerability studies were performed in naïve mice with the administration of up to 20 and 40 mg/kg eq. NT (i.e. 25.8 and 51.6 mg/kg of VH-N412) with n=3 for these doses. The rectal temperature of the animals did not fall below 32.5 to 33.2°C, similar to the temperature induced with the 4 mg/kg eq. NT dose. We observed no mortality or notable clinical signs other than those associated with the rapid HT effect such as a decrease in locomotor activity. We thus report a very interesting therapeutic index since the maximal tolerated dose (MTD) was > 40 mg/kg eq. NT, while the maximum effect is observed at a 10x lower dose of 4 mg/kg eq. NT and an ED50 established at 0.69 mg/kg as shown in Figure 1G.”

We propose to add a sentence in the Results section, page 11, relative to the fact that we can also induce severe hypothermia in rats using conjugates similar to VH-N412.

We also added in the Discussion section (page 38) that we could induce hypothermia with different conjugates in mice, rats and pigs.

Author response:

The following is the authors’ response to the original reviews.

Reviewer #1 (Public Review):

(1) Some of the figures are of rather poor quality. For example, the H&E and Sirius Red stainings in Figures 3 and 4 are quite poor so it is difficult to see what is going on in the muscles. The authors should take note of another publication on dy3K/dy3K mice of similar age (PMID: 31586140) where such images are of much higher quality. Similarly, the Western blot for laminin-alpha2 (Figure 4B) of the wild-type mouse needs improvement. If the single laminin-alpha2 protein is not detected, there is an issue with the denaturation buffer used to load the protein.

Thank you for the valuable suggestions. We have read the study on dy3K/dy3K mice of similar age (PMID: 31586140) which showed dystrophic changes in dy3K/dy3K muscle throughout the disease course with the whole muscle and representative muscle area. We have generated new figures with higher quality including the whole muscle and representative muscle area for the H&E and Sirius Red stainings.  However, due to the large images, we have added them in the new Figure supplement 2 and Figure supplement 3. Also, we have changed the denaturation buffer used to load the protein, and performed Western blot of laminin α2, the result of the laminin α2 protein of the wild-type mice (n =3) and dyH/dyH mice (n =3) detected by Western blot has been showed in Figure 4B.

(2) My biggest concern is, however, the many overstatements in the manuscript and the over-interpretation of the data. This already starts with the first sentence in the abstract where the authors write: "Understanding the underlying pathogenesis of LAMA2- related muscular dystrophy (LAMA2-MD) have been hampered by lack of genuine mouse model." This is not correct as the dy3K/dy3K, generated in 1997 (PMID: 9326364), are also Lama2 knockout mice; there are also other strains (dyW/dyW mice) that are severely affected and there are the dy2J/dy2J mice that represent a milder form of LAMA2-MD. Similarly, the last two sentences of the abstract "This is the first reported genuine model simulating human LAMA2-MD. We can use it to study the molecular pathogenesis and develop effective therapies." are a clear overstatement. The mechanisms of the disease are well studied and the above-listed mouse models have been amply used to develop possible treatment options. The overinterpretation concerns the results from transcriptomics. The fact that Lama2 is expressed in particular cell types of the brain does not at all imply that Lama2 knockout mice have a defect in the blood-brain barrier as the authors state. If there are no functional data, this cannot be stated. Indications for a blood-brain barrier defect come from work in dy3K/dy3K mice (PMID: 25392494) and this needs to be written like this.

Thank you for your comment and sorry for the overstatements in the manuscript. We have carefully considered our previous statements and corrected them accordingly. We have changed the first sentence in the abstract into "Our understanding of the molecular pathogenesis of LAMA2-related muscular dystrophy (LAMA2-MD) requires improving". Also, we have changed the last two sentences in the abstract with "In summary, this study provided useful information for understanding the molecular pathogenesis of LAMA2-MD".

We also agree that "Lama2 is expressed in particular cell types of the brain does not at all imply that Lama2 knockout mice have a defect in the blood-brain barrier", and the indications for a blood-brain barrier defect come from work in dy3K/dy3K mice (PMID: 25392494). Therefore, we have corrected the overstatement according to the suggestion with "It was reported that the deficiency of laminin α2 in astrocytes and pericytes was associated with a defective blood-brain barrier (BBB) in the dy3K/dy3K mice (Menezes et al., 2014). The defective BBB presented with altered integrity and composition of the endothelial basal lamina, reduced pericyte coverage, and hypertrophic astrocytic endfeet lacking appropriately polarized aquaporin4 channels."

(3) Finally, the bulk RNA-seq data also needs to be presented in a disease context. The authors, again, mix up changes in expression with functional impairment. All gene expression changes are interpreted as direct evidence of an involvement of the cytoskeleton. In fact, changes in the cytoskeleton are more likely a consequence of the severe muscle phenotype and the delay in muscle development. This is particularly possible as muscle samples from 14-day-old mice are compared; a stage at which muscle still develops and grows tremendously. Thus, all the data need to be interpreted with caution.

Thank you for your comment. We have changed the over-interpretation of the bulk RNA-seq data, and have corrected the last sentence in the Result with "These observations important data for the impaired muscle cytoskeleton and abnormal muscle development which were associated with the muscle pathology consequence of severe dystrophic changes in the dyH/dyH mice.".

(4) In summary, the authors need to improve data presentation and, most importantly, they need to tone down the interpretation and they must be fully aware that their work is not as novel as they present it.

Thank you for your comments and valuable suggestions, and we have changed the previous overstatements and interpretation of the results. We are sorry that we failed to clearly present our rational of making this mouse model. Indeed, there were many existing mouse models, which were all important to the research in the field. One of the reasons why we wished to create dyH/dyH is to make a mouse model without any trace of engineering (e.g., inserted bacterial elements for knockout). By doing so, we were hoping to provide a novel model suited for gene-editing-based gene therapy development. To this end, dyH/dyH was created to reflect the hot mutation region in the Chinese population. Hopefully, you will agree with our points and see that we were not trying to belittle previous models but were simply trying to provide a different option. The overstatements were largely rooted from language barriers, and we have tried to make our statements more cautious and acceptable to the readers.

Reviewer #2 (Public Review):

(1) The major weakness is the manuscript reads like this was the first-ever knockout mouse model generated for LAMA2-CMD. There are in fact many Lama2 knockout mice (dy, dy2J, dy3k, dyW, and more) which have all been extensively studied with publications. It is important for the authors to comment on these other published studies that have generated these well-studied mouse lines. Therefore, there is a lack of background information on these other Lama2 null mice.

Thank you for your comment. We have added background information on these other Lama2 null mice with the sentences "The most common mouse models for LAMA2-MD are the dy/dy, dy3k/dy3k, dyw/dyw and dy2J/dy2J mice (Xu et al., 1994; Michelson et al., 1995; Miyagoe et al., 1997; Kuang et al., 1998; Sunada et al., 1995). Among them, the dy/dy, dy3k/dy3k, dyw/dyw mice present severe muscular dystrophy, and dy2J/dy2J mice show mild muscular dystrophy and peripheral neuropathy (Gawlik and Durbeej, 2020). The mutation of the dy/dy mice has been still unclear (Xu et al., 1994; Michelson et al., 1995). The dy3k/dy3k mice were generated by inserting a reverse Neo element in the 3' end of exon 4 of Lama2 gene in 1997 (Miyagoe et al., 1997), and the dyw/dyw mice were created with an insertion of lacZ-neo in the exon 1 of Lama2 gene in 1998 (Kuang et al., 1998). The dy2J/dy2J mice were generated in 1970 by a spontaneous splice donor site mutation which resulted in a predominant transcript with a 171 base in-frame deletion, leading to the expression of a truncated laminin α2 with a 57 amino acid deletion (residues 34-90) and a substitution of Gln91Glu (Sunada et al., 1995). They were established in the pre-gene therapy era, leaving trace of engineering, such as bacterial elements in the Lama2 gene locus, thus unsuitable for testing various gene therapy strategies. Moreover, insufficient transcriptomic data of the muscle and brain of LAMA2-CMD mouse models limits the understanding of disease hallmarks. Therefore, there is a need to create new appropriate mouse models for LAMA2-CMD based on human high frequently mutated region using the latest gene editing technology such as clustered regularly interspaced short palindromic repeats (CRISPR)-Cas9."

(2) The phenotypes of dyH/dyH are similar to, if not identical to dy/dy, dy2J/dy2J, dy3k/dy3k, dyW/dyW including muscle wasting, muscle weakness, compromised blood-brain barrier, and reduced life expectancy. This should be addressed, and a comparison made with Lama2 deficient mice in published literature.

Thank you for your comment. We have added Table supplement 3 to make a comparison between dyH/dyH with other Lama2 deficient mice. We aslo have added the statement in Discussin with "Compared with other Lama2 deficient mice including dy/dy, dy2J/dy2J, dy3k/dy3k and dyW/dyW, the phenotype of the dyH/dyH mice presented with a very severe muscular dystrophy, which was similar to that of the dy3k/dy3k mice (Table supplement 3)."

(3) Recent published studies (Chen et al., Development (2023), PMID 36960827) show loss of Itga7 causes disruption of the brain-vascular basal lamina leading to defects in the blood-brain barrier. This should be referenced in the manuscript since this integrin is a major Laminin-211/221 receptor in the brain and the mouse model appears to phenocopy the dyH/dyH mouse model.

Thank you for your great suggestion. We have cited the published studies (Chen et al., Development (2023), PMID 36960827) and added statements in Discussion with "As reported, the aberrant BBB function was also associated with the adhesion defect of alpha7 integrin subunit in astrocytes to laminins in the Itga_7-/- mice (_Chen et al., 2023). In this study, loss of communications involving the laminins’ pathway between laminin α2 and integrins were predicted between vascular and leptomeningeal fibroblasts and astrocytes in the dyH/dyH brain, providing more evidence for the impaired BBB due to laminin α2 deficiency."

Recommendations for the authors:

Reviewer #1 (Recommendations For The Authors):

(1) Improve the data presentation (as mentioned above). Make a new picture of the histology; repeat the Western blots. Discuss the RNA-seq data with more caution and present it in a more attractive way. Tone down the wording.

Thank you for your recommendations. We have revised the overstatements and improved the RNA-seq data interpretation as suggested. Also,we have made a new picture of the histology, and repeated the Western blots.

Reviewer #2 (Recommendations For The Authors)

(1) There are many grammatical errors within the manuscript. The manuscript should be carefully proofread.

Thank you for your recommendations. We have carefully corrected the grammatical errors within the manuscript.

(2) Figure 2: The animal numbers used in this analysis were not indicated. Please include this number in the figure legend.

Thank you for your recommendations. We have added animal numbers in the figure legends wherever applicable.

(3) Figure 2: The forelimb grip strength is informative but has limitations. Ex vivo or in vivo muscle contractility is the gold standard for measuring muscle strength.

Thank you for your recommendations. We do agree that the ex vivo or in vivo muscle contractility is the gold standard for measuring muscle strength, and we really want to finish this experiment. However, we feel sorry that this test has not been finished due to the following reasons: (1) The forelimb grip strength for measuring muscle strength is a classic method and remains a commonly used method for measuring mouse muscle strength in the studies of different muscular dystrophies, such as LAMA2-MD (Amelioration of muscle and nerve pathology of Lama2-related dystrophy by AAV9-laminin-αLN linker protein. JCI Insight. 2022;7(13):e158397. PMID: 35639486), Duchenne muscular dystrophy (Investigating the role of dystrophin isoform deficiency in motor function in Duchenne muscular dystrophy. J Cachexia Sarcopenia Muscle. 2022;13(2):1360-1372. PMID: 35083887), facioscapulohumeral muscular dystrophy (Systemic delivery of a DUX4-targeting antisense oligonucleotide to treat facioscapulohumeral muscular dystrophy. Mol Ther Nucleic Acids. 2021;26:813-827. PMID: 34729250), and etc. (2) The forelimb grip strength for measuring muscle strength is also used in the human studies (PMID: 32366821; PMID: 29313844; PMID: 34499663, and etc). In view of reasons above, for measuring muscle strength, we used the forelimb grip strength, and have not finished the supplementary experiment of ex vivo or in vivo muscle contractility.

(4) Figure 3: Muscle fibrosis should be measured with a hydroxyproline assay.

Thank you for your recommendations. We do agree that the hydroxyproline assay is one of the most classic method to evaluate collagen content for measuring muscle fibrosis. However, we performed Sirius Red staining for measuring muscle fibrosis due to the following reasons: (1) Muscle fibrosis measured by Sirius Red staining can be observed more directly, and the other pathological features also can be observed, and compared through muscle pathology. (2) Sirius Red staining is also a classic method and remains a commonly used method for measuring muscle fibrosis, which has been previously reported in the mouse studies of muscle disorders, such as PMID: 22522482 (Losartan, a therapeutic candidate in congenital muscular dystrophy: studies in the dy(2J) /dy(2J) mouse. Ann Neurol. 2012;71(5):699-708.), PMID: 34337906 (Aging-related hyperphosphatemia impairs myogenic differentiation and enhances fibrosis in skeletal muscle. J Cachexia Sarcopenia Muscle. 2021;12(5):1266-1279.), PMID: 28798156 (Phosphodiesterase 4 inhibitor and phosphodiesterase 5 inhibitor combination therapy has antifibrotic and anti-inflammatory effects in mdx mice with Duchenne muscular dystrophy. FASEB J. 2017;31(12):5307-5320.), and etc. Therefore, we used Sirius Red staining to measure muscle fibrosis in this study.

(5) Figure 8: The N=3 is very low which could result in type I or II statistical errors. A larger sample size will reduce the chance of statistical errors.

Thank you for your recommendations. We have increased the number of animals to reduce the chance of statistical errors. We have performed the supplementary experiment, the number of animals for each group has been increased to 6 (3 male and female each).  The results were consistent with previous data in Figure 8.

(6) Power analysis to estimate experimental animal numbers should be reported in the manuscript.

Thank you for your recommendations. Refer to previous study (Power and sample size. Nature Methods. 2013;10:1139–1140), “The distributions show effect sizes d = 1, 1.5 and 2 for n = 3 and α = 0.05. Right, power as function of d at four different a values for n = 3”, and “If we average seven measurements (n = 7), we are able to detect a 10% increase in expression levels (μ_A = 11, _d = 1) 84% of the time with α = 0.05.”, the experimental animal numbers estimated were 3 to 7. Moreover, if the increased number of experimental animals could be available, we would retain data.

(7) It is unclear if the studies were performed with adequate rigor. Were those scoring outcome measures blinded to the treatment groups?

Thank you for your recommendations. We performed the studies with those scoring outcome measures not blinded to the treatment groups, the groups were based on their genotype. Actually, it was easy to discriminate the dyH/dyH groups from the WT/Het mice due to their small body shape.

(8) Authors should appropriately cite previous studies that have generated Lama2 null mice.

Thank you for your recommendations. We have cited previous studies that have generated Lama2 null mice with the sentence “The most common mouse models for LAMA2-MD are the dy/dy, dy3k/dy3k, dyw/dyw and dy2J/dy2J mice (Xu et al., 1994; Michelson et al., 1995; Miyagoe et al., 1997; Kuang et al., 1998; Sunada et al., 1995)”.

(9) The number of animals should be increased to reduce the chance of statistical error.

Thank you for your recommendations. We have performed the supplementary experiment, the number of animals for each group has been increased to reduce the chance of statistical error.

(10) A power analysis should be performed to determine the number of experimental animals.

Thank you for your recommendations. We have performed a power analysis to determine the number of experimental animals as mentioned above.

(11) There are many grammatical errors within the manuscript. The manuscript should be carefully proofread.

Thank you for your recommendations. We have carefully corrected the grammatical errors within the manuscript.

Author response:

The following is the authors’ response to the original reviews.

Public Reviews:

Reviewer 1:

(1) General comment: The evidence for these highly novel, potentially interesting roles (of the exocyst) would need to be more compelling to support direct involvement.

We wish to thank the reviewer for his/her comments, and for considering that the proposed functions are highly novel and potentially interesting. To strengthen the evidence supporting the new roles of the exocyst, we have performed a number of additional experiments that are depicted in novel figures or figure panels of the new version of the manuscript. Particularly, we aimed at providing further support of the direct involvement of the exocyst in different steps of the regulated secretory pathway. Please see the details below.

(2) For instance, the localization of exocyst to Golgi or to granule-granule contact sites does not seem substantial.

We have performed quantitative colocalization studies, as suggested by the reviewer to further substantiate our initial findings. We have carefully analysed GFP-Sec15 distribution in relation to the Golgi complex and secretory Glue granules at relevant time points of salivary gland development. Overall, we found that GFP-Sec15 distribution is dynamic during salivary gland development. Before Glue synthesis (72 h AEL), Sec15 was observed in close association (defined as a distance equal to, or less than 0.6 µm) with the Golgi complex (please see below Author response image 1). This association was lost once Glue granules have begun to form (96 h AEL). Importantly, we do not see relevant association between GFP-Sec15 and the ER (please see Author response image 2). These observations support our conclusion that the exocyst plays a role at the Golgi complex. New images supporting these conclusions, as well as quantitative data, have been included in Figure 5 of the new version of the manuscript. In addition, real time imaging, as well as 3D reconstruction analyses, confirming the close association between Sec15 and Golgi cisternae are now included in the manuscript. Please see Supplementary Videos 1-3. These new data are described in the text lines 200-210 of the Results section and text lines 359368 of the Discussion section.

Interestingly, at the time when Sec15-Golgi association is lost (96 h AEL), Sec15 foci associate instead with newly formed secretory granules (< 1µm diameter). This association persists during secretory granule maturation (100-116 h AEL), when Sec15 foci localize specifically in between neighbouring, immature secretory granules. When maturation has ended and Glue granule exocytosis begins (116-120 h AEL), this localization between granules is lost. These observations are consistent with a role of the exocyst in homotypic fusion during SG maturation. We have included new images showing that association between Sec15 and secretory granules is dynamic and depends on the developmental stage. We have quantified this association both during maturation and at a stage when SGs are already mature. We have in addition performed a 3D reconstruction analysis of these images to confirm the close association between Sec15 and immature SGs. These new data are now depicted in Figure 7BC, Supplementary Videos 4-5, and described in text lines 216-221 of the Results section. In addition, a lower magnification image is provided below in this letter (Author response image 3), quantifying the proportion of Sec15 foci localized in between SGs (yellow arrows) relative to the total number of Sec15 foci (yellow arrows + green arrowheads).

Author response image 1.

Criteria utilized to define Sec15 focithat were“associated” or“not associated” withthe trans-Golgi network in the experiments of Figure 5C-E of the manuscript.When the distance between maximal intensities of GFP-Sec15 and Golgi-RFP signals was equal or less than 0.6 m, the signals were considered “associated” (upper panels). When the distance was more than 0.6 m, the signals were considered “not associated” (lower panels).

Author response image 2.

Criteria utilized to define Sec15 focithat were“associated” or“not associated” withthe ERin the experiments of Figure 5A-Bof the manuscript.When the distance between maximal intensities of GFP-Sec15 and KDEL-RFP signals was equal or less than 0.6 m, the signals were considered “associated”. When the distance was more than 0.6 m, the signals were considered “not associated”.

Author response image 3.

(A) GFP-Sec15 foci (cyan) and SGs (red) are shown in cells bearing Immature SGs or (B) with mature SGs. Yellow arrows indicate GFP-Sec15 foci localized in between SGs; green arrowheads indicate GFP-Sec15 foci that arenot in between SGs. (C) Quantification of the percentage (%) of Sec15 foci localized in between SGs respect to the total number of Sec15 foci in cells filled with immature SGs (ISG)vs cells with mature SGs (MSG).

It is interesting to mention that previous evidence from mammalian cultured cells (Yeaman et al,  2001) show that the exocyst localizes both at the trans-Golgi network and at the plasma membrane, weighing in favour of our claim that the exocyst is required at various steps of the exocytic pathway. Thus, the exocyst may play multiple roles in the secretion pathway in other biological models as well. This concept has now been included at the Discussion section of the revised version of the manuscript (lines 359-368).

To make the conclusions of our work clearer, in the revised version of the manuscript, we have now included a graphical abstract, summarizing the dynamic localization of the exocyst in relation to the processes of SG biogenesis, maturation and exocytosis reported in our work. 

(3) Instead, it is possible that defects in Golgi traffic and granule homotypic fusion are not due to direct involvement of the exocyst in these processes, but secondary to a defect in canonical exocyst roles at the plasma membrane. A block in the last step of glue exocytosis could perhaps propagate backward in the secretory pathway to disrupt Golgi complexes or cause poor cellular health due to loss of cell polarity or autophagy.

We thank the reviewer for these thoughtful comments. We have performed a number of additional experiments to assess “cellular health” or to identify possible defects in cell polarity after knock-down of exocyst subunits. These new data have been included in new supplementary figures 5 and 6 of the revised version of the manuscript (please see below). 

In our view, the precise localization of GFP-Sec15 at the Golgi complex (Figure 5C-E), as well as in between immature secretory granules (Figure 7B-D), argues in favour of a direct involvement of the exocyst in SG biogenesis and homofusion respectively. 

We truly appreciate the comment of the reviewer raising the possibility that the defects that we observe at early steps of the pathway (SG biogenesis and SG maturation) may actually stem from a backward effect of the role of the exocyst in SG-plasma membrane tethering. We wish to respectfully point out that the processes of biogenesis, maturation and plasma membrane tethering/fusion of SGs do not occur simultaneously in the Drosophila larval salivary gland in vivo, as they do in other secretory model systems (i.e. cell culture). In this regard, the experimental model is unique in terms of synchronization. In each cell of the salivary gland, the three processes (biogenesis, maturation and exocytosis) occur sequentially, and controlled by developmental cues. At the developmental stage when SGs fuse with the plasma membrane, SG biogenesis has already ceased many hours earlier: SG biogenesis occurs at 96-100 hours after egg lay (AEL), SG maturation takes place at 100-112 hours AEL, and SG-plasma membrane fusion happens only when all SGs have undergone maturation and are ready to fuse with the plasma membrane at 116-120 h AEL. Thus, in our view it is not conceivable that a defect in SG-plasma membrane tethering/fusion (116-120 h AEL) may affect backwards the processes of SG biogenesis or SG maturation, which have occurred earlier in development (96-112 h AEL).

As suggested by the reviewer, we have analysed several markers of cellular health and cell polarity, comparing conditions of exocyst subunit silencing (exo70RNAi, sec3RNAi or exo84RNAi) with wild type controls (whiteRNAi). These new data are depicted in Supplementary Figures 5 and 6, and described in lines 172-179 of the Results section of the revised version of the manuscript. Noteworthy, for these experiments we have applied silencing conditions that block secretory granule maturation, bringing about mostly immature SGs. Our analyses included: 1) Subcellular distribution of PI(4,5)P2, 2) subcellular distribution of the tetraspanin CD63, 3) of Rab11, 4) of filamentous actin, and 5) of CD8. We have also compared 6) nuclear size and nuclear general morphology, 7) the number and distribution of mitochondria, 8) morphology and subcellular distribution of the cis- and 9) trans-Golgi networks. Finally, 10) we have compared basal autophagy in salivary cells with or without knocking down exocyst subunits. The markers that we have analysed behaved similarly to those of control salivary glands, suggesting that the observed defects in regulated exocytosis indeed reflect different roles of the exocyst in the secretory pathway, rather than poor cellular health or impaired cell polarity.  

Our conclusions are in line with previous studies in which apico-basal polarity, Golgi complex morphology and distribution, as well as apical membrane trafficking were also evaluated in exocyst mutant backgrounds, finding no anomalies (Jafar-Nejad et al, 2005). 

Conversely, in studies in which apical polarity was disturbed by interfering with Crumbs levels, SG biogenesis, maturation and exocytosis were not affected (Lattner et al, 2019), indicating that these processes not necessarily interfere with one another.  

(4) Final recommendation: In the absence of stronger evidence for these other exocyst roles, I would suggest focusing the study on the canonical role (interesting, as it was previously reported that Drosophila exocyst had no function in the salivary gland and limited function elsewhere [DOI: 10.1034/j.1600-0854.2002.31206.x]), and leave the alternative roles for discussion and deeper study in the future.  

We appreciate the reviewer´s recommendation. However, we believe that the major strength of our work is the discovery of non-canonical roles of the exocyst complex, unrelated to its function as a tethering complex for vesicle-plasma membrane fusion. We believe that in the new version of our manuscript, we provide stronger evidence supporting the two novel roles of the exocyst:

a) Its participation in maintaining the normal structure of the Golgi complex, and b) Its function in secretory granule maturation.

Reviewer 2:

(5) General comment: A key strength is the breadth of the assays and study of all 8 exocyst subunits in a powerful model system (fly larvae). Many of the assays are quantitated and roles of the exocyst in early phases of granule biogenesis have not been ascribed. 

We are grateful that the reviewer appreciates the novelty of our contribution.

(6) However there are several weaknesses, both in terms of experimental controls, concrete statements about the granules (better resolution), and making a clear conceptual framework. Namely, why do KD of different exocysts have different effects on presumed granule formation

The reviewer has raised a point that is central to the interpretation of all our data throughout the manuscript. The short answer is that the extent of RNAi-dependent silencing of exocyst subunits determines the phenotype: 

1) Maximum silencing affects Golgi complex morphology and prevents SG biogenesis. 2) Intermediate silencing blocks SG maturation, without affecting Golgi complex morphology and SG biogenesis. 3) Weak silencing blocks SG tethering and fusion with the plasma membrane, without affecting Golgi complex morphology, SG biogenesis or SG maturation. 

In other words, 1) Low levels of exocyst subunits are sufficient for normal Golgi complex morphology and SG biogenesis. 2) Intermediate levels of exocyst subunits are sufficient for SG maturation (and also sufficient for SG biogenesis). 3) High levels of exocyst subunits are required for SG tethering and subsequent fusion with the plasma membrane. 

Based on the above notion, we have exploited the fact that temperature can fine-tune the level of Gal4/UAS-dependent transcription, thereby achieving different levels of silencing, as shown by Norbert Perrimon et al in their seminal paper “the level of RNAi knockdown can also be altered by using Gal4 lines of various strengths, rearing flies at different temperatures, or via coexpression of UAS-Dicer2” (Perkins et al, 2015). 

We found in our system that indeed, by applying appropriate silencing conditions (RNAi line and temperature) to any of the eight subunits of the exocyst, we have been able to obtain one of the three alternative phenotypes: Impaired SG biogenesis, or impaired SG maturation, or impaired SG tethering/fusion with the plasma membrane.

These concepts are summarized below in Author response image 4. Please see also at point 26, the general comment of Reviewer #3. 

We have conducted qRT-PCR assays to provide experimental support to the notions summarized above in Author response image 4. We measured the remaining levels of mRNAs of some of the exocyst subunits, after inducing RNAi-mediated silencing at different temperatures, or with different RNAi transgenic lines. The remaining RNA levels after silencing correlate well with the observed phenotypes, following the predictions of Author response image 4 and summarized in Author response image 5. These new data are now shown in Supplementary Figure 2 of the revised version of the manuscript, and described in lines 153-159 at the Results section.

(7) Why does just overexpression of a single subunit (Sec15) induce granule fusion?

The reviewer raises a very important point. Based on available data from the literature, Sec15 behaves as a seed for assembly of the holocomplex and it also mediates the recruitment of the holocomplex to SGs through its interaction with Rab11 (Escrevente et al, 2021; Bhuin and Roy, 2019; Wu et al, 2005; Zhang et al, 2004; Guo et al, 1999). Thus, overexpression of Sec15 is expected to enhance exocyst assembly, thereby potentiating the activities carried out by the complex in the cell, including SG homofusion. In the revised version of the manuscript we have also performed the overexpression of Sec8, finding that, unlike Sec15, Sec8 fails to induce homotypic fusion. These results were expected, as they confirm that Sec8 does not behave as a seed for mounting the whole complex. These new data have been included in Figure 7E-H, and are described in text lines 221-229 of the Results section. 

Author response image 4.

Conceptual model of RNAi expression at different temperatures , remaining levels of mRNA/protein levels and phenotypes obtained at each temperature.

Author response image 5.

qRT-PCR assays presented in Supplementary Figure 2 are shown in combination with the phenotypes observed at each of the conditions analyzed. Note the correlation between phenotypes and the extent of mRNA downregulation.

(8) While the paper is fascinating, the major comments need to be addressed to really be able to make better sense of this work, which at present is hard to disentangle direct vs. secondary effects, especially as much of the TGN seems to be altered in the KDs.  

We hope that our response to point 6) has helped to clarify this important point raised by the Reviewer. After applying silencing conditions where normal structure of the trans-Golgi network is impaired, SG biogenesis does not occur. Thus, since SGs do not form, it is not conceivable to detect defects in SG maturation or SG fusion with the plasma membrane in the same cell.

(9) The authors conveniently ascribe many of the results to the holocomplex, but their own data (Fig. 4 and Fig. 6) are at odds with this.

This is another central point of our work, so we thank the reviewer for his/her comment. In Figures 4A, 7A and 9A of the revised version of the manuscript, we show that, by inducing appropriate levels of silencing of any of the 8 subunits of the exocyst, each of the three alternative phenotypic manifestations can occur. In our opinion, this argues in favour of a function for the whole exocyst complex in each of the three specific activities proposed in our study: 1) SG biogenesis, 2) SG maturation, and 3) SG tethering/fusion with the plasma membrane. In detailed characterizations of these three phenotypes performed throughout the study, we decided to induce silencing of just two or three of the subunits of the exocyst, assuming that the whole complex accounts the mechanisms involved.

Major comments

(10) Resolution not sufficient. Identification of "mature secretory granules" (MSG) in Fig. 3 is based on low-resolution images in which the MSG are not clearly seen (see control in Fig. 3A) and rather appear as a diffuse haze, and not as clear granules. There may be granules here, but as shown it is not clear. Thus it would be helpful to acquire images at higher resolution (at the diffraction limit, or higher) to see and count the MSG.

We thank the reviewer for raising this point, as it may not be straightforward to the reader to identify the SGs throughout the figures of our study. To make it clearer, in Figure 3A (magnified insets on the right), we have delimitated individual SGs with a green dotted line, and included diagrams (far right), which we hope will help the identification of SGs. In Figure 3B, we show that after silencing Sec84, a mosaic phenotype was observed: In some cells SGs fail to undergo maturation, and remain smaller than normal. In other cells of this mosaic phenotype, biogenesis of SGs was impaired and the fluorescent cargo remained trapped in a mesh-like structure (that we later show that corresponds to the ER). The dotted line marks individual SGs, and the diagrams included on the right intend to help the interpretation of the phenotype. The mesh-like structures where Sgs3-GFP was retained are also marked with dotted line, and schematized on the right. These new schemes are described in the Figure 3 caption of the revised version of the manuscript.

We wish to mention that all the confocal images depicted in this figure and throughout the manuscript  have been captured at high resolution, with a theoretical resolution limit of 168177nm (d = γ/2NA). Given that secretory granules range from 0.8-7µm in diameter, the resolution is more than sufficient to clearly resolve these structures. 

(11) Note: the authors are not clear on which objective was used. Maybe the air objective as the resolution appears poor).  

In this particular figure, we have utilized a Plan-Apochromat 63X/1.4NA oil objective of the inverted Carl Zeiss LSM 880 confocal microscope (mentioned in materials and methods).

(12) They need to prove that the diffuse Sgs3-GFP haze is indeed due to MSG.  

If we interpret correctly the concern of the reviewer, what he/she calls “diffuse haze” is actually the distribution of Sgs3-GFP within individual SGs, which, as previously reported by other authors, is not homogeneous at this stage (Syed et al. 2022). We hope that the diagrams that we have included in Figure 3 A, B (point 10) will help the readers interpreting the images.   

(13) Related it is unclear what are the granule structures that correspond to Immature secretory granules (ISG) and cells with mesh-like structures (MLS)?

We are confident that the diagrams now included in Figure 3A and B will help the interpretation, and particularly to identify immature granules and the mesh-like structure generated after silencing of exocyst subunits.

(14) Similarly, Sgs3 images of KD of 8 exocyst subunits were interpreted to be identical, in Fig. 4, but the resolution is poor.

We hope that the issue related to resolution of our images has been properly addressed in the response to point 10) of this letter. In Figure 4A, we show that after silencing of any of the 8 subunits (with the appropriate conditions), in all cases SG biogenesis was impaired, and Sgs3GFP was instead retained in a mesh-like structure. Images obtained after silencing different exocyst subunits are of course not identical, but in all cases, a mesh-like structure has replaced the formation of SGs (Figure 4A). Hopefully, the diagrams now included in Figure 3A and B help the correct interpretation of the phenotypes throughout the study.

To demonstrate that the structure in which Sgs3-GFP was retained upon exocyst complex knockdown corresponds to the ER, we performed a colocalization analysis between Sgs3-GFP and the ER markers GFP-KDEL or Bip-sfGFP-HDEL, after which we calculated the Pearsons Coefficient, which indicated substantial colocalization (Figure 4B-G and Supplementary Figures 7 and 8). These new data are described in lines 196-199 of the revised version of the manuscript. To facilitate the visualization of the results, in the revised version of the manuscript we have included magnified cropped areas of the images shown in Figure 4A.

(15) What is remarkable is a highly variable effect of different subunit KD on the percentage of cells with MLS (Fig. 4C). Controls = 100 %, Exo70=~75% (at 19 deg), Sec3 = ~30%, Sec10 = 0%, Exo84 = 100% ... This is interesting for the functional exocyst is an octameric holocomples, thus why the huge subunit variability in the phenotypes? The trivial explanation is either: i) variable exocyst subunit KD (not shown) or ii) variability between experiments (no error bars are shown). Both should be addressed by quantification of the KD of different proteins and secondly by replicating the experiments.

We agree with the reviewer statement. We believe that both, variability of KD efficiency (i) and variability between experiments (ii) contribute to the variable effect observed after knocking down the different subunits. As detailed in the response to point 6), we have performed qRT-PCR determinations to confirm that the severity of the phenotype depends on the efficiency of RNAimediated silencing. We chose to analyse in detail the effect on the subunits exo70 and sec3, which were those with the highest phenotypic differences between the three silencing temperatures utilized. We found that as expected, the levels of silencing were temperaturedependent, being higher at 29°C and lower at 19°C. These data were included in Supplementary Figure 2, and described lines 153-159 of the Results section and also summarized in Author response images 4 and 5 of this rebuttal letter.

We thank the reviewer for his/her comment on the replication of experiments and statistics. We failed to include detailed numerical information in the original submission, such as the number of replicas and standard deviations of the data depicted in Figure 3C and Supplementary Figure 1, so we apologize for this omission. In the revised version of the manuscript, we have included a table (Supplementary Table 3) in which all the raw data of Figure 3C and Supplementary Figure 1, including standard deviations, are now depicted.

(16) If their data holds up then the underlying mechanism here needs to be considered.

(Note: there is some precedent from the autophagy field of differential exocyst effects)

Our proposed mechanism is essentially that the holocomplex is required for multiple processes along the secretory pathway. Each of these actions (Golgi structure maintenance, SG maturation and SG tethering/fusion with the plasma membrane) requires different amounts of holocomplex activity, being this the reason why each phenotype manifests at different levels of RNAi-mediated silencing (Author response image 4 of this letter). The model predicts that Golgi structure maintenance requires minimal levels of complex activity, and that is why strong knock-down of exocyst subunits is required to obtain this phenotype. In line with our results, it has been reported that other tethering complexes of the CATCHR family are also required for maintaining Golgi cisternae stuck together (D'Souza et al, 2020; Khakurel and Lupashin, 2023; Liu et al, 2019). One possibility is that the exocyst may play a redundant role in the maintenance of the normal structure of the Golgi complex, along with other CATCHR complexes. This potential redundancy could explain why severe exocyst knock-down is required to observe structural anomalies at this organelle. On the other end of the spectrum, we propose that tethering/fusion with the plasma membrane is very susceptible to even slight reduction of complex activity, so that mild RNAi-mediated silencing is sufficient to provoke defects in this process. This proposed model is depicted in Author response image 4 and discussed in lines 395-405 of the Discussion section. 

(17) In the salivary glands the authors state that the exocyst is needed for Sgs3-GFP exit from the ER. First, Pearson's coefficient should be shown so as to quantitate the degree of ER localizations of all KDs.

We thank the reviewer for this comment that helped us to strengthen the observation that when SG biogenesis is impaired, Sgs3-GFP remains trapped in the ER. In the revised version of the manuscript, we have calculated Pearson´s coefficient to assess colocalization between ER markers (GFP-KDEL or Bip-sfGFP-HDEL) and Sgs3-GFP in salivary gland cells that express sec15RNAi. The Pearson’s coefficient was around 0.6 for both ER markers, indicating that colocalization with Sgs3-GFP was substantial (Supplementary Figure 8, text lines 196-199 of the Results section).

(18) Second, there should be some rescue performed (if possible) to support specificity. 

As suggested by the reviewer, we have performed a rescue experiment of the phenotype provoked by the expression of sec15 RNAi, which consisted on the retention of Sgs3-GFP in the endoplasmic reticulum: Expression of Sec15-GFP reverted substantially the ER retention phenotype, rescuing SG biogenesis and also SG maturation in most cells (over 60% of the cells). These new data are now shown in Supplementary Figure 4, and described in lines 168-171 of the Results section.

(19) Third, importantly other proteins that should traffic to the PM need to be shown to traffic normally so as to rule out a non-specific effect.

We have addressed this issue (also mentioned by Reviewer #1), by analyzing the localization of a number of polarization markers, finding that the overall polarization of the cell was not affected by loss of function of exocyst subunits. Please, see our response to the point 3) raised by Reviewer #1. The new data showing cell polarization markers are shown in Supplementary Figure 6 of the revised version of the manuscript, and described on text lines 172-179 of the Results section.

(20) It is unclear from their model (Fig. 5) why after exocyst KD of Sec15 the cis-Golgi is more preserved than the TGN, which appears as large vacuoles. This is not quantitated and not shown for the 8 subunits.

We thank the reviewer for this relevant comment. We agree that the phenotype of either, sec15 or sec3 loss-of-function cells manifests differently with cis-Golgi and trans-Golgi markers. While the cis-Golgi marker looked fragmented and aggregated, the trans-Golgi marker adopted a swollen appearance. However, in our view, the different appearance of the two markers does not necessarily imply that one compartment is more preserved than the other. In the revised version of the manuscript, we have quantified the penetrance of the phenotypes provoked by sec15 or sec3 silencing, using both cis-Golgi and trans-Golgi markers. In both cases, the penetrance was high, although even higher with the trans-Golgi marker. These new data are now depicted in Supplementary Figure 9 of the revised version of the manuscript. 

It is interesting to mention that in HeLa cells, as well as in the retinal epithelial cell line hTERT, Golgi phenotypes similar to those we have described here have been reported after loss-offunction of other tethering complexes, which were shown to maintain the Golgi cisternae stuck together, including the GOC and GARP complexes (D'Souza et al, 2020, Khakurel and Lupashin, 2023; Shijie Liu et al, 2019). As we did throughout our work, not every aspect of the analysis included the silencing of all eight subunits. In this case, we chose to silence Sec3 and Sec15. Please note that we have modified the model depicted in Figure 6E-F, to highlight the cis- and transGolgi phenotypes upon exocyst knock-down, as well as the localization of the exocyst in cisternae of the Golgi complex.

(21) Acute/Chronic control: It would be nice to acutely block the exocyst so as to better distinguish if the effects observed are primary or secondary effects (e.g. on a recycling pathway).

We thank the reviewer for raising this important issue. To address this point, and to be able to induce silencing of exocyst subunits at specific time intervals of larval development, we utilized a strategy based on a thermosensitive variant of the Gal4 inhibitor Gal80 (Gal80ts)(Lee and Luo, 1999). We blocked Gal4 activity (and therefore RNAi expression) by maintaining the larvae at 18 °C during the 1st and 2nd instars (until 120 hours after egg lay), and then induced the activity of Gal4 specifically at the 3rd larval instar by raising the temperature to 29 ºC, a condition in which Gal80ts becomes inactive. After silencing the expression of sec3 or sec15 at the 3rd larval instar only, the phenotype was very similar to that observed after chronic silencing of exocyst subunits (larvae maintained at 29 ºC all throughout development, where Gal4 was never inhibited). These observations suggest that the defects observed in the secretory pathway after knock down of exocyst subunits reflect genuine functions of the exocyst in this pathway, rather than a secondary effect derived from impaired development of the salivary glands at early larval stages. These new results are now shown in Supplementary Figure 3, and described in manuscript lines 160-171 of the Results section.   

(22) Granule homotypic fusion. Strangely over-expression of just one subunit, Sec15-GFP, made giant secretory granules (SG) that were over 8 microns big! Why is that, especially if normally the exocyst is normally a holocomplex. Was this an effect that was specific to Sec15 or all exocyst subunits? Is the Sec15 level rate limiting in these cells? It may be that a subcomplex of Sec15/10 plays earlier roles, but in any case this needs to be addressed across all (or many) of the exocyst subcomplex members.

Please, see our response to point 7) of this letter. Sec15 is believed to act as a seed for the formation of the whole complex.

(23) In summary, there are clearly striking effects on secretory granule biogenesis by dysfunction of the exocyst, however right now it is hard to disentangle effects on ERGolgi traffic, loss of the TGN, and a problem in maturation or fusion of granules. 

As discussed in detail in our response to the point 3 raised by Reviewer #1, the secretory pathway is highly synchronized in each of the cells of the Drosophila salivary gland. SG biogenesis, SG maturation and SG fusion with the plasma membrane never occur simultaneously in the same cell. Thus, in a cell in which ER-Golgi traffic is impaired (and SG biogenesis does not occur), SGs do not exist, and therefore, they cannot exhibit defects in the process of maturation or fusion with the plasma membrane. In summary, we believe that our work has shown that in Drosophila larval salivary glands the exocyst holocomplex is required for (at least) three functions along the secretory pathway: 1) To maintain the appropriate Golgi complex architecture, thus enabling ERGolgi transport; 2) For secretory granule maturation: both, homotypic fusion and acquisition of maturation factors; 3) For secretory granule exocytosis: secretory granule tethering to enable subsequent fusion with the plasma membrane. As mentioned above (point 6 of this letter), these three functions require different amounts of the holocomplex, and therefore can be revealed by inducing different levels of silencing.  

(24) It is also confusing if the entire exocyst holocomplex or subcomplex plays a key role 

The fact that, by silencing any of the subunits (with the appropriate conditions) it is possible obtain any of the 3 phenotypes (impaired SG biogenesis, impaired SG maturation or impaired SG fusion with the plasma membrane) argues in favour of a function of the complex as a whole in each of these three functions.

Reviewer 3:

(25) General comment: Freire and co-authors examine the role of the exocyst complex during the formation and secretion of mucins from secretory granules in the larval salivary gland of Drosophila melanogaster. Using transgenic lines with a tagged Sgs3 mucin the authors KD expression of exocyst subunit members and observe a defect in secretory granules with a heterogeneity of phenotypes. By carefully controlling RNAi expression using a Gal4-based system the authors can KD exocyst subunit expression to varying degrees. The authors find that the stronger the inhibition of expression of exocyst the earlier in the secretory pathway the defect. The manuscript is well written, the model system is physiological, and the techniques are innovative.

We appreciate the reviewer´s assessment of our work. 

(26) My major concern is that the evidence underlying the fundamental claim of the manuscript that "the exocyst complex participates" in multiple secretory processes lacks direct evidence.

We thank the reviewer for raising this important issue. We believe that the analysis of Sec15 subcellular localization during salivary gland development (Figures 5, 7B-D and 9E-F), in combination with the detailed analysis of the phenotypes provoked by loss-of-function of each of the exocyst subunits, provide evidence supporting multiple functions of the exocyst in the secretory pathway. We have also included 3D reconstructions and videos of GFP-Sec15 colocalization with Golgi and SG markers to support exocyst localization associated to these structures (Supplementary Videos 1-7), text lines 200-210; 216-221 and 303-305.

(27) It is clear from multiple lines of evidence, which are discussed by the authors, that exocyst is essential for an array of exocytic events. The fundamental concern is that loss of homeostasis on the plasma membrane proteome and lipidome might have severe pleiotropic effects on the cell.

We agree with the reviewer that this is an important point that needed to be addressed. As discussed in detail above at the response to point 3 raised by Reviewer #1, we have analysed several plasma membrane markers (including a PI(4,5)P2 lipid reporter), and found that overall, plasma membrane integrity and polarity were not substantially affected (Supplementary Figure 6). In addition, we have analyzed several markers of general cellular “health” that indicate that salivary gland cells do not seem to be distressed by the reduction of exocyst complex activity (Supplementary Figure 5). These new data are described in lines 172-179 of the Results section.

(28) Perhaps the authors have more evidence that exocyst is important for homeotypic fusion of the SGs, as supported by the localisation of Sec15 on the fusion sites.

We believe that the fact that, by silencing any of the exocyst subunits (with the appropriate conditions), immature smaller-than-normal granules were observed, argus in favour that the exocyst as a whole participates in SG homofusion (Figure 7A). In addition, we have included more images, quantifications, 3D reconstructions and videos of GFP-Sec15 localized just at the contact sites between immature SGs. We have quantified and compared GFP-Sec15 localization at immature SG vs its localization at mature SGs, finding that localizes preferentially at immature SGs, supporting a role of the exocyst as a tethering complex during homotypic fusion (shown Figure 7B-C and Supplementary Videos 4-6, and described in lines 216-221 of the Results section). Please see also our response to the point 2 raised by reviewer 1 in this rebuttal letter, and to Author response image 3 above in this letter.

(29) The second question that I think is important to address is, what exactly do the varying RNAi levels correspond to in terms of experiments, and have these been validated? Due to the fundamental claim being that the severity of the phenotype being correlated with the level of KD, I think validation of this model is absolutely essential.  

We thank the Reviewer for raising this important point, and agree it was lacking in the original version of our manuscript. As discussed in our response to the point 6) raised by Reviewer #2, we have performed qRT-PCR determinations for exo70 and sec3 mRNA levels after inducing silencing of these subunits at different temperatures, or with different RNAi transgenic lines. The remnant mRNA levels correlate well with the observed phenotypes. Please see Supplementary Figure 2 of the revised manuscript, and Author response image 5 of this rebuttal letter; described in lines 155-159 of the Results section. 

Recommendations for the authors:

Reviewer #1 (Recommendations For The Authors):

-  The authors assert in the discussion that exocyst involvement in constitutive secretion is well documented. This is based on a very recent study in mammalian culture cells. Therefore, I would not dismiss the issue as completely settled. Furthermore, a previous study of Drosophila sec10 reported no roles outside the ring gland (DOI: 10.1034/j.1600-0854.2002.31206.x).

We have included these observations in the Discussion section. Lines 326-329.

-  A salivary gland screening by Julie Brill's lab reported exocyst components as hits (DOI: 10.1083/jcb.201808017).

We have referred to this paper in the Discussion section. Lines 326-329.

-  It should be explained in more detail what is measured in graphs 7C, F, and others quantifying fluorescence around secretory granules. Looking at the images, the decrease in Rab1 and Rab11 seems less convincing.

We have made a clearer description of how fluorescence intensity was measured in the Methods section lines 558-561. Also, we have uploaded a source data file in which the raw data of each experiment used for quantifications are disclosed. 

Please note that the data indicates that Rab11 levels are higher in sec5 (Figure 8J-L) and sec3 (supplementary Figure 11M-R).

Reviewer #2 (Recommendations For The Authors):

No major issues.

Writing - The authors should better frame their interpretations of other studies of the exocyst that include the role in autophagy, Palade body trafficking, and differential roles of the subunits.

We have discussed these specific points in the Discussion section, lines 348-355 and 409-410.

Minor - Fig. 6A: Why are variable temperatures (19-29 deg C used for the 8 KD experiments)?

Please show it all at the same temperature (control too).

The need for the usage of specific temperatures to obtain specific phenotypes with each of the RNAi lines used was explained in point 6 of this letter.

Reviewer #3 (Recommendations For The Authors):

In the abstract, the authors refer to the exocytic process and go on to describe secretory granule biogenesis and exocytosis. However, there are many exocytic processes aside from secretory granule biogenesis, and I think the authors should clarify this.

Corrected in the Abstract. Lines 19-21

Page 17 Thomas, 2021 reference, there is a glitch with the reference.

Thanks for noticing. Fixed.

References

Bhuin T, Roy JK. Developmental expression, co-localization and genetic interaction of exocyst component Sec15 with Rab11 during Drosophila development. Exp Cell Res. 2019 Aug 1;381(1):94-104. doi: 10.1016/j.yexcr.2019.04.038. Epub 2019 May 7. PMID: 31071318.

D'Souza Z, Taher FS, Lupashin VV. Golgi inCOGnito: From vesicle tethering to human disease. Biochim Biophys Acta Gen Subj. 2020 Nov;1864(11):129694. doi: 10.1016/j.bbagen.2020.129694. Epub 2020 Jul 27. PMID: 32730773; PMCID: PMC7384418.

Escrevente C, Bento-Lopes L, Ramalho JS, Barral DC. Rab11 is required for lysosome exocytosis through the interaction with Rab3a, Sec15 and GRAB. J Cell Sci. 2021 Jun 1;134(11):jcs246694. doi: 10.1242/jcs.246694. Epub 2021 Jun 8. PMID: 34100549; PMCID: PMC8214760.

Guo W, Roth D, Walch-Solimena C, Novick P. The exocyst is an effector for Sec4p, targeting secretory vesicles to sites of exocytosis. EMBO J. 1999 Feb 15;18(4):1071-80. doi: 10.1093/emboj/18.4.1071. PMID: 10022848; PMCID: PMC1171198.

Jafar-Nejad H, Andrews HK, Acar M, Bayat V, Wirtz-Peitz F, Mehta SQ, Knoblich JA, Bellen HJ. Sec15, a component of the exocyst, promotes notch signaling during the asymmetric division of Drosophila sensory organ precursors. Dev Cell. 2005 Sep;9(3):351-63. doi: 10.1016/j.devcel.2005.06.010. PMID: 16137928.

Khakurel A, Lupashin VV. Role of GARP Vesicle Tethering Complex in Golgi Physiology. Int J Mol Sci. 2023 Mar 23;24(7):6069. doi: 10.3390/ijms24076069. PMID: 37047041; PMCID: PMC10094427.

Lattner J, Leng W, Knust E, Brankatschk M, Flores-Benitez D. Crumbs organizes the transport machinery by regulating apical levels of PI(4,5)P2 in Drosophila. Elife. 2019 Nov 7;8:e50900. doi: 10.7554/eLife.50900. PMID: 31697234; PMCID: PMC6881148.

Lee T, Luo L. Mosaic analysis with a repressible cell marker for studies of gene function in neuronal morphogenesis. Neuron. 1999 Mar;22(3):451-61. doi: 10.1016/s08966273(00)80701-1. PMID: 10197526.

Liu S, Majeed W, Grigaitis P, Betts MJ, Climer LK, Starkuviene V, Storrie B. Epistatic Analysis of the Contribution of Rabs and Kifs to CATCHR Family Dependent Golgi Organization. Front Cell Dev Biol. 2019 Aug 2;7:126. doi: 10.3389/fcell.2019.00126. PMID: 31428608; PMCID: PMC6687757.

Perkins LA, Holderbaum L, Tao R, Hu Y, Sopko R, McCall K, Yang-Zhou D, Flockhart I, Binari R, Shim HS, Miller A, Housden A, Foos M, Randkelv S, Kelley C, Namgyal P, Villalta C, Liu LP, Jiang X, Huan-Huan Q, Wang X, Fujiyama A, Toyoda A, Ayers K, Blum A, Czech B, Neumuller R, Yan D, Cavallaro A, Hibbard K, Hall D, Cooley L, Hannon GJ, Lehmann R, Parks A, Mohr SE, Ueda R, Kondo S, Ni JQ, Perrimon N. The Transgenic RNAi Project at Harvard Medical School: Resources and Validation. Genetics. 2015 Nov;201(3):843-52. doi: 10.1534/genetics.115.180208. Epub 2015 Aug 28. PMID: 26320097; PMCID: PMC4649654.

Wu S, Mehta SQ, Pichaud F, Bellen HJ, Quiocho FA. Sec15 interacts with Rab11 via a novel domain and affects Rab11 localization in vivo. Nat Struct Mol Biol. 2005 Oct;12(10):879-85. doi: 10.1038/nsmb987. Epub 2005 Sep 11. PMID: 16155582.

Yeaman C, Grindstaff KK, Wright JR, Nelson WJ. Sec6/8 complexes on trans-Golgi network and plasma membrane regulate late stages of exocytosis in mammalian cells. J Cell Biol. 2001 Nov 12;155(4):593-604. doi: 10.1083/jcb.200107088. Epub 2001 Nov 5. PMID: 11696560; PMCID: PMC2198873.

Zhang XM, Ellis S, Sriratana A, Mitchell CA, Rowe T. Sec15 is an effector for the Rab11 GTPase in mammalian cells. J Biol Chem. 2004 Oct 8;279(41):43027-34. doi: 10.1074/jbc.M402264200. Epub 2004 Jul 29. PMID: 15292201.

Author response:

The following is the authors’ response to the original reviews.

We summarized the main changes:

(1) In the Introduction part, we give a general definition of habitat fragmentation to avoid confusion, as reviewers #1 and #2 suggested.

(2) We clarify the two aspects of the observed “extinction”——“true dieback” and “emigration”, as reviewers #2 and #3 suggested.

(3) In the Methods part, we 1) clarify the reason for testing the temporal trend in colonization/extinction dynamics and describe how to select islands as reviewer #1 suggested; 2) describe how to exclude birds from the analysis as reviewer #2 suggested.

(4) In the Results part, we modified and rearranged Figure 4-6 as reviewers #1, #2 and #3 suggested.

(5) In the Discussion part, we 1) discuss the multiple aspects of the metric of isolation for future research as reviewer #3 suggested; 2) provide concrete evidence about the relationship between habitat diversity or heterogeneity and island area and 3) provide a wider perspective about how our results can inform conservation practices in fragmented habitats as reviewer #2 suggested.

eLife Assessment

This important study enhances our understanding of how habitat fragmentation and climate change jointly influence bird community thermophilization in a fragmented island system. The evidence supporting some conclusions is incomplete, as while the overall trends are convincing, some methodological aspects, particularly the isolation metrics and interpretation of colonization/extinction rates, require further clarification. This work will be of broad interest to ecologists and conservation biologists, providing crucial insights into how ecosystems and communities react to climate change.

We sincerely extend our gratitude to you and the esteemed reviewers for acknowledging the importance of our study and for raising these concerns. We have clarified the rationale behind our analysis of temporal trends in colonization and extinction dynamics, as well as the choice of distance to the mainland as the isolation metric. Additionally, we further discuss the multiple aspects of the metric of isolation for future research and provide concrete supporting evidence about the relationship between habitat diversity or heterogeneity and island area.

Incorporating these valuable suggestions, we have thoroughly revised our manuscript, ensuring that it now presents a more comprehensive and nuanced account of our research. We are confident that these improvements will further enhance the impact and relevance of our work for ecologists and conservation biologists alike, offering vital insights into the resilience and adaptation strategies of communities facing the challenges of climate change.

Reviewer #1 (Public Review):

Summary:

This study reports on the thermophilization of bird communities in a network of islands with varying areas and isolation in China. Using data from 10 years of transect surveys, the authors show that warm-adapted species tend to gradually replace cold-adapted species, both in terms of abundance and occurrence. The observed trends in colonisations and extinctions are related to the respective area and isolation of islands, showing an effect of fragmentation on the process of thermophilization.

Strengths:

Although thermophilization of bird communities has been already reported in different contexts, it is rare that this process can be related to habitat fragmentation, despite the fact that it has been hypothesized for a long time that it could play an important role. This is made possible thanks to a really nice study system in which the construction of a dam has created this incredible Thousand Islands lake. Here, authors do not simply take observed presence-absence as granted and instead develop an ambitious hierarchical dynamic multi-species occupancy model. Moreover, they carefully interpret their results in light of their knowledge of the ecology of the species involved.

Response: We greatly appreciate your recognition of our study system and the comprehensive approach and careful interpretation of results. 

Weaknesses:

Despite the clarity of this paper on many aspects, I see a strong weakness in the authors' hypotheses, which obscures the interpretation of their results. Looking at Figure 1, and in many sentences of the text, a strong baseline hypothesis is that thermophilization occurs because of an increasing colonisation rate of warm-adapted species and extinction rate of cold-adapted species. However, there does not need to be a temporal trend! Any warm-adapted species that colonizes a site has a positive net effect on CTI; similarly, any cold-adapted species that goes extinct contributes to thermophilization.

Thank you very much for these thoughtful comments. The understanding depends on the time frame of the study and specifically, whether the system is at equilibrium. We think your claim is based on this background: if the system is not at equilibrium, then CTI can shift simply by having differential colonization (or extinction) rates for warm-adapted versus cold-adapted species. We agree with you in this case.

On the other hand, if a community is at equilibrium, then there will be no net change in CTI over time. Imagine we have an archipelago where the average colonization of warm-adapted species is larger than the average colonization of cold-adapted species, then over time the archipelago will reach an equilibrium with stable colonization/extinction dynamics where the average CTI is stable over time. Once it is stable, then if there is a temporal trend in colonization rates, the CTI will change until a new equilibrium is reached (if it is reached).

For our system, the question then is whether we can assume that the system is or has ever been at equilibrium. If it is not at equilibrium, then CTI can shift simply by having differential colonization (or extinction) rates for warm-adapted versus cold-adapted species. If the system is at equilibrium (at the beginning of the study), then CTI will only shift if there is a temporal change or trend in colonization or extinction rates.

Habitat fragmentation can affect biomes for decades after dam formation. The “Relaxation effect” (Gonzalez, 2000) refers to the fact that the continent acts as a potential species pool for island communities. Under relaxation, some species will be filtered out over time, mainly through the selective extinction of species that are highly sensitive to fragmentation. Meanwhile, for a 100-hectare patch, it takes about ten years to lose 50% of bird species; The smaller the patch area, the shorter the time required (Ferraz et al., 2003; Haddad et al., 2015). This study was conducted 50 to 60 years after the formation of the TIL, making the system with a high probability of reaching “equilibrium” through “Relaxation effect”(Si et al., 2014). We have no way of knowing exactly whether “equilibrium” is true in our system. Thus, changing rates of colonization-extinction over time is actually a much stronger test of thermophilization, which makes our inference more robust.

We add a note to the legend of Figure 1 on Lines 781-786:

“CTI can also change simply due to differential colonization-extinction rates by thermal affinity if the system is not at equilibrium prior to the study. In our study system, we have no way of knowing whether our island system was at equilibrium at onset of the study, thus, focusing on changing rates of colonization-extinction over time presents a much stronger tests of thermophilization.”

We hope this statement can make it clear. Thank you again for this meaningful question.

Another potential weakness is that fragmentation is not clearly defined. Generally, fragmentation sensu lato involves both loss of habitat area and changes in the spatial structure of habitats (i.e. fragmentation per se). Here, both area and isolation are considered, which may be slightly confusing for the readers if not properly defined.

Thank you for reminding us of that. Habitat fragmentation in this study involves both habitat loss and fragmentation per se. We have clarified the general definition in the Introduction on Lines 61-63:

“Habitat fragmentation, usually defined as the shifts of continuous habitat into spatially isolated and small patches (Fahrig, 2003), in particular, has been hypothesized to have interactive effects with climate change on community dynamics.”

Reviewer #2 (Public Review):

Summary:

This study addresses whether bird community reassembly in time is related to climate change by modelling a widely used metric, the community temperature index (CTI). The authors first computed the temperature index of 60 breeding bird species thanks to distribution atlases and climatic maps, thus obtaining a measure of the species realized thermal niche.

These indices were aggregated at the community level, using 53 survey transects of 36 islands (repeated for 10 years) of the Thousand Islands Lake, eastern China. Any increment of this CTI (i.e. thermophilization) can thus be interpreted as a community reassembly caused by a change in climate conditions (given no confounding correlations).

The authors show thanks to a mix of Bayesian and frequentist mixed effect models to study an increment of CTI at the island level, driven by both extinction (or emigration) of cold-adapted species and colonization of newly adapted warm-adapted species. Less isolated islands displayed higher colonization and extinction rates, confirming that dispersal constraints (created by habitat fragmentation per se) on colonization and emigration are the main determinants of thermophilization. The authors also had the opportunity to test for habitat amount (here island size). They show that the lack of microclimatic buffering resulting from less forest amount (a claim backed by understory temperature data) exacerbated the rates of cold-adapted species extinction while fostering the establishment of warm-adapted species.

Overall these findings are important to range studies as they reveal the local change in affinity to the climate of species comprising communities while showing that the habitat fragmentation VS amount distinction is relevant when studying thermophilization. As is, the manuscript lacks a wider perspective about how these results can be fed into conservation biology, but would greatly benefit from it. Indeed, this study shows that in a fragmented reserve context, habitat amount is very important in explaining trends of loss of cold-adapted species, hinting that it may be strategic to prioritize large habitats to conserve such species. Areas of diverse size may act as stepping stones for species shifting range due to climate change, with small islands fostering the establishment of newly adapted warm-adapted species while large islands act as refugia for cold-adapted species. This study also shows that the removal of dispersal constraints with low isolation may help species relocate to the best suitable microclimate in a heterogenous reserve context.

Thank you very much for your valuable feedback. We greatly appreciate your recognition of the scientific question to the extensive dataset and diverse approach. In particular, you provided constructive suggestions and examples on how to extend the results to conservation guidance. This is something we can’t ignore in the manuscript. We have added a paragraph to the end of the Discussion, stating how our results can inform conservation, on Lines 339-347:

‘Overall, our findings have important implications for conservation practices. Firstly, we confirmed the role of isolation in limiting range shifting. Better connected landscapes should be developed to remove dispersal constraints and facilitate species’ relocation to the best suitable microclimate. Second, small patches can foster the establishment of newly adapted warm-adapted species while large patches can act as refugia for cold-adapted species. Therefore, preserving patches of diverse sizes can act as stepping stones or shelters in a warming climate depending on the thermal affinity of species. These insights are important supplement to the previous emphasis on the role of habitat diversity in fostering (Richard et al., 2021) or reducing (Gaüzère et al., 2017) community-level climate debt.’

Strength:

The strength of the study lies in its impressive dataset of bird resurveys, that cover 10 years of continued warming (as evidenced by weather data), 60 species in 36 islands of varying size and isolation, perfect for disentangling habitat fragmentation and habitat amount effects on communities. This distinction allows us to test very different processes mediating thermophilization; island area, linked to microclimatic buffering, explained rates for a variety of species. Dispersal constraints due to fragmentation were harder to detect but confirms that fragmentation does slow down thermophilization processes.

This study is a very good example of how the expected range shift at the biome scale of the species materializes in small fragmented regions. Specifically, the regional dynamics the authors show are analogous to what processes are expected at the trailing and colonizing edge of a shifting range: warmer and more connected places display the fastest turnover rates of community reassembly. The authors also successfully estimated extinction and colonization rates, allowing a more mechanistic understanding of CTI increment, being the product of two processes.

The authors showed that regional diversity and CTI computed only by occurrences do not respond in 10 years of warming, but that finer metrics (abundance-based, or individual islands considered) do respond. This highlights the need to consider a variety of case-specific metrics to address local or regional trends. Figure Appendix 2 is a much-appreciated visualization of the effect of different data sources on Species thermal Index (STI) calculation.

The methods are long and diverse, but they are documented enough so that an experienced user with the use of the provided R script can follow and reproduce them.

Thank you very much for your profound Public Review. We greatly appreciate your recognition of the scientific question, the extensive dataset and the diverse approach. 

Weaknesses:

While the overall message of the paper is supported by data, the claims are not uniformly backed by the analysis. The trends of island-specific thermophilization are very credible (Figure 3), however, the variable nature of bird observations (partly compensated by an impressive number of resurveys) propagate a lot of errors in the estimation of species-specific trends in occupancy, abundance change, and the extinction and colonization rates. This materializes into a weak relationship between STI and their respective occupancy and abundance change trends (Figure 4a, Figure 5, respectively), showing that species do not uniformly contribute to the trend observed in Figure 3. This is further shown by the results presented in Figure 6, which present in my opinion the topical finding of the study. While a lot of species rates response to island areas are significant, the isolation effect on colonization and extinction rates can only be interpreted as a trend as only a few species have a significant effect. The actual effect on the occupancy change rates of species is hard to grasp, and this trend has a potentially low magnitude (see below).

Thank you very much for pointing out this shortcoming. The R2 between STI and their respective occupancy trends is relatively small (R2\=0.035). But the R2 between STI and their respective abundance change trends are relatively bigger, in the context of Ecology research (R2\=0.123). The R2 between STI and their respective colonization rate (R2\=0.083) and extinction rate trends (R2\=0.053) are also relatively small. Low R2 indicates that we can’t make predictions using the current model, we must notice that except STI, other factors may influence the species-specific occupancy trend. Nonetheless, it is important to notice that the standardized coefficient estimates are not minor and the trend is also significant, indicating the species-specific response is as least related to STI.

The number of species that have significant interaction terms for isolation (Figure 6) is indeed low. Although there is uncertainty in the estimation of relationships, there are also consistent trends in response to habitat fragmentation of colonization of warm-adapted species and extinction of cold-adapted species. This is especially true for the effect of isolation, where on islands nearer to the mainland, warm-adapted species (15 out of 15 investigated species) increased their colonization probability at a higher rate over time, while most cold-adapted species (21 out of 23 species) increased their extinction probability at a higher rate. We now better highlight these results in the Results and Discussion.

While being well documented, the myriad of statistical methods used by the authors ampere the interpretation of the figure as the posterior mean presented in Figure 4b and Figure 6 needs to be transformed again by a logit-1 and fed into the equation of the respective model to make sense of. I suggest a rewording of the caption to limit its dependence on the method section for interpretation.

Thank you for this suggestion. The value on the Y axis indicates the posterior mean of each variable (year, area, isolation and their interaction effects) extracted from the MSOM model, where the logit(extinction rate) or logit(colonization rate) was the response variable. All variables were standardized before analysis to make them comparable so interpretation is actually quite straight forward: positive values indicate positive influence while negative values indicate negative influence. Because the goal of Figure 6 is to display the negative/positive effect, we didn’t back-transform them. Following your advice, we thus modified the caption of Figure 6 (now renumbered as Figure 5, following a comment from Reviewer #3, to move Figure 5 to Figure 4c). The modified title and legends of Figure 5 are on Lines 817-820:

“Figure 5. Posterior estimates of logit-scale parameters related to cold-adapted species’ extinction rates and warm-adapted species’ colonization rates. Points are species-specific posterior means on the logit-scale, where parameters >0 indicate positive effects (on extinction [a] or colonization [b]) and parameters <0 indicate negative effects...”

By using a broad estimate of the realized thermal niche, a common weakness of thermophilization studies is the inability to capture local adaptation in species' physiological or behavioral response to a rise in temperature. The authors however acknowledge this limitation and provide specific examples of how species ought to evade high temperatures in this study region.

We appreciate your recognition. This is a common problem in STI studies. We hope in future studies, researchers can take more details about microclimate of species’ true habitat across regions into consideration when calculating STI. Although challenging, focusing on a smaller portion of its distribution range may facilitate achievement.

Reviewer #3 (Public Review):

Summary:

Juan Liu et al. investigated the interplay between habitat fragmentation and climate-driven thermophilization in birds in an island system in China. They used extensive bird monitoring data (9 surveys per year per island) across 36 islands of varying size and isolation from the mainland covering 10 years. The authors use extensive modeling frameworks to test a general increase in the occurrence and abundance of warm-dwelling species and vice versa for cold-dwelling species using the widely used Community Temperature Index (CTI), as well as the relationship between island fragmentation in terms of island area and isolation from the mainland on extinction and colonization rates of cold- and warm-adapted species. They found that indeed there was thermophilization happening during the last 10 years, which was more pronounced for the CTI based on abundances and less clearly for the occurrence-based metric. Generally, the authors show that this is driven by an increased colonization rate of warm-dwelling and an increased extinction rate of cold-dwelling species. Interestingly, they unravel some of the mechanisms behind this dynamic by showing that warm-adapted species increased while cold-dwelling decreased more strongly on smaller islands, which is - according to the authors - due to lowered thermal buffering on smaller islands (which was supported by air temperature monitoring done during the study period on small and large islands). They argue, that the increased extinction rate of cold-adapted species could also be due to lowered habitat heterogeneity on smaller islands. With regards to island isolation, they show that also both thermophilization processes (increase of warm and decrease of cold-adapted species) were stronger on islands closer to the mainland, due to closer sources to species populations of either group on the mainland as compared to limited dispersal (i.e. range shift potential) in more isolated islands.

The conclusions drawn in this study are sound, and mostly well supported by the results. Only a few aspects leave open questions and could quite likely be further supported by the authors themselves thanks to their apparent extensive understanding of the study system.

Strengths:

The study questions and hypotheses are very well aligned with the methods used, ranging from field surveys to extensive modeling frameworks, as well as with the conclusions drawn from the results. The study addresses a complex question on the interplay between habitat fragmentation and climate-driven thermophilization which can naturally be affected by a multitude of additional factors than the ones included here. Nevertheless, the authors use a well-balanced method of simplifying this to the most important factors in question (CTI change, extinction, and colonization, together with habitat fragmentation metrics of isolation and island area). The interpretation of the results presents interesting mechanisms without being too bold on their findings and by providing important links to the existing literature as well as to additional data and analyses presented in the appendix.

We appreciate very much for your positive and constructive comments and suggestions. Thank you for your recognition of the scientific question, the modeling approach and the conclusions. 

Weaknesses:

The metric of island isolation based on the distance to the mainland seems a bit too oversimplified as in real life the study system rather represents an island network where the islands of different sizes are in varying distances to each other, such that smaller islands can potentially draw from the species pools from near-by larger islands too - rather than just from the mainland. Thus a more holistic network metric of isolation could have been applied or at least discussed for future research. The fact, that the authors did find a signal of island isolation does support their method, but the variation in responses to this metric could hint at a more complex pattern going on in real-life than was assumed for this study.

Thank you for this meaningful question. Isolation can be measured in different ways in the study region. We chose the distance to the mainland as a measure of isolation based on the results of a previous study. One study in our system provided evidence that the colonization rate and extinction rate of breeding bird species were best fitted using distance to the nearest mainland over other distance-based measures (distance to the nearest landmass, distance to the nearest bigger landmass)(Si et al., 2014). Besides, their results produced almost identical patterns of the relationship between isolation and colonization/extinction rate (Si et al., 2014). That’s why we only selected “Distance to the mainland” in our current analysis and we do find some consistent patterns as expected. The plants on all islands were cleared out about 60 years ago due to dam construction, with all bird species coming from the mainland as the original species pool through a process called “relaxation”. This could be the reason why distance to the nearest mainland is the best predictor.

We agree with you that it’s still necessary to consider more aspects of “isolation” at least in discussion for future research. In our Discussion, we address these on Lines 292-299:

“As a caveat, we only consider the distance to the nearest mainland as a measure of fragmentation, consistent with previous work in this system (Si et al., 2014), but we acknowledge that other distance-based metrics of isolation that incorporate inter-island connections could reveal additional insights on fragmentation effects. The spatial arrangement of islands, like the arrangement of habitat, can influence niche tracking of species (Fourcade et al., 2021). Future studies should take these metrics into account to thoroughly understand the influence of isolation and spatial arrangement of patches in mediating the effect of climate warming on species.”

Further, the link between larger areas and higher habitat diversity or heterogeneity could be presented by providing evidence for this relationship. The authors do make a reference to a paper done in the same study system, but a more thorough presentation of it would strengthen this assumption further.

Thank you very much for this question. We now add more details about the relationship between habitat diversity and heterogeneity based on a related study in the same system. The observed number of species significantly increased with increasing island area (slope = 4.42, R2 = 0.70, p < .001), as did the rarefied species richness per island (slope = 1.03, R2 = 0.43, p < .001), species density (slope = 0.80, R2 = 0.33, p = .001) and the rarefied species richness per unit area (slope = 0.321, R2 = 0.32, p = .001). We added this supporting evidence on Lines 317-321:

“We thus suppose that habitat heterogeneity could also mitigate the loss of these relatively cold-adapted species as expected. Habitat diversity, including the observed number of species, the rarefied species richness per island, species density and the rarefied species richness per unit area, all increased significantly with island area instead of isolation in our system (Liu et al., 2020)”

Despite the general clear patterns found in the paper, there were some idiosyncratic responses. Those could be due to a multitude of factors which could be discussed a bit better to inform future research using a similar study design.

Thank you for these suggestions. We added a summary statement about the reasons for idiosyncratic responses on Lines 334-338:

“Overall, these idiosyncratic responses reveal several possible mechanisms in regulating species' climate responses, including resource demands and biological interactions like competition and predation. Future studies are needed to take these factors into account to understand the complex mechanisms by which habitat loss meditates species range shifts.”

Reviewer #1 (Recommendations For The Authors):

(1) Figure 1: I disagree that there should be a temporal trend in colonisation/extinction dynamics.

Thank you again for these thoughtful comments. We have explained in detail in the response to the Public Review.

(2) L 485-487: As explained before I disagree. I don't see why there needs to be a temporal trend in colonization and extinction.

Thank you again for these thoughtful comments. Because we can’t guarantee that the study system has reached equilibrium, changing rates of colonization-extinction over time is actually a much stronger test of thermophilization. More detailed statement can be seen in the response to the Public Review.

(3) L 141: which species' ecological traits?

Sorry for the confusion. The traits included continuous variables (dispersal ability, body size, body mass and clutch size) and categorical variables (diet, active layer, residence type). Specifically, we tested the correlation between STI and dispersal ability, body size, body mass and clutch size using Pearson correlation test. We also tested the difference in STI between different trait groups using the Wilcoxon signed-rank test for three Category variables: diet (carnivorous/ omnivorous/ herbivory), active layer (canopy/mid/low), and residence type (resident species/summer visitor). There is no significant difference between any two groups for each of the three category variables (p > 0.2). We added these on Lines 141-145:

“No significant correlation was found between STI and species’ ecological traits; specifically, the continuous variables of dispersal ability, body size, body mass and clutch size (Pearson correlations for each, |r| < 0.22), and the categorial variables of diet (carnivorous/omnivorous/herbivory), active layer (canopy/mid/low), and residence type (resident species/summer visitor)”

(4) L 143: CTIoccur and CTIabun were not defined before.

Because CTIoccur and CTIabun were first defined in Methods part (section 4.4), we change the sentence to a more general statement here on Lines 147-150:

“At the landscape scale, considering species detected across the study area, occurrence-based CTI (CTIoccur; see section 4.4) showed no trend (posterior mean temporal trend = 0.414; 95% CrI: -12.751, 13.554) but abundance-based CTI (CTIabun; see section 4.4) showed a significant increasing trend.”

(5) Figure 4: what is the dashed vertical line? I assume the mean STI across species?

Sorry for the unclear description. The vertical dashed line indicates the median value of STI for 60 species, as a separation of warm-adapted species and cold-adapted species. We have added these details on Lines 807-809:

“The dotted vertical line indicates the median of STI values. Cold-adapted species are plotted in blue and warm-adapted species are plotted in orange.”

(6) Figure 6: in the legend, replace 'points in blue' with 'points in blue/orange' or 'solid dots' or something similar.

Thank you for this suggestion. We changed it to “points in blue/orange” on Lines 823.

(7) L 176-176: unclear why the interaction parameters are particularly important for explaining the thermophilization mechanism: if e.g. colonization rate of warm-adapted species is constantly higher in less isolated islands, (and always higher than the extinction rate of the same species), it means that thermophilization is increased in less isolated islands, right?

Thank you for this question. This is also related to the question about “Why use temporal trends in colonization/extinction rate to test for thermophilization mechanisms”. Colonization-extinction over time is actually a much stronger test of thermophilization (more details refer to response to Public Review and Recommendations 1&2).

Based on this, the two main driving processes of thermophilization mechanism include the increasing colonization rate of warm-adapted species and the increasing extinction rate of cold-adapted species with year. The interaction effect between island area (or isolation) and year on colonization rate (or extinction rate) can tell us how habitat fragmentation mediates the year effect. For example, if the interaction term between year and isolation is negative for a warm-adapted species that increased in colonization rate with year, it indicates that the colonization rate increased faster on less isolated islands. This is a signal of a faster thermophilization rate on less-isolated islands.

(8) L201-203: this is only little supported by the results that actually show that there is NO significant interaction for most species.

Thank you for this comment. Although most species showed non-significant interaction effect, the overall trend is relatively consistent, this is especially true for the effect of isolation. To emphasize the “trend” instead of “significant effect”, we slightly modified this sentence in more rigorous wording on Lines 205-208: 

“We further found that habitat fragmentation influences two processes of thermophilization: colonization rates of most warm-adapted species tended to increase faster on smaller and less isolated islands, while the loss rates of most cold-adapted species tended to be exacerbated on less isolated islands.”

(9) Section 2.3: can't you have a population-level estimate? I struggled a bit to understand all the parameters of the MSOM (because of my lack of statistical/mathematical proficiency) so I cannot provide more advice here.

Thank you for raising this advice. We think what you are mentioning is the overall estimate across all species for each variable. From MSOM, we can get a standardized estimate of every variable (year, area, isolation, interaction) for each species, separately. Because the divergent or consistent responses among species are what we are interested in, we didn’t calculate further to get a population-level estimate.

(10) L 291: a dot is missing.

Done. Thank you for your correction.

(11) L 305, 315: a space is missing

Done

(12) L 332: how were these islands selected?

Thank you for this question. The 36 islands were selected according to a gradient of island area and isolation, spreading across the whole lake region. The selected islands guaranteed there is no significant correlation between island area and isolation (the Pearson correlation coefficient r = -0.21, p = 0.21). The biggest 7 islands among the 36 islands are also the only several islands larger than 30 ha in the whole lake region. We have modified this in the Method part on Lines 360-363.

“We selected 36 islands according to a gradient of island area and isolation with a guarantee of no significant correlation between island area and isolation (Pearson r = -0.21, p = 0.21). For each island, we calculated island area and isolation (measured in the nearest Euclidean distance to the mainland) to represent the degree of habitat fragmentation.”

(13) L 334: "Distance to the mainland" was used as a metric of isolation, but elsewhere in the text you argue that the observed thermophilization is due to interisland movements. It sounds contradictory. Why not include the average or shortest distance to the other islands?

Thank you very much for raising this comment. Yes, “Distance to the mainland” was the only metric we used for isolation. We carefully checked through the manuscript where the “interisland movement” comes from and induces the misunderstanding. It must come from Discussion 3.1 (n Lines 217-221): “Notably, when tested on the landscape scale (versus on individual island communities), only the abundance-based thermophilization trend was significant, indicating thermophilization of bird communities was mostly due to inter-island occurrence dynamics, rather than exogenous community turnover.”

Sorry, the word “inter-island” is not exactly what we want to express here, we wanted to express that “the thermophilization was mostly due to occurrence dynamics within the region, rather than exogenous community turnover outside the region”. We have changed the sentence in Discussion part on Lines 217-221:

“Notably, when tested on the landscape scale (versus on individual island communities), only the abundance-based thermophilization trend was significant, indicating thermophilization of bird communities was mostly due to occurrence dynamics within the region, rather than exogenous community turnover outside the region.”

Besides, I would like to explain why we use distance to the mainland. We chose the distance to the mainland as a measure of isolation based on the results of a previous study. One study in our system provided evidence that the colonization rate and extinction rate of breeding bird species were best fitted using distance to the nearest mainland over other distance-based measures (distance to the nearest landmass, distance to the nearest bigger landmass)(Si et al., 2014). Besides, their results produced almost identical patterns of the relationship between isolation and colonization/extinction rate(Si et al., 2014). That’s why we only selected “Distance to the mainland” in our current analysis and we do find some consistent patterns as expected. The plants on all islands were cleared out about 60 years ago due to dam construction, with all bird species coming from the mainland as the original species pool through a process called “relaxation”. This may be the reason why distance to the nearest mainland is the best predictor.

In Discussion part, we added the following discussion and talked about the other measures on Lines 292-299:

“As a caveat, we only consider the distance to the nearest mainland as a measure of fragmentation, consistent with previous work in this system (Si et al., 2014), but we acknowledge that other distance-based metrics of isolation that incorporate inter-island connections could reveal additional insights on fragmentation effects. The spatial arrangement of islands, like the arrangement of habitat, can influence niche tracking of species (Fourcade et al., 2021). Future studies should take these metrics into account to thoroughly understand the influence of isolation and spatial arrangement of patches in mediating the effect of climate warming on species.”

(14) L 347: you write 'relative' abundance but this measure is not relative to anything. Better write something like "we based our abundance estimate on the maximum number of individuals recorded across the nine annual surveys".

Thank you for this suggestion, we have changed the sentence on Lines 377-379:

“We based our abundance estimate on the maximum number of individuals recorded across the nine annual surveys.”

(15) L 378: shouldn't the formula for CTIoccur be (equation in latex format):

CTI{occur, j, t} =\frac{\sum_{i=1}^{N_{j,t}}STI_{i}}{N_{j,t}}

Where Nj,t is the total number of species surveyed in the community j in year t

Thank you very much for this careful check, we have revised it on Lines 415, 417:

“where Nj,t is the total number of species surveyed in the community j in year t.”

Reviewer #2 (Recommendations For The Authors):

(1) Line 76: "weakly"

Done. Thank you for your correction.

(2) Line 98: I suggest a change to this sentence: "For example, habitat fragmentation renders habitats to be too isolated to be colonized, causing sedentary butterflies to lag more behind climate warming in Britain than mobile ones"

Thank you for this modification, we have changed it on Lines 99-101.

(3) Line 101: remove either "higher" or "increasing"

Done, we have removed “higher”. Thank you for this advice.

(4) Line 102: "benefiting from near source of"

Done.

(5) Line 104: "emigrate"

Done.

(6) Introduction: I suggest making it more explicit what process you describe under the word "extinction". At first read, I thought you were only referring to the dieback of individuals, but you also included emigration as an extinction process. It also needs to be reworded in Fig 1 caption.

Thank you for this suggestion. Yes, we can’t distinguish in our system between local extinction and emigration. The observed “extinction” of cold-adapted species over 10 years may involve two processes that usually occur in order: first “emigration” and then if can’t emigrate or withstand, “real local dieback”. It should also be included in the legend of Figure 1, as you said. We have modified the legend in Lines 780-781:

“Note that extinction here may include both the emigration of species and then the local extinction of species.”

There is also one part in the Discussion that mentions this on Lines 287-291: “While we cannot truly distinguish in our system between local extinction and emigration, we suspect that given two islands equal except in isolation, and if both lose suitability due to climate change, individuals can easily emigrate from the island nearer to the mainland, while individuals on the more isolated island would be more likely to be trapped in place until the species went locally extinct due to a lack of rescue”.

(7) I also suggest differentiating habitat fragmentation (distances between islands) and habitat amount (area) as explained in Fahrig 2013 (Rethinking patch size and isolation effects: the habitat amount hypothesis) and her latter paper. This will help the reader what lies behind the general trend of fragmentation: fragmentation per se and habitat amount reduction.

Thank you for this suggestion! Habitat fragmentation in this study involves both habitat loss and fragmentation per se. We now give a general definition of habitat fragmentation on Lines 61-63:

“Habitat fragmentation, usually defined as the shifts of continuous habitat into spatially isolated and small patches (Fahrig, 2003), in particular, has been hypothesized to have interactive effects with climate change on community dynamics.”

(8) Line 136: is the "+-" refers to the standard deviation or confidence interval, I suggest being explicit about it once at the start of the results.

Thank you for reminding this. The "+-" refers to the standard deviation (SD). The modified sentence is now on Lines 135-139:

“The number of species detected in surveys on each island across the study period averaged 13.37 ± 6.26 (mean ± SD) species, ranging from 2 to 40 species, with an observed gamma diversity of 60 species. The STI of all 60 birds averaged 19.94 ± 3.58 ℃ (mean ± SD) and ranged from 9.30 ℃ (Cuculus canorus) to 27.20 ℃ (Prinia inornate), with a median of STI is 20.63 ℃ (Appendix 1—figure 2; Appendix 1—figure 3).”

(9) Line 143: please specify the unit of thermophilization.

The unit of thermophilization rate is the change in degree per unit year. Because in all analyses, predictor variables were z-transformed to make their effect comparable. We have added on Line 151:

“When measuring CTI trends for individual islands (expressed as °/ unit year)”

(10) Line 289: check if no word is missing from the sentence.

The sentence is: “In our study, a large proportion (11 out of 15) of warm-adapted species increasing in colonization rate and half (12 out of 23) of cold-adapted species increasing in extinction rate were changing more rapidly on smaller islands.”

Given that we have defined the species that were included in testing the third prediction in both Methods part and Result part: 15 warm-adapted species that increased in colonization rate and 23 cold-adapted species that increased in extinction rate. We now remove this redundant information and rewrote the sentence as below on Lines 300-302:

“In our study, the colonization rate of a large proportion of warm-adapted species (11 out of 15) and the extinction rate of half of old-adapted species (12 out of 23) were increasing more rapidly on smaller islands.”

(11) Line 319: I really miss a concluding statement of your discussion, your results are truly interesting and deserve to be summarized in two or three sentences, and maybe a perspective about how it can inform conservation practices in fragmented settings.

Thank you for this profound suggestion both in Public Review and here. We have added a paragraph to the end of the Discussion, stating how our results can inform conservation, on Lines 339-347:

“Overall, our findings have important implications for conservation practices. Firstly, we confirmed the role of isolation in limiting range shifting. Better connected landscapes should be developed to remove dispersal constraints and facilitate species’ relocation to the best suitable microclimate. Second, small patches can foster the establishment of newly adapted warm-adapted species while large patches can act as refugia for cold-adapted species. Therefore, preserving patches of diverse sizes can act as stepping stones or shelters in a warming climate depending on the thermal affinity of species. These insights are important supplement to the previous emphasis on the role of habitat diversity in fostering (Richard et al., 2021) or reducing (Gaüzère et al., 2017) community-level climate debt.”

(12) Line 335: I suggest " ... the islands has been protected by forbidding logging, ..."

Thanks for this wonderful suggestion. Done. The new sentence is now on Lines 365-366:

“Since lake formation, the islands have been protected by forbidding logging, allowing natural succession pathways to occur.”

(13) Line 345: this speed is unusually high for walking, check the speed.

Sorry for the carelessness, it should be 2.0 km/h. It has been corrected on Lines 375-376:

“In each survey, observers walked along each transect at a constant speed (2.0 km/h) and recorded all the birds seen or heard on the survey islands.”

(14) Line 351: you could add a sentence explaining why that choice of species exclusion was made. Was made from the start of the monitoring program or did you exclude species afterward?

We excluded them afterward. We excluded non-breeding species, nocturnal and crepuscular species, high-flying species passing over the islands (e.g., raptors, swallows) and strongly water-associated birds (e.g., cormorants). These records were recorded during monitoring, including some of them being on the shore of the island or high-flying above the island, and some nocturnal species were just spotted by accident.

We described more details about how to exclude species on Lines 379-387:

“We excluded non-breeding species, nocturnal and crepuscular species, high-flying species passing over the islands (e.g., raptors, swallows) and strongly water-associated birds (e.g., cormorants) from our record. First, our surveys were conducted during the day, so some nocturnal and crepuscular species, such as the owls and nightjars were excluded for inadequate survey design. Second, wagtail, kingfisher, and water birds such as ducks and herons were excluded because we were only interested in forest birds. Third, birds like swallows, and eagles who were usually flying or soaring in the air rather than staying on islands, were also excluded as it was difficult to determine their definite belonging islands. Following these operations, 60 species were finally retained.”

(15) Line 370: I suggest adding the range and median of STI.

Thanks for this good suggestion. The range, mean±SD of STI were already in the Results part, we added the median of STI there as well. The new sentence is now in Results part on Lines 137-139:

“The STI of all 60 birds averaged 19.94 ± 3.58 ℃ (mean ± SD) and ranged from 9.30 ℃ (Cuculus canorus) to 27.20 ℃ (Prinia inornate), with a median of 20.63 ℃ (Appendix 1—figure 2; Appendix 1—figure 3).”

(16) Figure 4.b: Is it possible to be more explicit about what that trend is? the coefficient of the regression Logit(ext/col) ~ year + ...... ?

Thank you for this advice. Your understanding is right: we can interpret it as the coefficient of the ‘year’ effect in the model. More specifically, the ‘year’ effect or temporal trend here is the ‘posterior mean’ of the posterior distribution of ‘year’ in the MSOM (Multi-species Occupancy Model), in the context of the Bayesian framework. We modified this sentence on Lines 811-813:

“ Each point in (b) represents the posterior mean estimate of year in colonization, extinction or occupancy rate for each species.”

(17) Figure 6: is it possible to provide an easily understandable meaning of the prior presented in the Y axis? E.g. "2 corresponds to a 90% probability for a species to go extinct at T+1", if not, please specify that it is the logit of a probability.

Thank you for this question both in Public Review and here. The value on the Y axis indicates the posterior mean of each variable (year, area, isolation and their interaction effects) extracted from the MSOM model, where the logit(extinction rate) or logit(colonization rate) was the response variable. All variables were standardized before analysis to make them comparable. So, positive values indicate positive influence while negative values indicate negative influence. Because the goal of Figure 6 is to display the negative/positive effect, we didn’t back-transform them. Following your advice, we thus modified the caption of Figure 6 (now renumbered as Figure 5, following a comment from Reviewer #3, to move Figure 5 to Figure 4c). The modified title and legends of Figure 5 are on Lines 817-820:

“Figure 5. Posterior estimates of logit-scale parameters related to cold-adapted species’ extinction rates and warm-adapted species’ colonization rates. Points are species-specific posterior means on the logit-scale, where parameters >0 indicate positive effects (on extinction [a] or colonization [b]) and parameters <0 indicate negative effects.”

(18) Line 773: points in blue only are significant? I suggest "points in color".

Thank you for your reminder. Points in blue and orange are all significant. We have revised the sentence on Line 823:

“Points in blue/orange indicate significant effects.”

These are all small suggestions that may help you improve the readability of the final manuscript. I warmly thank you for the opportunity to review this impressive study.

We appreciate your careful review and profound suggestions. We believe these modifications will improve the final manuscript.

Reviewer #3 (Recommendations For The Authors):

I have a few minor suggestions for paper revision for your otherwise excellent manuscript. I wish to emphasize that it was a pleasure to read the manuscript and that I especially enjoyed a very nice flow throughout the ms from a nicely rounded introduction that led well into the research questions and hypotheses all the way to a good and solid discussion.

Thank you very much for your review and recognition. We have carefully checked all recommendations and addressed them in the manuscript.

(1) L 63: space before the bracket missing and I suggest moving the reference to the end of the sentence (directly after habitat fragmentation does not seem to make sense).

Thank you very much for this suggestion. The missed space was added, and the reference has been moved to the end of the sentence. We also add a general definition of habitat fragmentation. The new sentence is on Lines 61-64:

“Habitat fragmentation, usually defined as the shifts of continuous habitat into spatially isolated and small patches (Fahrig, 2003), in particular, has been hypothesized to have interactive effects with climate change on community dynamics.”

(2) L 102: I suggest to write "benefitting ..." instead.

Done.

(3) L 103: higher extinction rates (add "s").

Done.

(4) L 104: this should probably say "emigrate" and "climate warming".

Done.

(5) L 130-133: this is true for emigration (more isolated islands show slower emigration). But what about increased local extinction, especially for small and isolated islands? Especially since you mentioned later in the manuscript that often emigration and extinction are difficult to identify or differentiate. Might be worth a thought here or somewhere in the discussion?

Thank you for this good question. I would like to answer it in two aspects:

Yes, we can’t distinguish between true local extinction and emigration. The observed local “extinction” of cold-adapted species over 10 years may involve two processes that usually occur in order: first “emigration” and then, if can’t emigrate or withstand, “real local dieback”. Over 10 years, the cold-adapted species would have to tolerate before real extinction on remote islands because of disperse limitation, while on less isolated islands it would be easy to emigrate and find a more suitable habitat for the same species. Consequently, it’s harder for us to observe “extinction” of species on more isolated islands, while it’s easier to observe “fake extinct” of species on less isolated islands due to emigration. As a result, the observed extinction rate is expected to increase more sharply for species on less remote islands, while the observed extinction rate is expected to increase relatively moderately for the same species on remote islands.

We have modified the legend of Figure 1 on Lines 780-781:

“Note that extinction here may include both the emigration of species and then the local extinction of species.”

There is also one part in the Discussion that mentions this on Lines 287-291: “While we cannot truly distinguish in our system between local extinction and emigration, we suspect that given two islands equal except in isolation, if both lose suitability due to climate change, individuals can easily emigrate from the island nearer to the mainland, while individuals on the more isolated island would be more likely to be trapped in place until the species went locally extinct due to a lack of rescue”.

Besides, you said “But what about increased local extinction, especially for small and isolated islands?”, I think you are mentioning the “high extinction rate per se on remote islands”. We want to test the “trend” of extinction rate on a temporal scale, rather than the extinction rate per se on a spatial scale. Even though species have a high extinction rate on remote islands, it can also show a slower changing rate in time.

I hope these answers solve the problem.

(6) L 245: I think this is the first time the acronym appears in the ms (as the methods come after the discussion), so please write the full name here too.

Thank you for pointing out this. I realized “Thousand Island Lake” appears for the first time in the last paragraph of the Introduction part. So we add “TIL” there on Lines 108-109:

“Here, we use 10 years of bird community data in a subtropical land-bridge island system (Thousand Island Lake, TIL, China, Figure 2) during a period of consistent climatic warming.”

(7) L 319: this section could end with a summary statement on idiosyncratic responses (i.e. some variation in the responses you found among the species) and the potential reasons for this, such as e.g. the role of other species traits or interactions, as well as other ways to measure habitat fragmentation (see main comments in public review).

Thank you for this suggestion both in Public Review and here. We added a summary statement about the reasons for idiosyncratic responses on Lines 334-338:

“Overall, these idiosyncratic responses reveal several possible mechanisms in regulating species' climate responses, including resource demands and biological interactions like competition and predation. Future studies are needed to take these factors into account to understand the complex mechanisms by which habitat loss meditates species range shifts.”

We only strengthen “habitat loss” here, because idiosyncratic responses mainly come from the mediating effect of habitat loss. For the mediating effect of isolation, the response is relatively consistent (see Page 8, Lines 183-188): “In particular, the effect of isolation on temporal dynamics of thermophilization was relatively consistent across cold- and warm-adapted species (Figure 5a, b); specifically, on islands nearer to the mainland, warm-adapted species (15 out of 15 investigated species) increased their colonization probability at a higher rate over time, while most cold-adapted species (21 out of 23 species) increased their extinction probability at a higher rate”.

(8) L 333: what about the distance to other islands? it's more of a network than a island-mainland directional system (Figure 2). You could address this aspect in the discussion.

Thank you for this good question again. Isolation can be measured in different ways in the study region. We chose distance to the mainland because it was the best predictor of colonization and extinction rate of breeding birds in the study region, and produced similar results like the other distance-based measures, including distance to the nearest landmass, distance to the nearest larger landmass (Si et al., 2014). We still agree with you that it’s necessary to consider more aspects of “isolation” at least in discussion for future research. In Discussion part, we addressed these on Lines 292-299. For more details refer to the response to Public Review.

(9) Figure 2: Is B1 one of the sampled islands? It is clearly much larger than most other islands and I think it could thus serve as an important population source for many of the adjacent smaller islands? Thus, the nearest neighbor distance to B1 could be as important in addition to the distance to the mainland?

Yes, B1 is one of the sampled islands and is also the biggest island. In previous research in our study system, we tried distance to the nearest landmass, to the nearest larger landmass and the nearest mainland, they produced similar results (For more details refer to the response to Public Review). We agree with you that the nearest neighbor distance to B1 could be a potentially important measure, but need further research. In our Discussion, we address these on Lines 292-299:

“As a caveat, we only consider the distance to the nearest mainland as a measure of fragmentation, consistent with previous work in this system (Si et al., 2014), but we acknowledge that other distance-based metrics of isolation that incorporate inter-island connections could reveal additional insights on fragmentation effects. The spatial arrangement of islands, like the arrangement of habitat, can influence niche tracking of species (Fourcade et al., 2021). Future studies should take these metrics into account to thoroughly understand the influence of isolation and spatial arrangement of patches in mediating the effect of climate warming on species.”

(10) L 345: 20km/h walking seems impressively fast? I assume this is a typo.

Sorry for the carelessness, it should be 2.0 km/h. it has been corrected on Lines 375-376:

“In each survey, observers walked along each transect at a constant speed (2.0 km/h) and recorded all the birds seen or heard on the survey islands.”

(11) L 485: I had difficulties fully understanding the models that were fitted here and could not find them in the codes you provided (which were otherwise very well documented!). Could you explain this modeling step in a bit more detail?

Thank you for your recognition! According to Line 485 in the online PDF version (Methods part 4.6.3), it says: “An increasing colonization trend of warm-adapted species and increasing extinction trend of cold-adapted species are two main expected processes that cause thermophilization (Fourcade et al., 2021). To test our third prediction about the mediating effect of habitat fragmentation, we selected warm-adapted species that had an increasing trend in colonization rate (positive year effect in colonization rate) and cold-adapted species that had an increasing extinction rate (positive year effect in extinction rate)…..”

We carefully checked the code in Figshare link and found that the MOSM JAGS code was not uploaded before. Very sorry for that. Now it can be found in the document [MOSM.R] at https://figshare.com/s/7a16974114262d280ef7. Hope the code, together with the modeling process in section 4.5 in the Methods can help to understand the whole modeling process. Besides, we would like to explain how to decide the temporal trend in colonization or extinction of each species related to Line 485. Let’s take the model of species-specific extinction rate for example:

In this model, “Island” was a random effect, “Year” is added as a random slope, thus allowing “year effect” (that is: the temporal trend) of extinction rate of species to vary with “island”. Further, the interaction effect between island variables (isolation, area) was added to test if the “year effect” was related to island area or isolation.

Because we are only interested in warm-adapted species that have a positive temporal trend in colonization and cold-adapted species that have a positive temporal trend in extinction, which are two main processes underlying thermophilizaiton, we choose warm-adapted species that have a positive year-effect in colonization, and cold-adapted species that has a positive year-effect in extinction. Hope this explanation and the JAGS code can help if you are confused about this part.

Hope these explanations can make it clearer.

(12) Figure 1: to me, it would be more intuitive to put the landscape configuration in the titles of the panels b, c, and d instead of "only" the mechanisms. E.g. they could be: a) fragmented islands with low climate buffering; b) small islands with low habitat heterogeneity; c) isolated islands with dispersal limitations?

It is also slightly confusing that the bird communities are above "island" in the middle of the three fragmented habitats - which all look a bit different in terms of tree species and structure which makes the reader first think that it has something to do with the "new" species community. so maybe worth rethinking how to illustrate the three fragmented islands?

We would like to thank you for your nice proposition. Firstly, it’s a good idea to put the landscape configuration in the title of the panels b, c, d. The new title (a) is “Fragmented islands with low climate buffering”, title (b) is “Small islands with low habitat heterogeneity”, and title (c) is “Isolated patches with dispersal limitations”.

Second, we realized that putting the “bird community” above “island” in the middle of the three patches is a bit confusing. Actually, we wanted to show bird communities only on that one island in the middle. The other two patches are only there to represent a fragmented background. To avoid misunderstanding, we added a sentence in the legend of Figure 1 on Lines 778-780:

“The three distinct patches signify a fragmented background and the community in the middle of the three patches was selected to exhibit colonization-extinction dynamics in fragmented habitats.”

(13) Figure 4: please add the description of the color code for panel a.

Sorry for the unclear description. The vertical dashed line indicates the median value of STI for 60 species, as a separation of warm-adapted species and cold-adapted species. We have added these details on Lines 807-809:

“The dotted vertical line indicates the median of STI values. Cold-adapted species are plotted in blue and warm-adapted species are plotted in orange.”

(14) Figure 5: You could consider adding this as panel c to Figure 4 as it depicts the same thing as in 4a but for CTI-abundance.

Thank you for this advice. We have moved the original Figure 5 to Figure 4c. Previous Figure 6 thus turned into Figure 5. All corresponding citations in the main text were checked to adapt to the new index. The new figure is now on Lines 801-815:

References

Ferraz, G., Russell, G. J., Stouffer, P. C., Bierregaard Jr, R. O., Pimm, S. L., & Lovejoy, T. E. (2003). Rates of species loss from Amazonian forest fragments. Proceedings of the National Academy of Sciences, 100(24), 14069-14073. doi:10.1073/pnas.2336195100

Fourcade, Y., WallisDeVries, M. F., Kuussaari, M., van Swaay, C. A., Heliölä, J., & Öckinger, E. (2021). Habitat amount and distribution modify community dynamics under climate change. Ecology Letters, 24(5), 950-957. doi:10.1111/ele.13691

Gaüzère, P., Princé, K., & Devictor, V. (2017). Where do they go? The effects of topography and habitat diversity on reducing climatic debt in birds. Global Change Biology, 23(6), 2218-2229. doi:10.1111/gcb.13500

Gonzalez, A. (2000). Community relaxation in fragmented landscapes: the relation between species richness, area and age. Ecology Letters, 3(5), 441-448. doi:10.1046/j.1461-0248.2000.00171.x

Haddad, N. M., Brudvig, L. A., Clobert, J., Davies, K. F., Gonzalez, A., Holt, R. D., . . . Collins, C. D. (2015). Habitat fragmentation and its lasting impact on Earth’s ecosystems. Science advances, 1(2), e1500052. doi:10.1126/sciadv.1500052

Richard, B., Dupouey, J. l., Corcket, E., Alard, D., Archaux, F., Aubert, M., . . . Macé, S. (2021). The climatic debt is growing in the understorey of temperate forests: Stand characteristics matter. Global Ecology and Biogeography, 30(7), 1474-1487. doi:10.1111/geb.13312

Si, X., Pimm, S. L., Russell, G. J., & Ding, P. (2014). Turnover of breeding bird communities on islands in an inundated lake. Journal of Biogeography, 41(12), 2283-2292. doi:10.1111/jbi.12379

Author Response:

Reviewer #1 (Public review):

Summary:

Fallah and colleagues characterize the connectivity between two basal ganglia output nuclei, the SNr and GPe, and the pedunculopontine nucleus, a brainstem nucleus that is part of the mesencephalic locomotor region. Through a series of systematic electrophysiological studies, they find that these regions target and inhibit different populations of neurons, with anatomical organization. Overall, SNr projects to PPN and inhibits all major cell types, while the GPe inhibits glutamatergic and GABAergic PPN neurons, and preferentially in the caudal part of the nucleus. Optogenetic manipulation of these inputs had opposing effects on behavior - SNr terminals in the PPN drove place aversion, while GPe terminals drove place preference.

Strengths:

This work is a thorough and systematic characterization of a set of relatively understudied circuits. They build on the classic notions of basal ganglia connectivity and suggest a number of interesting future directions to dissect motor control and valence processing in brainstem systems.

We thank the reviewers for these positive comments.

Weaknesses:

Characterization of the behavioral effects of manipulations of these PPN input circuits could be further parsed, for a better understanding of the functional consequences of the connections demonstrated in the ephys analyses.

We will further analyze our behavioral data to reveal more nuanced functional effects.

All the cell type recording studies showing subtle differences in the degree of inhibition and anatomical organization of that inhibition suggest a complex effect of general optogenetic manipulation of SNr or GPe terminals in the PPN. It will be important to determine if SNr or GPe inputs onto a particular cell type in PPN are more or less critical for how the locomotion and valence effects are demonstrated here.

This is a really interesting future direction and we will expand on these points in the discussion.

Reviewer #2 (Public review):

Summary:

Fallah et al carefully dissect projections from SNr and GPe - two key basal ganglia nuclei - to the PPN, an important brainstem nucleus for motor control. They consider inputs from these two areas onto 3 types of downstream PPN neurons: GABAergic, glutamatergic, and cholinergic neurons. They also carefully map connectivity along the rostrocaudal axis of the PPN.

Strengths:

The slice electrophysiology work is technically well done and provides useful information for further studies of PPN. The optogenetics and behavioral studies are thought-provoking, showing that SNr and GPe projections to PPN play distinct roles in behavior.

We appreciate the reviewer’s positive evaluation.

Weaknesses:

Although the optogenetics and behavioral studies are intriguing, they are somewhat difficult to fit together into a specific model of circuit function. Perhaps the authors can work to solidify the connection between these two arms of the work.

We will expand on these topics in the discussion.

(1) Male and female mice are used, but the authors do not discuss any analysis of sex differences. If there are no sex differences, it is still useful to report data disaggregated by sex in addition to pooled data.

While we do not have sufficient n for a well-powered analysis of sex differences in behavior, we find that both male and female mice increase movement in response to SNr axon stimulation and decrease movement in response to GPe axon stimulation. We will expand on this further in the revised manuscript.

(2) There is some lack of clarity in the current manuscript on the ages used - 2-5 months vs "at least 7 weeks." Is 7 weeks the time of virus injection surgery, then recordings 3 weeks later (at least 10 weeks)? Please clarify if these ages apply equally to electrophysiological and behavioral studies. If the age range used for the test is large, it may be useful to analyze and report if there are age-related effects.

7 weeks is the youngest age at which mice used for electrophysiology were injected, and all were used for electrophysiology between 2-5 months. For behavior, the youngest mice used were 11 weeks old at time of behavior (8 weeks old at injection). Mice in the GPe-stimulated condition were 110 ± 7.4 SEM days old and mice in the SNr-stimulated condition 132 ± 23.4 SEM days old. We will add these details to the revised manuscript.

In addition, we have correlated distance traveled at baseline and during stimulation with age for both SNr and GPe stimulated conditions. Baseline distance traveled did not correlate with age, but there was a trend toward more movement during stimulation with older mice in the SNr axon stimulation group. We will discuss this in the revised manuscript.

(3) Were any exclusion criteria applied, e.g. to account for missed injections?

All injection sites and implant sites were within our range of acceptability, so we did not exclude any mice for missed injections.

(4) 28-34degC is a fairly wide range of temperatures for electrophysiological recording, which could affect kinetics.

This is an important consideration. We have checked our main measurement of current amplitude in the condition where we found significant differences between rostral and caudal PPN (SNr to Vglut2 PPN neurons) against temperature and found no correlation (Pearson’s r value = -0.0076). Similarly, we found no correlation between baseline (pre-opto) firing frequency and temperature (r = -0.068).

(5) It would be good to report the number of mice used for each condition in addition to n=cells. Statistically, it would be preferable not to assume that each cell from the same mouse is an independent measurement and to use a nested ANOVA.

For electrophysiology, the number of mice used in each experiment was 6 (3 male, 3 female). In the manuscript ‘N’ represents number of mice and ‘n’ represents number of cells. Because of the unpredictability of how many healthy cells can be recorded from one mouse, our data were planned to be collected with n=cells, and are underpowered for a nested ANOVA. However, rostral and caudal data were collected from the same mice. While we do not have sufficient paired data for each parameter, analyzing one of our main and most important findings with a paired comparison (with biological replicates being mice) shows a statistically significant difference in the inhibitory effect of SNr axon stimulation on firing rate between rostral and caudal glutamatergic neurons (p=0.031, Wilcoxon signed rank test).

Reviewer #3 (Public review):

Summary:

The study by Fallah et al provides a thorough characterization of the effects of two basal ganglia output pathways on cholinergic, glutamatergic, and GABAergic neurons of the PPN. The authors first found that SNr projections spread over the entire PPN, whereas GPe projections are mostly concentrated in the caudal portion of the nucleus. Then the authors characterized the postsynaptic effects of optogenetically activating these basal ganglia inputs and identified the PPN's cell subtypes using genetically encoded fluorescent reporters. Activation of inputs from the SNr inhibited virtually all PPN neurons. Activation of inputs from the GPe predominantly inhibited glutamatergic neurons in the caudal PPN, and to a lesser extent GABAergic neurons. Finally, the authors tested the effects of activating these inputs on locomotor activity and place preference. SNr activation was found to increase locomotor activity and elicit avoidance of the optogenetic stimulation zone in a real-time place preference task. In contrast, GPe activation reduced locomotion and increased the time in the RTPP stimulation zone.

Strengths:

The evidence of functional connectivity of SNr and GPe neurons with cholinergic, glutamatergic, and GABAergic PPN neurons is solid and reveals a prominent influence of the SNr over the entire PPN output. In addition, the evidence of a GPe projection that preferentially innervates the caudal glutamatergic PPN is unexpected and highly relevant for basal ganglia function.

Opposing effects of two basal ganglia outputs on locomotion and valence through their connectivity with the PPN.

Overall, these results provide an unprecedented cell-type-specific characterization of the effects of basal ganglia inputs in the PPN and support the well-established notion of a close relationship between the PPN and the basal ganglia.

We thank the reviewer for their positive comments.

Weaknesses:

The behavioral experiments require further analysis as some motor effects could have been averaged out by analyzing long segments.

We will further analyze our motor effects in the revised manuscript.

Additional controls are needed to rule out a motor effect in the real-time place preference task.

This is an important point. Our use of unilateral stimulation in the RTPP task reduces potential motor effects, and our supplemental videos show that the mice can easily escape and enter the stimulated zone. However, we can't completely rule out a motor component. To delve into this further, we analyzed mouse speed in the RTPP task. We find that in both SNr and GPe stimulation conditions, the maximum speed of the mouse is not different in the stimulated vs unstimulated zone. We will further analyze mouse speed at the transition into and out of the stimulated zone to identify any acute motor effects in this experiment.

Importantly, the location of the stimulation is not reported even though this is critical to interpret the behavioral effects.

The implant locations were generally over the middle-to-rostral PPN and we will clarify this in the revised manuscript. These locations are shown in figure 7B.

There are some concerns about the possible recruitment of dopamine neurons in the SNr experiments.

We are very interested in this possibility and plan to discuss this with more clarity in a revised manuscript.

Author response:

The following is the authors’ response to the previous reviews.

Reviewer #1 (Recommendations For The Authors): 

This is not a recommendation. While reading old literature, I found some interesting facts. The shape of the neurocranium in monotremes, birds, and mammals, at least in early stages, resembles the phenotype of 'dact'1/2, wnt11f2, or syu mutants. For more details, see DeBeer's: 'The Development of the Vertebrate Skull, !937' Plate 137. 

Thank you for pointing this out. It is indeed interesting.

Minor Comments: 

• Lines 64, 66, and 69: same citation without interruption: Heisenberg, Brand et al. 1996

Revised line 76. 

• Lines 101 and 102: same citation without interruption: Li, Florez et al. 2013 

Revised line 118.

• Lines 144, 515, 527, and 1147: should be wnt11f2 instead of wntllf2 - if not, then explain 

Revised lines 185, 625, 640,1300.

• Lines 169 and 171: incorrect figure citation: Fig 1D - correct to Fig 1F 

Revised lines 217, 219.

• Line 173: delete (Fig. S1) 

Revised line 221.

• Line 207: indicate that both dact1 and dact2 mRNA levels increased, noting a 40% higher level of dact2 mRNA after deletion of 7 bp in the dact2 gene 

Revised line 265.

• Line 215: Fig 1F instead of Fig 1D 

Revised line 217.

• Line 248: unify naming of compound mutants to either dact1/2 or dact1/dact2 compound mutants 

Revised to dact1/2 throughout.

• Line 259: incorrect figure citation: Fig S1 - correct to Fig S2D/E 

Revised line 324.

• Line 302: correct abbreviation position: neural crest (NCC) cell - change to neural crest cell (NCC) population 

Revised line 380.

• Line 349: repeating kny mut definition from line 70 may be unnecessary 

Revised line 434.

• Line 351: clarify distinction between Fig S1 and Fig S2 in the supplementary section 

Revised line 324.

• Line 436: refer to the correct figure for pathways associated with proteolysis (Fig 7B) 

Revised line 530.

• Line 446-447: complete the sentence and clarify the relevance of smad1 expression, and correct the use of "also" in relation to capn8 

Revised line 567.

• Line 462: clarify that this phenotype was never observed in wildtype larvae, and correct figure reference to exclude dact1+/- dact2+/- 

Revised line 563, 568.

• Line 463: explain the injection procedure into embryos from dact1/2+/- interbreeding 

Revised line 565.

• Lines 488 and 491: same citation without interruption: Waxman, Hocking et al. 2004 

Revised line 591.

• Line 502: maintain consistency in referring to TGF-beta signaling throughout the article 

Revised throughout.

• Line 523: define CNCC; previously used only NCC 

Revised to cranial NCC throughout.

• Line 1105: reconsider citing another work in the figure legend 

Revised line 1249.

• Line 1143: consider using "mutant" instead of "mu" 

Revised line 1295.

• Fig 2A/B: indicate the number of animals used ("n") 

N is noted on line 1274.

• Fig 2C, D, E: ensure uniform terminology for control groups ("wt" vs. "wildtype") 

Revised in figure.

• Fig 7C: clarify analysis of dact1/2-/- mutant in lateral plate mesoderm vs. ectoderm 

Revised line 1356.

• Fig 8A: label the figure to indicate it shows capn8, not just in the legend 

Revised.

• Fig 8D: explain the black/white portions and simplify to highlight important data 

Revised.

• Fig S2: add the title "Figure S2" 

Revised.

• Consider omitting the sentence: "As with most studies, this work has contributed some new knowledge but generated more questions than answers." 

Revised line 720.

Reviewer #2 (Recommendations For The Authors): 

Major comments: 

(1) The authors have addressed many of the questions I had, including making the biological sample numbers more transparent. It might be more informative to use n = n/n, e.g. n = 3/3, rather than just n = 3. Alternatively, that information can be given in the figure legend or in the form of penetrance %. 

The compound heterozygote breeding and phenotyping analyses were not carried out in such a way that we can comment on the precise % penetrance of the ANC phenotype, as we did not dissect every ANC and genotype every individual that resulted from the triple heterozygote in crossings. We collected phenotype/genotype data until we obtained at least three replicates.

We did genotype every individual resulting from dact1/2 dHet crosses to correlate genotype to the phenotype of the embryonic convergent extension phenotype and narrowed ethmoid plate (Fig. 2A, Fig. 3) which demonstrated full penetrance.

(2) The description of the expression of dact1/2 and wnt11f2 is not consistent with what the images are showing. In the revised figure 1 legend, the author says "dact2 and wnt11f2 transcripts are detected in the anterior neural plate" (line 1099)", but it's hard to see wnt11f2 expression in the anterior neural plate in 1B. The authors then again said " wnt11f2 is also expressed in these cells", referring to the anterior neural plate and polster (P), notochord (N), paraxial and presomitic mesoderm (PM) and tailbud (TB). However, other than the notochord expression, other expression is actually quite dissimilar between dact2 and wnt11f2 in 1C. The authors should describe their expression more accurately and take that into account when considering their function in the same pathway. 

We have revised these sections to more carefully describe the expression patterns. We have added references to previous descriptions of wnt11 expression domains.

(3) Similar to (2), while the Daniocell was useful in demonstrating that expression of dact1 and dact2 are more similar to expression of gpc4 and wnt11f2, the text description of the data is quite confusing. The authors stated "dact2 was more highly expressed in anterior structures including cephalic mesoderm and neural ectoderm while dact1 was more highly expressed in mesenchyme and muscle" (lines 174-176). However, the Daniocell seems to show more dact1 expression in the neural tissues than dact2, which would contradict the in situ data as well. I think the problem is in part due to the dataset contains cells from many different stages and it might be helpful to include a plot of the cells at different stages, as well as the cell types, both of which are available from the Daniocell website. 

We have revised the text to focus the Daniocell analysis on the overall and general expression patterns. Line 220.

(4) The authors used the term "morphological movements" (line 337) to describe the cause of dact1/2 phenotypes. Please clarify what this means. Is it cell movement? Or is it the shape of the tissues? What does "morphological movements" really mean and how does that affect the formation of the EP by the second stream of NCCs? 

We have revised this sentence to improve clarity. Line 416.

(5) In the first submission, only 1 out of 142 calpain-overexpressing animals phenocopied dact1/2 mutants and that was a major concern regarding the functional significance of calpain 8 in this context. In the revised manuscript, the authors demonstrated that more embryos developed the phenotype when they are heterozygous for both dact1/2. While this is encouraging, it is interesting that the same phenomenon was not observed in the dact1-/-; dact2+/- embryos (Fig. 6D). The authors did not discuss this and should provide some explanation. The authors should also discuss sufficiency vs requirement tested in this experiment. However, given that this is the most novel aspect of the paper, performing experiments to demonstrate requirements would be important. 

We have added a statement regarding the non-effect in dact1-/-;dact2+/- embryos. Line 568-570. We have also added discussion of sufficiency vs necessity/requirement testing. Line 676-679.

(6) Related to (5), the authors cited figure 8c when mentioning 0/192 gfp-injected embryos developed EP phenotypes. However, figure 8c is dact1/2 +/- embryos. The numbers also doesn't match the numbers in Figure 8d either. Please add relevant/correct figures. 

The text has been revised to distinguish between our overexpression experiment in wildtype embryos (data not shown) versus overexpression in dact1/2 double het in cross embryos (Fig 8).

Minor comments: 

(1) Fig 1 legend line 1106 "the midbrain (MP)" should be MB 

Revised line 1250.

(2) Wntllf2, instead of wnt11f2, (i.e. the letter "l" rather than the number "1") was used in 4 instances, line 144, 515, 527, 1147 

Revised lines 185, 625, 640,1300.

(3) The authors replaced ANC with EP in many instances, but ANC is left unchanged in some places and it's not defined in the text. It's first mentioned in line 170.

Revised line 218.

Author response:

The following is the authors’ response to the original reviews.

Public Reviews:

Reviewer #1 (Public Review):

Summary:

The manuscript gives a broad overview of how to write NeuroML, and a brief description of how to use it with different simulators and for different purposes - cells to networks, simulation, optimization, and analysis. From this perspective, it can be an extremely useful document to introduce new users to NeuroML.

We are glad the reviewer found our manuscript useful.

However, the manuscript itself seems to lose sight of this goal in many places, and instead, the description at times seems to target software developers. For example, there is a long paragraph on the board and user community. The discussion on simulator tools seems more for developers, not users. All the information presented at the level of a developer is likely to be distracting to eLife readership.

To make the paper less developer focussed and more accessible to the end user we have shortened the long paragraphs on the board and user community (and moved some of this text to the Methods section; lines: 524-572 in the document with highlighted changes). We have also made the discussion on simulator tools more focussed on the user (lines 334-406). However, we believe some information on the development and oversight of NeuroML and its community base are relevant to the end user, so we have not removed these completely from the main text.

Strengths:

The modularity of NeuroML is indeed a great advantage. For example, the ability to specify the channel file allows different channels to be used with different morphologies without redundancy. The hierarchical nature of NeuroML also is commendable, and well illustrated in Figures 2a through c.

The number of tools available to work with NeuroML is impressive.

The abstract, beginning, and end of the manuscript present and discuss incorporating NeuroML into research workflows to support FAIR principles.

Having a Python API and providing examples using this API is fantastic. Exporting to NeuroML from Python is also a great feature.

We are glad the reviewer appreciated the design of NeuroML and its support for FAIR principles.

Weaknesses:

Though modularity is a strength, it is unclear to me why the cell morphology isn't also treated similarly, i.e., specify the morphology of a multi-compartmental model in a separate file, and then allow the cell file to specify not only the files containing channels, but also the file containing the multi-compartmental morphology, and then specify the conductance for different segment groups. Also, after pynml_write_neuroml2_file, you would not have a super long neuroML file for each variation of conductances, since there would be no need to rewrite the multi-compartmental morphology for each conductance variation.

We thank the reviewer for highlighting this shortcoming in NeuroML2. We have now added the ability to reference externally defined (e.g. in another file) <morphology> and <biophysicalProperties> elements from <cells>. This has enabled the morphologies and/or specification of ionic conductances to be separated out and enables more streamlined analysis of cells with different properties, as requested. Simulators NEURON, NetPyNE and EDEN already support this new form. Information on this feature has been added to https://docs.neuroml.org/Userdocs/ImportingMorphologyFiles.html#neuroml2 and also mentioned in the text (lines 188-190).

This would be especially important for optimizations, if each trial optimization wrote out the neuroML file, then including the full morphology of a realistic cell would take up excessive disk space, as opposed to just writing out the conductance densities. As long as cell morphology must be included in every cell file, then NeuroML is not sufficiently modular, and the authors should moderate their claim of modularity (line 419) and building blocks (551).

We believe the new functionality outlined above addresses this issue, as a single file containing the <morphology> element could be referenced, while a much smaller file, containing the channel distributions in a <biophysicalProperties> element would be generated and saved on each iteration of the optimisation.

In addition, this is very important for downloading NeuroML-compliant reconstructions from NeuroMorpho.org. If the cell morphology cannot be imported, then the user has to edit the file downloaded from NeuroMorpho.org, and provenance can be lost.

While the NeuroMorpho.Org website does support converting reconstructed morphologies in SWC format to NeuroML, this export feature is no longer supported on most modern browsers due to it being based on Java Applet technologies. However, a desktop version of this application, CVApp, is actively maintained

(https://github.com/NeuroML/Cvapp-NeuroMorpho.org), and we have updated it to support export of the SWC to the standalone <morphology> element form of NeuroML discussed above. Additionally, a new Python application for conversion of SWC to NeuroML is in development and will be incorporated into PyNeuroML (Google Summer of Code 2024). Our documentation has been updated with the recommended use of SWC in NeuroML based modelling here: https://docs.neuroml.org/Userdocs/Software/Tools/SWC.html

We have also included URLs to the tool and the documentation in the paper (lines: 473-474).

SWC files, however, cannot be used “as is” for modelling since they only include information (often incomplete—for example a single point may represent a soma in SWC files) on the points that make the cell, but not on the sections/segments/cables that these form. Therefore, NeuroML and other simulation tools, including NEURON, must convert these into formats suitable for simulation. The suggested pipeline for use of NeuroMorpho SWC files would therefore be to convert them to NeuroML, check that they represent the intended compartmentalisation of the neuron and then use them in models.

To ensure that provenance is maintained in all NeuroML models (including conversions from other formats), NeuroML supports the addition of RDF annotations using the COMBINE annotation specifications in model files:

https://docs.neuroml.org/Userdocs/Provenance.html. We have added this information to the paper (lines: 464-465).

Also, Figure 2d loses the hierarchical nature by showing ion channels, synapses, and networks as separate main branches of NeuroML.

While an instance of an ion channel is on a segment, in a cell, in a population (and hence there is a hierarchy between them), in terms of layout in a NeuroML file the ion channel is defined at the “top level” so that it can be referenced and used by multiple cells, the cell definitions are also defined top level, and used in multiple populations, etc. There are multiple ways to depict these relationships between entities, and we believe Fig 2d complements Fig 2a-c (which is more hierarchical), by emphasising the different categories of entities present in NeuroML files. We have modified the caption of Figure 2d to clarify that it shows the main categories of elements included in the NeuroML standard in their respective hierarchies.

In Figure 5, the difference between the core and native simulator is unclear.

We have modified the figure and text (lines: 341) to clarify this. We now say “reference” simulators instead of “core”. This emphasises that jNeuroML and pyLEMS are intended as reference implementations in each of their languages of how to interpret NeuroML models, as opposed to high performance simulators for research use. We have also updated the categorization of the backends in the text accordingly.

What is involved in helper scripts?

Simulators such as NetPyNE can import NeuroML into their own internal format, but require some boilerplate code to do this (e.g. the NetPyNE scripts calls the importNeuroML2SimulateAnalyze() method with appropriate parameters). The NeuroML tools generate short scripts that use this boilerplate code. We have renamed “helper scripts” to “import scripts'' for clarity (Figure 5 and its caption).

I thought neurons could read NeuroML? If so, why do you need the export simulator-specific scripts?

The NEURON simulator does have some NeuroML functionality (it can export cells, though not the full network, to NeuroML 2 through its ModelView menu), but does not natively support reading/importing of NeuroML in its current version. But this is not a problem as jNeuroML/PyNeuroML translates the NeuroML model description into NEURON’s formats: Python scripts/HOC/Nmodl which NEURON then executes.

As NEURON is the simulator which allows simulation of the widest range of NeuroML elements, we have (in agreement with the NEURON developers) concentrated on incorporating the best support for NeuroML import/export in the latest (easy to install/update) releases of PyNeuroML, rather than adding this to the Neuron source code. NEURON’s core features have been very stable for years and many versions of the simulator are used by modellers - installing the latest PyNeuroML gives them the latest NEURON support without having to reinstall the latter.

In addition, it seems strange to call something the "core" simulation engine, when it cannot support multi-compartmental models. It is unclear why "other simulators" that natively support NeuroML cannot be called the core.

We agree that this terminology was confusing. As mentioned above, we have changed “core simulator” to “reference simulator”, to emphasise the roles of these simulation engine options.

It might be more helpful to replace this sort of classification with a user-targeted description. The authors already state which simulators support NeuroML and which ones need code to be exported. In contrast, lines 369-370 mention that not all NeuroML models are supported by each simulator. I recommend expanding this to explain which features are supported in each simulator. Then, the unhelpful separation between core and native could be eliminated.

As suggested, we have grouped the simulators in terms of function and removed the core/ non-core distinction. We have also added a table (Table 3) in the appendices that lists what features each simulation engine supports and updated the text to be more user focussed (lines: 348-394).

The body of the manuscript has so much other detail that I lose sight of how NeuroML supports FAIR. It is also unclear who is the intended audience. When I get to lines 336-344, it seems that this description is too much detail for the eLife audience. The paragraph beginning on line 691 is a great example of being unclear about who is the audience. Does someone wanting to develop NeuroML models need to understand XSD schema? If so, the explanation is not clear. XSD schema is not defined and instead explains NeuroML-specific aspects of XSD. Lines 734-735 are another example of explaining to code developers (not model developers).

We have modified these sentences to be more suitable for the general eLife audience: we have moved the explanation of how the different simulator backends are supported to the more technically detailed Methods section (lines 882-942).

While the results sections focus on documenting what users can do with NeuroML, the Methods sections include information on “how” the NeuroML and software ecosystem function. While the information in the methods sections may not be required by users who want to use the standard NeuroML model elements, those users looking to extend NeuroML with their own model entities and/or contribute these for inclusion in the NeuroML standard will require some understanding of how the schema and component types work.

We have tried to limit this information to the bare minimum, pointing to online documentation where appropriate. XSD schemas are, for example, briefly introduced at the beginning of the section “The NeuroML XML Schema”. We have also included a link to the W3C documentation on XSD schemas as a footnote (line 724).

Reviewer #2 (Public Review):

Summary:

Developing neuronal models that are shareable, reproducible, and interoperable allows the neuroscience community to make better use of published models and to collaborate more effectively. In this manuscript, the authors present a consolidated overview of the NeuroML model description system along with its associated tools and workflows. They describe where different components of this ecosystem lay along the model development pathway and highlight resources, including documentation and tutorials, to help users employ this system.

Strengths:

The manuscript is well-organized and clearly written. It effectively uses the delineated model development life cycle steps, presented in Figure 1, to organize its descriptions of the different components and tools relating to NeuroML. It uses this framework to cover the breadth of the software ecosystem and categorize its various elements. The NeuroML format is clearly described, and the authors outline the different benefits of its particular construction. As primarily a means of describing models, NeuroML also depends on many other software components to be of high utility to computational neuroscientists; these include simulators (ones that both pre-date NeuroML and those developed afterwards), visualization tools, and model databases.

Overall, the rationale for the approach NeuroML has taken is convincing and well-described. The pointers to existing documentation, guides, and the example usages presented within the manuscript are useful starting points for potential new users. This manuscript can also serve to inform potential users of features or aspects of the ecosystem that they may have been unaware of, which could lower obstacles to adoption. While much of what is presented is not new to this manuscript, it still serves as a useful resource for the community looking for information about an established, but perhaps daunting, set of computational tools.

We are glad the reviewer appreciated the utility of the manuscript.

Weaknesses:

The manuscript in large part catalogs the different tools and functionalities that have been produced through the long development cycle of NeuroML. As discussed above, this is quite useful, but it can still be somewhat overwhelming for a potential new user of these tools. There are new user guides (e.g., Table 1) and example code (e.g. Box 1), but it is not clear if those resources employ elements of the ecosystem chosen primarily for their didactic advantages, rather than general-purpose utility. I feel like the manuscript would be strengthened by the addition of clearer recommendations for users (or a range of recommendations for users in different scenarios).

To make Table 1 more accessible to users and provide recommendations we have added the following new categories: Introductory guides aimed at teaching the fundamental

NeuroML concepts; Advanced guides illustrating specific modelling workflows; and Walkthrough guides discussing the steps required for converting models to NeuroML. Box 1 has also been improved to clearly mark API and command line examples.

For example, is the intention that most users should primarily use the core NeuroML tools and expand into the wider ecosystem only under particular circumstances? What are the criteria to keep in mind when making that decision to use alternative tools (scale/complexity of model, prior familiarity with other tools, etc.)? The place where it seems most ambiguous is in the choice of simulator (in part because there seem to be the most options there) - are there particular scenarios where the authors may recommend using simulators other than the core jNeuroML software?

The interoperability of NeuroML is a major strength, but it does increase the complexity of choices facing users entering into the ecosystem. Some clearer guidance in this manuscript could enable computational neuroscientists with particular goals in mind to make better strategic decisions about which tools to employ at the outset of their work.

As mentioned in the response to Reviewer 1, the term “core simulator” for jNeuroML was confusing, as it suggested that this is a recommended simulation tool. We have changed the description of jNeuroML to a “reference simulator” to clarify this (Figure 5 and lines 341, 353).

In terms of giving specific guidance on which simulator to use, we have focussed on their functionality and limitations rather than recommending a specific tool (as simulator independent standards developers we are not in a position to favour particular simulators). While NEURON is the most widely used simulator currently, other simulation opinions (e.g. EDEN) have emerged recently which provide quite comprehensive NeuroML support and similar performance. Our approach is to document and promote all supported tools, while encouraging innovation and new developments. The new Table 3 in the Appendix gives a guide to assist users in choosing which simulator may best suit their needs and we have updated the text to include a brief description (lines 348-394).

Recommendations for the authors:

Reviewer #1 (Recommendations For The Authors):

I do not understand what the $comments mean in Box 1. It isn't until I get further in the text that I realize that those are command line equivalents to the Python commands.

We thank the reviewer for highlighting this confusion. We’ve now explicitly marked the API usage and command line usage example columns to make this clearer. We have also used “>” instead of “$” now to indicate the command line,

In Figure 9 Caption "Examples of analysis functions ..", the word analysis seems a misnomer, as these graphs all illustrate the simulation output and graphing of existing variables. I think analysis typically refers to the transformation of variables, such as spike counts and widths.

To clarify this we have changed the caption to “Examples of visualizing biophysical properties of a NeuroML model neuron”.

Figure 10: Why is the pulse generator part of a model? Isn't that the input to a model?

Whether the input to the model is described separately from the NeuroML biophysical description or combined with it is a choice for the researcher. This is possible because in NeuroML any entity which has time varying states can be a NeuroML element, including the current pulse generator. In this simple example the input is contained within the same file (and therefore <neuroml> element) as the cell. However, this does not need to be the case. The cell could be fully specified in its own NeuroML file and then this can be included in other files which add different inputs to facilitate different simulation scenarios. The Python scripting interface facilitates these types of workflows.

In the interest of modularity, can stim information be stored in a separate file and "included"?

Yes, as mentioned above, the stimulus could be stored in a separate file.

I find it strange to use a cell with mostly dimensionless numbers as an example. I think it would be more helpful to use a model that was more physiological.

In choosing an example model type to use to illustrate the use of LEMS (Fig 12), NeuroML (Fig 10), XML Schema (Fig 11), the Python API (Fig 13) and online documentation (Fig 15), we needed an example which showed a sufficiently broad range of concepts (dimensional parameters, state variables, time derivatives), but which is sufficiently compact to allow a concise depiction of the key elements in figures, that fit in a single page (e.g. Fig 12). We felt that the Hindmarsh Rose model, while not very physiological, was well suited for this purpose (explaining the underlying technologies behind the NeuroML specification). The simplicity of the Hindmarsh Rose model is counterbalanced in the manuscript by the detailed models of neurons and circuits in Figures 7 & 9. The latter shows a morphologically and biophysically detailed cortical L5b pyramidal cell model.

In lines 710-714, it is unclear what is being validated. That all parameters are defined? Using the units (or lack thereof) defined in the schema?

Validation against the schema is “level 1” validation where the model structure, parameters, parameter values and their units, cardinality, and element positioning in the model hierarchy are checked. We have updated the paragraph to include this information and to also point to Figure 6 where different levels of validation are explained.

Lines 740 to 746 are confusing. If 1-1 between XSD and LEMS (1st sentence) then how can component types be defined in LEMS and NOT added to the standard? Which is it? 1-1 or not 1-1?

For the curated model elements included in the NeuroML standard, there will be a 1-1 correspondence between their component type definitions in LEMS and type definitions in the XSD schema. New user defined component types (e.g. a new abstract cell model) can be specified in LEMS as required, and these do not need to be included in the XSD schema to be loaded/simulated. However, since they are not present in the schema definition of the core/curated elements, they cannot be validated against it (level 1 validation). We have modified the text to make this clearer (line: 778).

Nonetheless, if the new type is useful for the wider community, it can be accepted by the Editorial Board, and at that stage it will be incorporated into the core types, and added to the Schema, to be part of “valid NeuroML”.

Figure 12. select="synapses[*]/i" is not explained. Does /i mean that iSyn is divided by i, which is current (according to the sentence 3 lines after 766) or perhaps synapse number?

We thank the reviewer for highlighting this confusion. We have now explained the construct in the text (lines 810-812). It denotes “select the i (current) values from all Attachments which have the id ‘synapses’”. These multiple values should be reduced down to a single value through addition, as specified by the attribute: reduce=”add”.

The line after 766 says that "DerivedVariables, variables whose values depend on other variables". You should add "and that are not derivatives, which are handled separately" because by your definition derivatives are derived variables.

Thank you. We have updated the text with your suggestion

Reviewer #2 (Recommendations For The Authors):

- Figure 9: I found it somewhat confusing to have the header from the screenshot at the top ("Layer 5 Burst Accommodating Double Bouquet Cell (5)") not match the morphology shown at the bottom. It's not visually clear that the different panels in Figure 9 may refer to unrelated cells/models.

Thank you for pointing this out. We have replaced the NeuroML-DB screenshot with one of the same Layer 5b pyramidal cells shown in the panels below it.

Additional change:

Figure 7c (showing the NetPyNE-UI interface) has been replaced. Previously, this displayed a 3D model which had been created in NetPyNE itself, but now shows a model which has been created in NeuroML and imported for display/simulation in NetPyNE-UI, and therefore better illustrates NeuroML functionality.

Author response:

To Reviewer #1:

Thank you for your kind words regarding the novelty, study design, and evidence presented. We will clarify our language when describing fuzzy local-linear regression discontinuity analysis. We thank you for this feedback as our goals are to introduce these methods to a neuroscientific audience. Lastly, we will respond and clarify the methodological points, including post-selection inference, bandwidths, and Bayesian analysis in version 2.

To Reviewers #2 and #3:

We thank you both for your constructive feedback, specifically in highlighting 1) the scope of the intervention and 2) the UKB-neuro healthy volunteer bias. In the next manuscript version, we will expand our discussion of plausible reasons for not finding an effect – weighing up the strengths and limitations of our study in 3 aspects; statistical (RD power), design-based (lack of representativeness vs. large sample), and mechanistic (the impact/or lack thereof of one-year of education on neural plasticity decades later). As we believe the approach of natural experiments with RD designs has considerable promise for the field of population cognitive neuroscience beyond this particular study, we will address each of these points within a broader section focused on considerations on how to optimize the insight, power, and inferences gained in future work within and beyond Biobank. Moreover, we will situate our discussion on the magnitude of the educational intervention among a broader discussion of cognitive training versus education, and short - versus long-term effects. We believe revising the manuscript will improve interpretation for the reader and thank you for your in-depth feedback. Lastly, we will provide a point-by-point response in the next version.

Author response:

Public Reviews: 

Reviewer #1 (Public review): 

The conserved AAA-ATPase PCH-2 has been shown in several organisms including C. elegans to remodel classes of HORMAD proteins that act in meiotic pairing and recombination. In some organisms the impact of PCH-2 mutations is subtle but becomes more apparent when other aspects of recombination are perturbed. Patel et al. performed a set of elegant experiments in C. elegans aimed at identifying conserved functions of PCH-2. Their work provides such an opportunity because in C. elegans meiotically expressed HORMADs localize to meiotic chromosomes independently of PCH-2. Work in C. elegans also allows the authors to focus on nuclear PCH-2 functions as opposed to cytoplasmic functions also seen for PCH-2 in other organisms. 

The authors performed the following experiments: 

(1) They constructed C. elegans animals with SNPs that enabled them to measure crossing over in intervals that cover most of four of the six chromosomes. They then showed that doublecrossovers, which were common on most of the four chromosomes in wild-type, were absent in pch-2. They also noted shifts in crossover distribution in the four chromosomes. 

(2) Based on the crossover analysis and previous studies they hypothesized that PCH-2 plays a role at an early stage in meiotic prophase to regulate how SPO-11 induced double-strand breaks are utilized to form crossovers. They tested their hypothesis by performing ionizing irradiation and depleting SPO-11 at different stages in meiotic prophase in wild-type and pch-2 mutant animals. The authors observed that irradiation of meiotic nuclei in zygotene resulted in pch-2 nuclei having a larger number of nuclei with 6 or greater crossovers (as measured by COSA-1 foci) compared to wildtype. Consistent with this observation, SPO11 depletion, starting roughly in zygotene, also resulted in pch-2 nuclei having an increase in 6 or more COSA-1 foci compared to wild type. The increased number at this time point appeared beneficial because a significant decrease in univalents was observed. 

(3) They then asked if the above phenotypes correlated with the localization of MSH-5, a factor that stabilizes crossover-specific DNA recombination intermediates. They observed that pch-2

mutants displayed an increase in MSH-5 foci at early times in meiotic prophase and an unexpectedly higher number at later times. They conclude based on the differences in early MSH-5 localization and the SPO-11 and irradiation studies that PCH-2 prevents early DSBs from becoming crossovers and early loading of MSH-5. By analyzing different HORMAD proteins that are defective in forming the closed conformation acted upon by PCH-2, they present evidence that MSH-5 loading was regulated by the HIM-3 HORMAD. 

(4) They performed a crossover homeostasis experiment in which DSB levels were reduced. The goal of this experiment was to test if PCH-2 acts in crossover assurance. Interestingly, in this background PCH-2 negative nuclei displayed higher levels of COSA-1 foci compared to PCH-2 positive nuclei. This observation and a further test of the model suggested that "PCH-2's presence on the SC prevents crossover designation." 

(5) Based on their observations indicating that early DSBS are prevented from becoming crossovers by PCH-2, the authors hypothesized that the DNA damage kinase CHK-2 and PCH2 act to control how DSBs enter the crossover pathway. This hypothesis was developed based on their finding that PCH-2 prevents early DSBs from becoming crossovers and previous work showing that CHK-2 activity is modulated during meiotic recombination progression. They tested their hypothesis using a mutant synaptonemal complex component that maintains high CHK-2 activity that cannot be turned off to enable crossover designation. Their finding that the pch-2 mutation suppressed the crossover defect (as measured by COSA-1 foci) supports their hypothesis. 

Based on these studies the authors provide convincing evidence that PCH-2 prevents early DSBs from becoming crossovers and controls the number and distribution of crossovers to promote a regulated mechanism that ensures the formation of obligate crossovers and crossover homeostasis. As the authors note, such a mechanism is consistent with earlier studies suggesting that early DSBs could serve as "scouts" to facilitate homolog pairing or to coordinate the DNA damage response with repair events that lead to crossing over. The detailed mechanistic insights provided in this work will certainly be used to better understand functions for PCH-2 in meiosis in other organisms. My comments below are aimed at improving the clarity of the manuscript. 

We thank the reviewer for their concise summary of our manuscript and their assessment of our work as “convincing” and providing “detailed mechanistic insight.”

Comments 

(1) It appears from reading the Materials and Methods that the SNPs used to measure crossing over were obtained by mating Hawaiian and Bristol strains. It is not clear to this reviewer how the SNPs were introduced into the animals. Was crossing over measured in a single animal line? Were the wild-type and pch-2 mutations made in backgrounds that were isogenic with respect to each other? This is a concern because it is not clear, at least to this reviewer, how much of an impact crossing different ecotypes will have on the frequency and distribution of recombination events (and possibly the recombination intermediates that were studied). 

We will clarify these issues in the Materials and Methods of an updated preprint. The control and pch-2 mutants were isogenic in either the Bristol or Hawaiian backgrounds. Control lines were the original Bristol and Hawaiian lines and pch-2 mutants were originally made in the Bristol line and backcrossed at least 3 times before analysis. Hawaiian pch-2 mutants were made by backcrossing pch-2 mutants at least 7 times to the Hawaiian background and verifying the presence of Hawaiian SNPs on all chromosomes tested in the recombination assay. To perform the recombination assays, these isogenic lines were crossed to generate the relevant F1s.

(2) The authors state that in pch-2 mutants there was a striking shift of crossovers (line 135) to the PC end for all of the four chromosomes that were tested. I looked at Figure 1 for some time and felt that the results were more ambiguous. Map distances seemed similar at the PC end for wildtype and pch-2 on Chrom. I. While the decrease in crossing over in pch-2 appeared significant for Chrom. I and III, the results for Chrom. IV, and Chrom. X. seemed less clear. Were map distances compared statistically? At least for this reviewer the effects on specific intervals appear less clear and without a bit more detail on how the animals were constructed it's hard for me to follow these conclusions. 

We hope that the added details above makes the results of these assays more clear. Map distances were compared and did not satisfy statistical significance, except where indicated. While we agree that the comparisons between control animals and pch-2 mutants may seem less clear with individual chromosomes, we argue that more general patterns become clear when analyzing multiple chromosomes. Indeed, this is why we expanded our recombination analysis beyond Chromosome III and the X Chromosomes, as reported in Deshong, 2014. 

(3) Figure 2. I'm curious why non-irradiated controls were not tested side-by-side for COSA-1 staining. It just seems like a nice control that would strengthen the authors' arguments. 

We will add these controls in the updated preprint.

(4) Figure 3. It took me a while to follow the connection between the COSA-1 staining and DAPI staining panels (12 hrs later). Perhaps an arrow that connects each set of time points between the panels or just a single title on the X-axis that links the two would make things clearer. 

We will make changes in the updated preprint to make this figure more clear.

Reviewer #2 (Public review): 

Summary: 

This paper has some intriguing data regarding the different potential roles of Pch-2 in ensuring crossing over. In particular, the alterations in crossover distribution and Msh-5 foci are compelling. My main issue is that some of the models are confusingly presented and would benefit from some reframing. The role of Pch-2 across organisms has been difficult to determine, the ability to separate pairing and synapsis roles in worms provides a great advantage for this paper. 

Strengths: 

Beautiful genetic data, clearly made figures. Great system for studying the role of Pch-2 in crossing over. 

We thank the reviewers for their constructive and useful summary of our manuscript and the analysis of its strengths. 

Weaknesses: 

(1) For a general audience, definitions of crossover assurance, crossover eligible intermediates, and crossover designation would be helpful. This applies to both the proposed molecular model and the cytological manifestation that is being scored specifically in C. elegans. 

We will make these changes in an updated preprint.

(2) Line 62: Is there evidence that DSBs are introduced gradually throughout the early prophase? Please provide references. 

We will reference Woglar and Villeneuve 2018 and Joshi et. al. 2015 to support this statement in the updated preprint.

(3) Do double crossovers show strong interference in worms? Given that the PC is at the ends of chromosomes don't you expect double crossovers to be near the chromosome ends and thus the PC? 

Despite their rarity, double crossovers do show interference in worms. However, the PC is limited to one end of the chromosome. Therefore, even if interference ensures the spacing of these double crossovers, the preponderance of one of these crossovers toward one end (and not both ends) suggest something functionally unique about the PC end.

(4) Line 155 - if the previous data in Deshong et al is helpful it would be useful to briefly describe it and how the experimental caveats led to misinterpretation (or state that further investigation suggests a different model etc.). Many readers are unlikely to look up the paper to find out what this means. 

We will add this to the updated preprint.

(5) Line 248: I am confused by the meaning of crossover assurance here - you see no difference in the average number of COSA-1 foci in Pch-2 vs. wt at any time point. Is it the increase in cells with >6 COSA-1 foci that shows a loss of crossover assurance? That is the only thing that shows a significant difference (at the one time point) in COSA-1 foci. The number of dapi bodies shows the loss of Pch-2 increases crossover assurance (fewer cells with unattached homologs). So this part is confusing to me. How does reliably detecting foci vs. DAPI bodies explain this? 

We apologize for the confusion and will make this more clear in an updated perprint. The reviewer is correct that we do not see a difference in the average number of GFP::COSA1 foci at all time points in this experiment, even though we do see a difference in the number of DAPI stained bodies (an increase in crossover assurance in pch-2 mutants). What we meant to convey is that because of PCH-2’s dual role in regulating crossover formation (inhibiting it in early prophase, guaranteeing assurance later), the average number of GFP::COSA-1 foci at all time points also reflects this later role, resulting in this average being lower than if PCH-2 only inhibited crossovers early in meiotic prophase. We have shown that this later role does not significantly affect the average number of DAPI stained bodies, allowing us to see the role of PCH-2 in early meiotic prophase on crossover formation more clearly.

(6) Line 384: I am confused. I understand that in the dsb-2/pch2 mutant there are fewer COSA-1 foci. So fewer crossovers are designated when DSBs are reduced in the absence of PCH-2.

How then does this suggest that PCH-2's presence on the SC prevents crossover designation? Its absence is preventing crossover designation at least in the dsb-2 mutant. 

We will also make this more clear in an updated preprint, as well as provide additional evidence to support this claim. In this experiment, we had identified three possible explanations for why PCH-2 persists on some nuclei that do not have GFP::COSA-1 foci: 1) PCH-2 removal is coincident with crossover designation; 2) PCH-2 removal depends on crossover designation; and 3) PCH-2 removal facilitates crossover designation. The decrease in the number of GFP::COSA-1 foci in dsb-2::AID;pch-2 mutants argues against the first two possibilities, suggesting that the third might be correct. We have additional evidence that we will include in an updated preprint that should provide stronger support and make this more clear.

(7) Discussion Line 535: How do you know that the crossovers that form near the PCs are Class II and not the other way around? Perhaps early forming Class I crossovers give time for a second Class II crossover to form. In budding yeast, it is thought that synapsis initiation sites are likely sites of crossover designation and class I crossing over. Also, the precursors that form class I and II crossovers may be the same or highly similar to each other, such that Pch-2's actions could equally affect both pathways. 

We do not know that the crossovers that form near the PC are Class II but hypothesize that they are based on the close, functional relationship that exists between Class I crossovers and synapsis and the apparent antagonistic relationship that exists between Class II crossovers and synapsis. We agree that Class I and Class II crossover precursors are likely to be the same or highly similar, exhibit extensive crosstalk that may complicate straightforward analysis and PCH-2 is likely to affect both, as strongly suggested by our GFP::MSH-5 analysis. We present this hypothesis based on the apparent relationship between PCH-2 and synapsis in several systems but agree that it needs to be formally tested. We will make this argument more clear in an updated preprint.

Reviewer #3 (Public review): 

Summary: 

This manuscript describes an in-depth analysis of the effect of the AAA+ ATPase PCH-2 on meiotic crossover formation in C. elegant. The authors reach several conclusions, and attempt to synthesize a 'universal' framework for the role of this factor in eukaryotic meiosis. 

Strengths: 

The manuscript makes use of the advantages of the 'conveyor' belt system within the c.elegans reproductive tract, to enable a series of elegant genetic experiments. 

We thank this reviewer for the useful assessment of our manuscript and the articulation of its strengths.

Weaknesses: 

A weakness of this manuscript is that it heavily relies on certain genetic/cell biological assays that can report on distinct crossover outcomes, without clear and directed control over other aspects and variables that might also impact the final repair outcome. Such assays are currently out of reach in this model system. 

In general, this manuscript could be more generally accessible to non-C.elegans readers. Currently, the manuscript is hard to digest for non-experts (even if meiosis researchers). In addition, the authors should be careful to consider alternative explanations for certain results. At several steps in the manuscript, results could ostensibly be caused by underlying defects that are currently unknown (for example, can we know for sure that pch-2 mutants do not suffer from altered DSB patterning, and how can we know what the exact functional and genetic interactions between pch-2 and HORMAD mutants tell us?). Alternative explanations are possible and it would serve the reader well to explicitly name and explain these options throughout the manuscript. 

We will make the manuscript more accessible to non-C. elegans readers and discuss alternate explanations for specific results in an updated preprint.

Author response:

Reviewer 1:

There are no significant weaknesses to signal in the manuscript. However, in order to fully conclude that there is no obvious advantage for the linguistic dimension in neonates, it would have been most useful to test a third condition in which the two dimensions were pitted against each other, that is, in which they provide conflicting information as to the boundaries of the words comprised in the artificial language. This last condition would have allowed us to determine whether statistical learning weighs linguistic and non-linguistic features equally, or whether phonetic content is preferentially processed.

We appreciate the reviewers' suggestion that a stream with conflicting information would provide valuable insights. In the present study, we started with a simpler case involving two orthogonal features (i.e., phonemes and voices), with one feature being informative and the other uninformative, and we found similar learning capacities for both. Future work should explore whether infants—and humans more broadly—can simultaneously track regularities in multiple speech features. However, creating a stream with two conflicting statistical structures is challenging. To use neural entrainment, the two features must lead to segmentation at different chunk sizes so that their effects lead to changes in power/PLV at different frequencies—for instance, using duplets for the voice dimension and triplets for the linguistic dimension  (or vice versa). Consequently, the two dimensions would not be directly comparable within the same participant in terms of the number of distinguishable syllables/voices, memory demand, or SNR given the 1/F decrease in amplitude of background EEG activity. This would involve comparisons between two distinct groups counter-balancing chunk size and linguistic non-linguistic dimension. Considering the test phase, words for one dimension would have been part-words for the other dimension. As we are measuring differences and not preferences, interpreting the results would also have been difficult. Additionally, it may be difficult to find a sufficient number of clearly discriminable voices for such a design (triplets imply 12 voices). Therefore, an entirely different experimental paradigm would need to be developed.

If such a design were tested, one possibility is that the regularities for the two dimensions are calculated in parallel, in line with the idea that the calculation of statistical regularities is a ubiquitous implicit mechanism (see Benjamin et al., 2024, for a proposed neural mechanism). Yet, similar to our present study, possibly only phonetic features would be used as word candidates. Another possibility is that only one informative feature would be explicitly processed at a time due to the serial nature of perceptual awareness, which may prioritise one feature over the other.

Note: The reviewer’s summary contains a typo: syllabic rate (4 Hz) –not 2 Hz, and word rate (2 Hz) –not 4 Hz.

Reviewer 2:

N400: I am skeptical regarding the interpretation of the phoneme-specific ERP effect as a precursor of the N400 and would suggest toning it down. While the authors are correct in that infant ERP components are typically slower and more posterior compared to adult components, and the observed pattern is hence consistent with an adult N400, at the same time, it could also be a lot of other things. On a functional level, I can't follow the author's argument as to why a violation in phoneme regularity should elicit an N400, since there is no evidence for any semantic processing involved. In sum, I think there is just not enough evidence from the present paradigm to confidently call it an N400.

The reviewer is correct that we cannot definitively determine the type of processing reflected by the ERP component that appears when neonates hear a triplet after exposure to a stream with phonetic regularities. We interpreted this component as a precursor to the N400, based on prior findings in speech segmentation tasks without semantic content, where a ~400 ms component emerged when adult participants recognised pseudowords (Sander et al., 2002) or during structured streams of syllables (Cunillera et al., 2006, 2009). Additionally, the component we observed had a similar topography and timing to those labelled as N400 in infant studies, where semantic processing was involved (Parise et al., 2010; Friedrich & Friederici, 2011).

Given our experimental design, the difference we observed must be related to the type of regularity during familiarisation (either phonemes or voices). Thus, we interpreted this component as reflecting lexical search— a process which could be triggered by a linguistic structure but which would not be relevant to a non-linguistic regularity such as voices. However, we are open to alternative interpretations. In any case, this difference between the two streams reveals that computing regularities based on phonemes versus voices does not lead to the same processes. We will revise and tone down the corresponding part of the discussion to clarify that it is just a possible interpretation of the results.  

Female and male voices: Why did the authors choose to include male and female voices? While using both female and male stimuli of course leads to a higher generalizability, it also introduces a second dimension for one feature that is not present for this other (i.e., phoneme for Experiment 1 and voice identity plus gender for Experiment 2). Hence, couldn't it also be that the infants extracted the regularity with which one gender voice followed the other? For instance, in List B, in the words, one gender is always followed by the other (M-F or F-M), while in 2/3 of the part-words, the gender is repeated (F-F and M-M). Wouldn't you expect the same pattern of results if infants learned regularities based on gender rather than identity?

We used three female and three male voices to maximise acoustic variability. The streams were synthesised using MBROLA, which provides a limited set of artificial voices. Indeed, there were not enough French voices of acceptable quality, so we also used two Italian voices (the phonemes used existed in both Italian and French).

Voices differ in timbre, and female voices tend to be higher pitched. However, it is sometimes difficult to categorise low-pitched female voices and high-pitched male voices. Given that gender may be an important factor in infants' speech perception (newborns, for instance, prefer female voices at birth), we conducted tests to assess whether this dimension could have influenced our results.  

We first quantified the transitional probabilities matrices during the structured stream of Experiment 2, considering that there are only two types of voices: Female and Male.  

For List A, all transition probabilities are equal to 0.5 (P(M|F), P(F|M), P(M|M), P(F|F)), resulting in flat TPs throughout the stream (see Author response image 1, top). Therefore, we would not expect neural entrainment at the word rate (2 Hz), nor would we anticipate ERP differences between the presented duplets in the test phase.

For List B, P(M|F)=P(F|M)=0.66 while P(M|M)=P(F|F)=0.33. However, this does not produce a regular pattern of TP drops throughout the stream (see Author response image 1, bottom). As a result, strong neural entrainment at 2 Hz was unlikely, although some degree of entrainment might have occasionally occurred due to some drops occurring at a 2 Hz frequency. Regarding the test phase, all three Words and only one Part-word presented alternating patterns (TP=0.6). Therefore, the difference in the ERPs between Words and Partwords in List B might be attributed to gender alternation.  

However, it seems unlikely that gender alternation alone explains the entire pattern of results, as the effect is inconsistent and appears in only one of the lists. To rule out this possibility, we analysed the effects in each list separately.

Author response image 1.

Transition probabilities (TPs) across the structured stream in Experiment 2, considering voices processed by gender (Female or Male). Top: List A. Bottom: List B.

We computed the mean activation within the time windows and electrodes of interest and compared the effects of word type and list using a two-way ANOVA. For the difference between Words and Part-words over the positive cluster, we observed a main effect of word type (F(1,31) = 5.902, p = 0.021), with no effects of list or interactions (p > 0.1). Over the negative cluster, we again observed a main effect of word type (F(1,31) = 10.916, p = 0.0016), with no effects of list or interactions (p > 0.1). See Author response image 2.  

Author response image 2.

Difference in ERP voltage (Words – Part-words) for the two lists (A and B); W=Words; P=Part-Words, 

We conducted a similar analysis for neural entrainment during the structured stream on voices. A comparison of entrainment at 2 Hz between participants who completed List A and List B showed no significant differences (t(30) = -0.27, p = 0.79). A test against zero for each list indicated significant entrainment in both cases (List A: t(17) = 4.44, p = 0.00036; List B: t(13) = 3.16, p = 0.0075). See Author response image 3.

Author response image 3.

Neural entrainment at 2Hz during the structured stream of Experiment 2 for Lists A and B.

Words entrainment over occipital electrodes: Do you have any idea why the duplet entrainment effect occurs over the electrodes it does, in particular over the occipital electrodes (which seems a bit unintuitive given that this is a purely auditory experiment with sleeping neonates).

Neural entrainment might be considered as a succession of evoked response induced by the stream. After applying an average reference in high-density EEG recordings, the auditory ERP in neonates typically consists of a central positivity and a posterior negativity with a source located at the electrical zero in a single-dipole model (i.e. approximately in the superior temporal region (Dehaene-Lambertz & Dehaene, 1994). In adults, because of the average reference (i.e. the sum of voltages is equal to zero at each time point) and because the electrodes cannot capture the negative pole of the auditory response, the negativity is distributed around the head. In infants, however, the brain is higher within the skull, allowing for a more accurate recording of the negative pole of the auditory ERP (see Author response image 4 for the location of electrodes in an infant head model).  

Besides the posterior electrodes, we can see some entrainment on more anterior electrodes that probably corresponds to the positive pole of the auditory ERP.

Author response image 4.

International 10–20 sensors' location on the skull of an infant template, with the underlying 3-D reconstruction of the grey-white matter interface and projection of each electrode to the cortex. Computed across 16 infants (from Kabdebon et al, Neuroimage, 2014). The O1, O2, T5, and T6 electrodes project lower than in adults.

Reviewer 3:

(1) While it's true that voice is not essential for language (i.e., sign languages are implemented over gestures; the use of voices to produce non-linguistic sounds, like laughter), it is a feature of spoken languages. Thus I'm not sure if we can really consider this study as a comparison between linguistic and non-linguistic dimensions. In turn, I'm not sure that these results show that statistical learning at birth operates on non-linguistic features, being voices a linguistic dimension at least in spoken languages. I'd like to hear the authors' opinions on this.

On one hand, it has been shown that statistical learning (SL) operates across multiple modalities and domains in human adults and animals. On the other hand, SL is considered essential for infants to begin parsing speech. Therefore, we aimed to investigate whether SL capacities at birth are more effective on linguistic dimensions of speech, potentially as a way to promote language learning.

We agree with the reviewer that voices play an important role in communication (e.g., for identifying who is speaking); however, they do not contribute to language structure or meaning, and listeners are expected to normalize across voices to accurately perceive phonemes and words. Thus, voices are speech features but not linguistic features. Additionally, in natural speech, there are no abrupt voice changes within a word as in our experiment; instead, voice changes typically occur on a longer timescale and involve only a limited number of voices, such as in a dialogue. Therefore, computing regularities based on voice changes would not be useful in real-life language learning. We considered that contrasting syllables and voices was an elegant way to test SL beyond its linguistic dimension, as the experimental paradigm is identical in both experiments.  

Along the same line, in the Discussion section, the present results are interpreted within a theoretical framework showing statistical learning in auditory non-linguistic (string of tones, music) and visual domains as well as visual and other animal species. I'm not sure if that theoretical framework is the right fit for the present results.

(2) I'm not sure whether the fact that we see parallel and independent tracking of statistics in the two dimensions of speech at birth indicates that newborns would be able to do so in all the other dimensions of the speech. If so, what other dimensions are the authors referring to?

The reviewer is correct that demonstrating the universality of SL requires testing additional modalities and acoustic dimensions. However, we postulate that SL is grounded in a basic mechanism of long-term associative learning, as proposed in Benjamin et al. (2024), which relies on a slow decay in the representation of a given event. This simple mechanism, capable of operating on any representational output, accounts for many types of sequence learning reported in the literature (Benjamin et al., in preparation). We will revise the discussion section to clarify this theoretical framework.

(3) Lines 341-345: Statistical learning is an evolutionary ancient learning mechanism but I do not think that the present results are showing it. This is a study on human neonates and adults, there are no other animal species involved therefore I do not see a connection with the evolutionary history of statistical learning. It would be much more interesting to make claims on the ontogeny (rather than philogeny) of statistical learning, and what regularities newborns are able to detect right after birth. I believe that this is one of the strengths of this work.

We did not intend to make claims about the phylogeny of SL. Since SL appears to be a learning mechanism shared across species, we use it as a framework to suggest that SL may arise from general operational principles applicable to diverse neural networks. Thus, while it is highly useful for language acquisition, it is not specific to it. We will revise this section to tone down our claims.  

(4) The description of the stimuli in Lines 110-113 is a bit confusing. In Experiment 1, e.g., "pe" and "tu" are both uttered by the same voice, correct? ("random voice each time" is confusing). Whereas in Experiment 2, e.g., "pe" and "tu" are uttered by different voices, for example, "pe" by yellow voice and "tu" by red voice. If this is correct, then I recommend the authors to rephrase this section to make it more clear.

To clarify, in Experiment 1, the voices were randomly assigned to each syllable, with the constraint that no voice was repeated consecutively. This means that syllables within the same word were spoken by different voices, and each syllable was heard with various voices throughout the stream. As a result, neonates had to retrieve the words based solely on syllabic patterns, without relying on consistent voice associations or specific voice relationships.

In Experiment 2, the design was orthogonal: while the syllables were presented in a random order, the voices followed a structured pattern. Similar to Experiment 1, each syllable (e.g., “pe” and “tu”) was spoken by different voices. The key difference is that in Experiment 2, the structured regularities were applied to the voices rather than the syllables. In other words, the “green” voice was always followed by the “red” voice for example but uttered different syllables.

We will revise the methods section to clarify these important points.

(5) Line 114: the sentence "they should compute a 36 x 36 TPs matrix relating each acoustic signal, with TPs alternating between 1/6 within words and 1/12 between words" is confusing as it seems like there are different acoustic signals. Can the authors clarify this point?

Thank you for highlighting this point. To clarify, our suggestion is that neonates might not track regularities between phonemes and voices as separate features. Instead, they may treat each syllable-voice combination as a distinct item—for example, "pe" spoken by the "yellow" voice is one item, while "pe" spoken by the "red" voice is another. Under this scenario, there would be a total of 36 unique items (6 syllables × 6 voices), and infants would need to track regularities between these 36 combinations.

We will rephrase this sentence in the manuscript to make it clearer.

Author response:

The following is the authors’ response to the original reviews.

A summary of changes

(1) Line 93: “positive effect” to “positive contribution”, as suggested by reviewer 2.

(2) Line 147-148: the null hypothesis to test “equal interspecific and intraspecific interactions”, as indicated by reviewers 2 and 4.

(3) Lines 155-162: removed to reduce duplication with the additive partitioning, as suggested by reviewer 2.

(4) Lines 186-188: added “the estimated competitive growth response would also include the effects of density-dependent pests, pathogens, or microclimates”, as suggested by reviewer 3.  

(5) Lines 219-222: added “The community positive effect can be further partitioned by mechanisms of positive interactions (resource partitioning and facilitation), and facilitative effect can be classified as mutualism (+/+), commensalism (+/0), or parasitic (+/–) based on species specific assessments”.  

(6) Lines 377-386: added options for determining maximum competitive growth response in some extreme scenarios of species mixtures.

(7) Figure 1: modified to show the variations of competitive growth response with relative competitive ability from minimum (null expectation) to maximum (competitive exclusion).    

A summary of four reviewers’ questions and authors’ response

(1) A summary of authors’ responses. Reviewers did not seem to understand our work. They indicated that our model is inadequate for hypothesis testing. The fact is, as we note below, that our model allows for more hypothesis testing than the additive partitioning model. They suggested that one of our model components, the competitive growth response, needs to be further partitioned. However, this term represents only the competition effect and can not be split any further. Reviewers criticized us for misunderstanding the additive components while they suggested the same logic to test some intuitive ideas. They did not seem to know that the effects of competitive interactions vary with assessment methods, which differ between competition and biodiversity research. Our work seeks to harmonise definitions between these two fields and bridge the gap. The reviewers acknowledged that the additive components (i.e., the selection effect and complementarity effect) do not have clear biological meanings; however, they did not acknowledge that the additive components are used extensively for determining mechanisms of species interactions in biodiversity research. There is hardly any research that uses the additive partitioning model without linking the additive components to specific mechanisms of species interactions (i.e., positive SE to competition and positive CE to positive interactions).

(2) Additive partitioning and underlying mechanisms. Some reviewers acknowledged that additive partitioning is not meant for determining mechanisms of species interactions and therefore argued that the additive partitioning should not be criticized for lack of biological meanings with the additive components. However, they insisted that additive partitioning is useful in quantifying net biodiversity effects against the null hypothesis that there is no difference between intraspecific and interspecific interactions or testing the idea that “niche complementarity mitigates competition” or “competitively superior species dominate mixtures”. Are these views contradictory each other? How can the additive partitioning that is not designed for determining mechanisms of species interactions provide meaningful explanations for outputs of species interactions, e.g., “niche complementarity mitigates competition” or “competitively superior species dominate mixtures”?

Reviewers did not seem to realize that these ideas are equivalent to the suggestions that CE represents for the effects of positive interactions and SE for the effects of competitive interactions, that the quantification of net biodiversity effects does not require the two additive components, and that the null hypothesis exists long before the additive partitioning (see de Wit, 1960, de Wit et al., 1966). It is generally agreed that CE and SE result from mathematical calculations and do not have clear biological meanings in terms of linkages to specific mechanisms of species interactions responsible for observed net biodiversity effects or changes in ecosystem function (Loreau and Hector, 2012; Bourrat et al., 2023). Calling some mixed effects of species interactions as mechanisms (e.g., CE and SE) is misleading.        

Model structure: incomplete or inadequate for hypothesis testing. Other than positive, negative, and competition interactions, two reviewers wanted to have more specific interactions such as microclimate amelioration and negative feedback from species-specific pests and pathogens. The determination of these specific mechanisms requires more investigations and cannot be simply made through partitioning growth and yield data. However, the effects of these interactions will be captured in our definition of species interactions.  Reviewers did not seem to know that the additive partitioning would also not allow identifying these specific positive species interactions.

Inspired by the mathematical form of additive partitioning, two reviewers suggested that our model (presumably equation 4) is incomplete and the second term, i.e., competitive growth response needs to be further explored or partitioned. The second term represents deviations from the null expectation, due to species differences in growth and competitive ability or competition effect. We do not know why and how this term can be further partitioned and what any subcomponents would mean.   

Our competitive partitioning model is based on two hypotheses: first, the null hypothesis to test the equivalence of interspecific and intraspecific interactions. This hypothesis is the same as the additive partitioning model. Second, the competitive hypothesis, which tests the dominance of positive or negative species interactions in a community. Thus, our model allows for more hypothesis testing than the current additive partitioning model.     

(3) Types of species interactions. We follow the definition of species interactions generally used in biodiversity research (see Loreau and Hector, 2001), i.e., positive interactions (or complementarity) include resource partitioning and facilitation, negative interactions include interference competition, and competitive interactions include resource competition. One reviewer suggested that resource partitioning is byproduct of competition and should not be part of positive species interactions, which may be true for long-term evolution of species co-existence but not for biodiversity experiments of decade duration at most. Two reviewers suggested that positive interactions should also include microclimate amelioration or negative feedback from species-specific pests and pathogens. We agree and these are included in our definition. 

(4) Significance of partial density monocultures. We used partial and full density monocultures and species competitive ability to determine what species can possibly achieve in mixture under the competitive hypothesis that constituent species share an identical niche but differ in growth and competitive ability. We did not use partial monocultures to test the effects of density on biodiversity effects. As with the additive partitioning, the competitive partitioning model is not designed for comparing yields across different densities. We added at lines 186-188 to indicate that the estimated competitive growth response would also include the effects of density-dependent pests, pathogens, or microclimates.  

Similarly, we do not use the partial density monoculture to  supplant the replacement series design. Partial density monocultures only supplement the “replacement series” design that does not provides estimates of facilitative effects and competitive growth responses that would occur in mixtures. It is crucial to know that one experimental approach is simply not enough for determining underlying mechanisms of species interactions responsible for changes in ecosystem function.  

(5) Competition effect in competition and biodiversity research. Due to different methods used, competition effect in competition research has different ecological meanings from that in biodiversity research. In competition research, species performance in mixture are compared with their partial density monocultures and therefore competition effect is generally negative, as suggested by reviewer 4. In biodiversity research, comparison is between mixture and full density monocultures. The resulting competition effect can be positive or negative for both individual species and community productivity defined by species composition and full density monoculture yields.     

Therefore, we cannot use the results of competition research based on additive series design to describe effects of competitive interactions on ecosystem productivity based replacement series design.

Reviewer #1 (Public Review):

[Editors' note: this is an overall synthesis from the Reviewing Editor in consultation with the reviewers.]

The three reviews expand our critique of this manuscript in some depth and complementary directions. These can be synthesized in the following main points (we point out that there is quite a bit more that could be written about the flaws with this study; however, time constraints prevented us from further elaborating on the issues we see):

(1) It is unclear what the authors want to do.

As indicate by the title, our objective is to “partition changes in ecosystem productivity by effects of species interactions”, i.e., partitioning net biodiversity effects estimated from the null expectation into components associated with positive, negative, or competition interspecific interactions.

It seems their main point is that the large BEF literature and especially biodiversity experiments overstate the occurrence of positive biodiversity effects because some of these can result from competition.

We demonstrated through ecological theories and simulation/experiment data that competition is a major source of the net biodiversity effects estimated with additive partitioning model. We know that competition effect varies with mixture attributes. Future research will determine average effect of competitive interactions on biodiversity effects in large BEF literature.   

Because reduced interspecific relative to intraspecific competition in mixture is sufficient to produce positive effects in mixtures (if interspecific competition = 0 then RYT = S, where S is species richness in mixture -- this according to the reciprocal yield law = law of constant final yield), they have a problem accepting NE > 0 as true biodiversity effect (see additive partitioning method of Loreau & Hector 2001 cited in manuscript).

We have no problem to accept NE>0 as true positive biodiversity effect. However, NE>0 can also result from competitive interactions based on the null expectation and needs to be partitioned by effects of species interactions.

(2) The authors' next claim, without justification, that additive partitioning of NE is flawed and theoretically and biologically meaningless.

The additive partitioning model is based on Covariance equation (or Price equation) that has nothing to do with biodiversity partitioning (Bourrat et al., 2023). Biological meaning was arbitrarily assigned to CE and SE. We made clear that the additive partitioning model is mathematically sound but does not have biological meanings that it has been used for.   

They misinterpret the CE component as biological niche partitioning and the SE component as biological dominance.

We did not. Loreau and Hector (2001) clearly indicated positive CE for positive interactions and positive SE for competitive interactions, which is generally what has been used for in the last twenty years.

They do not seem to accept that the additive partitioning is a logically and mathematically sound derivation from basic principles that cannot be contested.

We do not have problem with mathematical form of additive partitioning but only oppose ecological meanings assigned to CE and SE, simply because CE and SE both result from all species interactions (see Loreau and Hector, 2001; Bourrat et al., 2023). The reviewer seemed to have a contradictory thinking that the additive components are biologically meaningless but derived from biological basic principles.       

(3) The authors go on to introduce a method to calculate species-level overyielding (RY > 1/S in replacement series experiments) as a competitive growth response and multiply this with the species monoculture biomass relative to the maximum to obtain competitive expectation. This method is based on resource competition and the idea that resource uptake is fully converted into biomass (instead of e.g. investing it in allelopathic chemical production).

Correct, but we did not assume “resource uptake is fully converted into biomass”.

(4) It is unclear which experiments should be done, i.e. are partial-density monocultures planted or simply calculated from full-density monocultures? At what time are monocultures evaluated? The framework suggests that monocultures must have the full potential to develop, but in experiments, they are often performing very poorly, at least after some time. I assume in such cases the monocultures could not be used.

Both partial and full density monocultures are needed, along with mixtures to separate NE by species interactions. Calculating competitive growth responses from density-size relationships can be an alternative, given the lack of partial density monocultures in current biodiversity experiments, but is not preferred.

Similar to additive partitioning, our model can (and should) be applied to all developmental stages of an experiment to examine how interactions evolve through time.   

(5) There are many reasons why the ideal case of only resource competition playing a role is unrealistic. This excludes enemies but also differential conversion factors of resources into biomass and antagonistic or facilitative effects. Because there are so many potential reasons for deviations from the null model of only resource competition, a deviation from the null model does not allow conclusions about underlying mechanisms.

The competitive expectation is only a hypothesis, just as the null expectation. The difference between competitive and null expectations represents a competitive effect resulting from species differences in growth and competitive ability, while the deviation of observed yields from the competitive expectation indicates positive or negative effect (see lines 201-219).

Furthermore, this is not a systematically developed partitioning, but some rather empirical ad hoc formulation of a first term that is thought to approximate competitive effects as understood by the authors (but again, there already are problems here). The second residual term is not investigated. For a proper partitioning approach, one would have to decompose overyielding into two (or more) terms and demonstrate (algebraically) that under some reasonable definitions of competitive and non-competitive interactions, these end up driving the respective terms.

The first term represents the null expectation assuming equal interspecific and intraspecific interactions, i.e., absence of positive, negative, and competition effects. The second residual term represents competition effect, due to species differences in growth and competitive ability. The meaning of second residual term is clear and does not need to be further partitioned or investigated.

In fact, our competitive partitioning also has several components including null expectation, competitive growth response, and observed yield, plus partial density monocultures for species assessment, or null expectations, competitive expectations, and observed yields for community level assessment, although different from the additive partitioning.

(6) Using a simplistic simulation to test the method is insufficient. For example, I do not see how the simulation includes a mechanism that could create CE in additive partitioning if all species would have the same monoculture yield. Similarly, they do not include mechanisms of enemies or antagonistic interactions (e.g. allelopathy).

The simulation model we used is developed from real world data and can only do what are available in the model in terms of species and their growth under different conditions. We can not go beyond data limitation. The model is empirical and has been shown to accurately estimate yield in the aspen-spruce forest condition. We would also note that we do also use experimental data (Table 2).  

(7) The authors do not cite relevant literature regarding density x biodiversity experiments, competition experiments, replacement-series experiments, density-yield experiments, additive partitioning, facilitation, and so on.

We cited literature relevant to biodiversity partitioning since we are not aiming to cover everything. The reviewer may not be aware that most of the research areas listed are actually included in our work, such as additive and replacement-series experiment designs, additive partitioning, facilitation, competition studies, and density-yield relationships. Our competitive model partitioning is based on biological principles, while the additive partitioning model is based only on a mathematical equation.   

Overall, this manuscript does not lead further from what we have already elaborated in the broad field of BEF and competition studies and rather blurs our understanding of the topic.

The results of competition studies based on additive series design are not really used in the broad field of BEF based on replacement series design. The effects of competitive interactions on BEF are never clearly defined using the results of competition studies. Our work is filling that gap.  

Reviewer #2 (Public Review):

This manuscript is motivated by the question of what mechanisms cause overyielding in mixed-species communities relative to the corresponding monocultures. This is an important and timely question, given that the ultimate biological reasons for such biodiversity effects are not fully understood.

As a starting point, the authors discuss the so-called "additive partitioning" (AP) method proposed by Loreau & Hector in 2001. The AP is the result of a mathematical rearrangement of the definition of overyielding, written in terms of relative yields (RY) of species in mixtures relative to monocultures. One term, the so-called complementarity effect (CE), is proportional to the average RY deviations from the null expectations that plants of both species "do the same" in monocultures and mixtures. The other term, the selection effect (SE), captures how these RY deviations are related to monoculture productivity. Overall, CE measures whether relative biomass gains differ from zero when averaged across all community members, and SE, whether the "relative advantage" species have in the mixture, is related to their productivity. In extreme cases, when all species benefit, CE becomes positive. When large species have large relative productivity increases, SE becomes positive. This is intuitively compatible with the idea that niche complementarity mitigates competition (CE>0), or that competitively superior species dominate mixtures and thereby driver overyielding (SE>0).

The reviewer needs to know that these ideas are based on the same logic that positive CE represents the effects of positive interactions and positive SE represents the effects of competitive interactions. CE>0 or SE>0 can result from many different scenarios of species interactions, not necessarily “niche complementarity mitigates competition” or “competitively superior species dominate mixtures”. CE>0 and SE>0 can occur alone or together. We simply can not tell underlying mechanisms of overyielding from mathematical calculations (CE and SE), as suggested by this reviewer later.

The reviewer criticizes us while using the same logic themselves.

However, it is very important to understand that CE and SE capture the "statistical structure" of RY that underlies overyielding. Specifically, CE and SE are not the ultimate biological mechanisms that drive overyielding, and never were meant to be. CE also does not describe niche complementarity. Interpreting CE and SE as directly quantifying niche complementarity or resource competition, is simply wrong, although it sometimes is done. The criticism of the AP method thus in large part seems unwarranted. The alternative methods the authors discuss (lines 108-123) are based on very similar principles.

The reviewer actually supports our point. However, CE and SE have been largely used as biological mechanisms, positive CE as the results of complementary interactions and positive SE as the results of competitive interactions (see Loreau and Hector, 2001).  

We do not have problem with the "statistical structure" of AP; it is simply a covariance equation. It is important to know that CE and SE do not provide additional information on overyielding than NE in terms of underlying mechanisms of species interactions. Any attempt to investigate mechanism of overyielding with CE or SE can easily go wrong.

Our competitive partitioning model incorporates effects of competitive interactions into the conventional null expectation and allows for separating different effects of species interactions. In comparison, the additive partitioning model does not have this capacity, not even designed for this purpose, as suggested by this and other reviewers.         

The authors now set out to develop a method that aims at linking response patterns to "more true" biological mechanisms.

Assuming that "competitive dominance" is key to understanding mixture productivity, because "competitive interactions are the predominant type of interspecific relationships in plants", the authors introduce "partial density" monocultures, i.e. monocultures that have the same planting density for a species as in a mixture. The idea is that using these partial density monocultures as a reference would allow for isolating the effect of competition by the surrounding "species matrix".

Correct.

The authors argue that "To separate effects of competitive interactions from those of other species interactions, we would need the hypothesis that constituent species share an identical niche but differ in growth and competitive ability (i.e., absence of positive/negative interactions)." - I think the term interaction is not correctly used here, because clearly competition is an interaction, but the point made here is that this would be a zero-sum game.

We did not say that competition is not an interaction; we only want to separate the effect of competition from those of other species interactions.

The authors use the ratio of productivity of partial density and full-density monocultures, divided by planting density, as a measure of "competitive growth response" (abbreviated as MG). This is the extra growth a plant individual produces when intraspecific competition is reduced.

Correct.

We added at lines 377-386 to discuss options to determine MG in some uncommon scenarios of species mixtures.

Here, I see two issues: first, this rests on the assumption that there is only "one mode" of competition if two species use the same resources, which may not be true, because intraspecific and interspecific competition may differ. Of course, one can argue that then somehow "niches" are different, but such a niche definition would be very broad and go beyond the "resource set" perspective the authors adopt. Second, this value will heavily depend on timing and the relationship between maximum initial growth rates and competitive abilities at high stand densities.

First, the "competitive effect" focusses on resource competition and other forms of competition (presumably interference competition) are included in the negative interactions.

Second, competitive growth response varies over time and with density, and so do NE, CE, SE, and interspecific interactions.

The authors then progress to define relative competitive ability (RC), and this time simply uses monoculture biomass as a measure of competitive ability. To express this biomass in a standardized way, they express it as different from the mean of the other species and then divide by the maximum monoculture biomass of all species.

I have two concerns here: first, if competitive ability is the capability of a species to preempt resources from a pool also accessed by another species, as the authors argued before, then this seems wrong because one would expect that a species can simply be more productive because it has a broader niche space that it exploits. This contradicts the very narrow perspective on competitive ability the authors have adopted. This also is difficult to reconcile with the idea that specialist species with a narrow niche would outcompete generalist species with a broad niche. Second, I am concerned by the mathematical form. Standardizing by the maximum makes the scaling dependent on a single value.

First, growth conditions are controlled in biodiversity experiments, i.e., both monocultures and mixtures are the same in resource space. Species do not have opportunity to exploit resources outside experimental area. For example, if less productive species on normal soils outperform more competitive species on saline/alkaline soil, these “less productive species” are considered “more productive”.    

Second, as discussed in our paper (lines 367-376; Figure 1), more research is needed to determine relationships between species traits (biomass or height) and relative competitive ability. By then, scaling by the maximum would not be needed. There has been quite a lot of research on such relationships; we should leave this to subject experts to determine what would be mostly appropriate for species studied.

As a final step, the authors calculate a "competitive expectation" for a species' biomass in the mixture, by scaling deviations from the expected yield by the product MG ⨯ RC. This would mean a species does better in a mixture when (1) it benefits most from a conspecific density reduction, and (2) has a relatively high biomass.

Put simply, the assumption would be that if a species is productive in monoculture (high RC), it effectively does not "see" the competitors and then grows like it would be the sole species in the community, i.e. like in the partial density monoculture.

Correct, if species competitive ability differs substantially, the more competitive species in the mixture would grow like partial density monoculture. This extra growth should not be treated as sources of positive biodiversity effects, simply because it does not result from positive species interactions.   

Overall, I am not very convinced by the proposed method.

(1) The proposed method seems not very systematic but rather "ad hoc". It also is much less a partitioning method than the AP method because the other term is simply the difference. It would be good if the authors investigated the mathematical form of this remainder and explored its properties.. when does complementarity occur? Would it capture complementarity and facilitation?

AP is, by no means, systematic. Remember, AP is based on covariance equation (or Price equation) that has nothing to do with species interactions, other than nice-looking mathematical form (Bourrat et al., 2023). Ecological meanings are subjectively given to CE and SE. Therefore,  CE and SE reflect what we call them, not what they really mean.    

The remainder measures deviations from the null expectation, due to only competition effect, and can not be partitioned any further. The remainder would be positive for more competitive species and negative for less competitive species in mixture relative to their full density monoculture. The deviation of observed yields from competitive expectations indicates dominance of positive or negative species interactions. All these are clearly outlined at lines 201-221.   

(2) The justification for the calculation of MG and RC does not seem to follow the very strict assumptions of what competition (in the absence of complementarity) is. See my specific comments above.

We do not see why not.

(3) Overall, the manuscript is hard to read. This is in part a problem of terminology and presentation, and it would be good to use more systematic terms for "response patterns" and "biological mechanisms".

To help understand the variations of competitive growth response with relative competitive ability, the x axis of Figure 1 is labelled with null expectation, competitive expectation, and competitive exclusion from minimum to maximum deviation of competitive ability from community average.

We have followed terms used in biodiversity partitioning and changing terms can be confusing.  

Examples:

- on line 30, the authors write that CE is used to measure "positive" interactions and SE to measure "competitive interactions", and later name "positive" and "negative" interactions "mechanisms of species interactions". Here the authors first use "positive interaction" as any type of effect that results in a community-level biomass gain, but then they use "interaction" with reference to specific biological mechanisms (e.g. one species might attract a parasite that infests another species, which in turn may cause further changes that modify the growth of the first and other species).

There are some differences in meaning, but that is what CE and SE have been generally used for. Using different terms can be confusing and does not help understanding the problems with AP.

- on line 70, the authors state that "positive interaction" increases productivity relative to the null expectation, but it is clear that an interaction can have "negative" consequences for one interaction partner and "positive" ones for the other. Therefore, "positive" and "negative" interactions, when defined in this way, cannot be directly linked to "resource partitioning" and "facilitation", and "species interference" as the authors do. Also, these categories of mechanisms are still simple. For example, how do biotic interactions with enemies classify, see above?

We are explaining effects of competitive interactions on species yield, and ultimately on community yield that can be linked to “resource partitioning" and "facilitation", and "species interference".

More specific species interactions require detailed biological investigation and cannot be determined through partitioning of biomass production.  

- line 145: "Under the null hypothesis, species in the mixture are assumed to be competitively equivalent (i.e., absence of interspecific interactions)". This is wrong. The assumption is that there are interspecific interactions, but that these are the same as the intraspecific ones. Weirdly, what follows is a description of the AP method, which does not belong here. This paragraph would better be moved to the introduction where the AP method is mentioned. Or omitted, since it is basically a repetition of the original Loreau & Hector paper.

As suggested, “absence of interspecific interactions” was replaced with “equal interspecific and intraspecific interactions”.

We have removed lines 155-162 to reduce duplication. However, our method is based on null expectation that needs to be introduced, despite it is part of AP.

Other points:

- line 66: community productivity, not ecosystem productivity.

Both community productivity and ecosystem productivity are used in biodiversity research, although meaning can be slightly different. Comparatively, ecosystem productivity is more common.

- line 68: community average responses are with respect to relative yields - this is important!

- line 64: what are "species effects of species interactions"?

We searched and did not find “species effects of species interactions”.

- line 90: here "competitive" and "productive" are mixed up, and it is important to state that "suffers more" refers to relative changes, not yield changes.

It, in fact, refers to yield changes. For example, less productive species, at active growth, are more responsive to changes in competition, while more productive species, at inactive growth (i.e., aging), are less responsive to changes in competition.   

- line 92: "positive effect of competitive dominance": I don't understand what is meant here.

The phrase was modified to “positive contribution of competitive dominance to ecosystem productivity based on the null expectation”.

Reviewer #3 (Public Review):

Summary:

This manuscript by Tao et al. reports on an effort to better specify the underlying interactions driving the effects of biodiversity on productivity in biodiversity experiments. The authors are especially concerned with the potential for competitive interactions to drive positive biodiversity-ecosystem functioning relationships by driving down the biomass of subdominant species. The authors suggest a new partitioning schema that utilizes a suite of partial density treatments to capture so-called competitive ability. While I agree with the authors that understanding the underlying drivers of biodiversity-ecosystem functioning relationships is valuable - I am unsure of the added value of this specific approach for several reasons.

Strengths:

I can find a lot of value in endeavouring to improve our understanding of how biodiversity-ecosystem functioning relationships arise. I agree with the authors that competition is not well integrated into the complementarity and selection effect and interrogating this is important.

Weaknesses:

(1) The authors start the introduction very narrowly and do not make clear why it is so important to understand the underlying mechanisms driving biodiversity-ecosystem functioning relationships until the end of the discussion.

There are different ways to start introduction; we believe that starting with the problems of the current approach is the most effective for outlining the study’s objective.  

(2) The authors criticize the existing framework for only incorporating positive interactions but this is an oversimplification of the existing framework in several ways:

We did not criticize the existing framework for only incorporating positive interactions. We criticize the existing framework, because it is not based on mechanisms of species interactions, but is extensively used to determine underlying mechanisms driving biodiversity-ecosystem functioning relationships.

a. The existing partitioning scheme incorporates resource partitioning which is an effect of competition.

Resource partitioning means that species utilize resources differently, while competition means species use the same resources. “resource partitioning is an effect of competition” is not true in biodiversity experiments that are often short in duration and controlled in conditions.  

b. The authors neglect the potential that negative feedback from species-specific pests and pathogens can also drive positive BEF and complementarity effects but is not a positive interaction, necessarily. This is discussed in Schnitzer et al. 2011, Maron et al. 2011, Hendriks et al. 2013, Barry et al. 2019, etc.

We did not. The feedback effect will be reflected in the differences between observed yields and competitive expectations if species in mixtures have different pests and pathogens relative to monocultures. The additive partitioning does not identify these feedback effects either.

c. Hector and Loreau (and many of the other citations listed) do not limit competition to SE because resource partitioning is a byproduct of competition.

Positive SE has been largely interpreted as the result of competition including Hector and Loreau (2001) and many others. It needs to be clear that neither of the additive components can be linked to specific mechanisms of species interactions. 

Does “resource partitioning is a byproduct of competition” mean that species change their niche to avoid competition? If this is what the reviewer means, it may occur through long-term evolution, but not in short-term biodiversity experiments. Hector and Loreau (2001) clearly indicated that their complementarity effect includes both resource partitioning and facilitation.   

(3) It is unclear how this new measure relates to the selection effect, in particular. I would suggest that the authors add a conceptual figure that shows some scenarios in which this metric would give a different answer than the traditional additive partition. The example that the authors use where a dominant species increases in biomass and the amount that it increases in biomass is greater than the amount of loss from it outcompeting a subdominant species is a general example often used for a selection effect when exactly would you see a difference between the two?:<br /> a. Just a note - I do think you should see a difference between the two if the species suffers from strong intraspecific competition and has therefore low monoculture biomass but this would tend to also be a very low-density monoculture in practice so there would potentially be little difference between a low density and high-density monoculture because the individuals in a high-density monoculture would die anyway. So I am not sure that in practice you would really see this difference even if partial density plots were incorporated.

Linking new measure to SE or CE would be difficult (see many comparisons in Tables and Figures in our manuscript), as SE and CE are derived from mathematical equation and do not represent specific mechanisms of species interactions (Hector and Loreau 2012; Bourrat et al., 2023).

(4) One of the tricky things about these endeavors is that they often pull on theory from two different subfields and use similar terminology to refer to different things. For example - in competition theory, facilitation often refers to a positive relative interaction index (this seems to be how the authors are interpreting this) while in the BEF world facilitation often refers to a set of concrete physical mechanisms like microclimate amelioration. The truth is that both of these subfields use net effects. The relative interaction index is also a net outcome as is the complementarity effect even if it is only a piece of the net biodiversity effect. Trying to combine these two subfields to come up with a new partitioning mechanism requires interrogating the underlying assumptions of both subfields which I do not see in this paper.

Agree, microclimate amelioration is also part of positive effect and will be reflected in the difference between observed yield and competitive expectation. We can not separate the two mechanisms of positive species interactions without investigating influences of microclimate on growth and yield.

(5) The partial density treatment does not isolate competition in the way that the authors indicate. All of the interactions that the authors discuss are density-dependent including the mechanism that is not discussed (negative feedback from species-specific pests and pathogens). These partial density treatment effects therefore cannot simply be equated to competition as the authors indicate.:

We use partial density monoculture to determine maximum competitive growth response, effect of density-dependent intraspecific interactions, and species competitive ability to determine the level of maximum competitive growth response species can achieve in mixtures. There may be changes in species-specific pests and pathogens from partial to full density monocultures, which will be captured in competitive growth responses of individuals. We added at lines 186-188 to indicate that the maximum competitive growth response estimated would also include the effects of density-dependent pests, pathogens, or microclimates.   

a. Additionally - the authors use mixture biomass as a stand-in for competitive ability in some cases but mixture biomass could also be determined by the degree to which a plant is facilitated in the mixture (for example).

We used monoculture biomass, not mixture biomass, to assess competitive ability

(6) I found the literature citation to be a bit loose. For example, the authors state that the additive partition is used to separate positive interactions from competition (lines 70-76) and cite many papers but several of these (e.g. Barry et al. 2019) explicitly do not say this.

Barry et al. (2019) defined CE as overproduction from monocultures, an effect of positive interactions.  

(7) The natural take-home message from this study is that it would be valuable for biodiversity experiments to include partial density treatments but I have a hard time seeing this as a valuable addition to the field for two reasons:

a. In practice - adding in partial density treatments would not be feasible for the vast majority of experiments which are already often unfeasibly large to maintain.

The reviewer suggested that quantity is more important than quality. Without partial density monocultures no one can separate different effects of species interactions, as suggested by Loreau and Hector, reviewers, and many others that effects of species interactions can not be clearly differentiated with replacement series design. Unreliable scientific findings are not valuable.

b. The density effect would likely only be valuable during the establishment phase of the experiment because species that are strongly limited by intraspecific competition will die in the full-density plots resulting in low-density monocultures. You can see this in many biodiversity experiments after the first years. Even though they are seeded (or rarely planted) at a certain density, the density after several years in many monocultures is quite low.

True. High or low density also depends on individual size; if individuals do not get enough resources, density is high. Therefore, density effect can be strong even as density drops substantially from initial levels.  

Reviewer #4 (Public Review):

Summary:

This manuscript claims to provide a new null hypothesis for testing the effects of biodiversity on ecosystem functioning. It reports that the strength of biodiversity effects changes when this different null hypothesis is used. This main result is rather inevitable. That is, one expects a different answer when using a different approach. The question then becomes whether the manuscript’s null hypothesis is both new and an improvement on the null hypothesis that has been in use in recent decades.

It needs to be clear that we use two hypotheses, null hypothesis that is currently used with AP, and competitive hypothesis that is new with this manuscript. The null hypothesis helps determine changes in ecosystem productivity from all species interactions, while the competitive hypothesis helps partition changes in ecosystem productivity by mechanisms of species interactions, i.e., positive, negative, or competitive interactions.    

Strengths:

In general, I appreciate studies like this that question whether we have been doing it all wrong and I encourage consideration of new approaches.

Weaknesses:

Despite many sweeping critiques of previous studies and bold claims of novelty made throughout the manuscript, I was unable to find new insights. The manuscript fails to place the study in the context of the long history of literature on competition and biodiversity and ecosystem functioning. The Introduction claims the new approach will address deficiencies of previous approaches, but after reading further I see no evidence that it addresses the limitations of previous approaches noted in the Introduction. Furthermore, the manuscript does not reproducibly describe the methods used to produce the results (e.g., in Table 1) and relies on simulations, claiming experimental data are not available when many experiments have already tested these ideas and not found support for them. Finally, it is unclear to me whether rejecting the ‘new’ null hypothesis presented in the manuscript would be of interest to ecologists, agronomists, conservationists, or others. I will elaborate on each of these points below.

First, there are many biodiversity experiments but those with partial density monocultures are rare. We found only one greenhouse experiment. We have to use simulation to illustrate different scenarios of species interactions to demonstrate how our approach works and how different it is from the AP.  

Because of different methods used, the results of long history competition research (generally based on additive series design) cannot be used to define effects of competitive interactions in biodiversity research (generally based on replacement series design). This may be the reason that few competition researchers were cited in Loreau and Hector (2001).

Our approach requires two hypotheses, null and competitive, and the meaning of deviation from these hypotheses are outlined at lines 201-221 for both individual species and community level assessments. Distinguishing changes in ecosystem productivity by species interactions would be of great interest to “ecologists, agronomists, conservationists, or others”.

The critiques of biodiversity experiments and existing additive partitioning methods are overstated, as is the extent to which this new approach addresses its limitations. For example, the critique that current biodiversity experiments cannot reveal the effects of species interactions (e.g., lines 37-39) isn't generally true, but it could be true if stated more specifically. That is, this statement is incorrect as written because comparisons of mixtures, where there are interspecific and intraspecific interactions, with monocultures, where there are only intraspecific interactions, certainly provide information about the effects of species interactions (interspecific interactions). These biodiversity experiments and existing additive partitioning approaches have limits, of course, for identifying the specific types of interactions (e.g., whether mediated by exploitative resource competition, apparent competition, or other types of interactions). However, the approach proposed in this manuscript gets no closer to identifying these specific mechanisms of species interactions. It has no ability to distinguish between resource and apparent competition, for example. Thus, the motivation and framing of the manuscript do not match what it provides. I believe the entire Introduction would need to be rewritten to clarify what gap in knowledge this proposed approach is addressing and what would be gained by filling this knowledge gap.

Our approach helps determine underlying mechanisms of species interactions, i.e., positive (resources partitioning or facilitation), negative, or competitive interactions. I am not sure how much we need to go further in identifying more specific mechanisms. If resource and apparent competition refers to resource and interference competition, our approach can tease apart them.

I recommend that the Introduction instead clarify how this study builds on and goes beyond many decades of literature considering how competition and biodiversity effects depend on density. This large literature is insufficiently addressed in this manuscript. This fails to give credit to previous studies considering these ideas and makes it unclear how this manuscript goes beyond the many previous related studies. For example, see papers and books written by de Wit, Harper, Vandermeer, Connolly, Schmid, and many others. Also, note that many biodiversity experiments have crossed diversity treatments with a density treatment and found no significant effects of density or interactions between density and diversity (e.g., Finn et al. 2013 Journal of Applied Ecology). Thus, claiming that these considerations of density are novel, without giving credit to the enormous number of previous studies considering this, is insufficient.

A misunderstanding here. Our approach is not designed to test density effect. The same density is held across full density monocultures and mixtures. We use partial density monocultures to determine what species may competitively achieve in full density mixture, without positive or negative interspecific interactions.  

Replacement series designs emerged as a consensus for biodiversity experiments because they directly test a relevant null hypothesis. This is not to say that there are no other interesting null hypotheses or study designs, but one must acknowledge that many designs and analyses of biodiversity experiments have already been considered. For example, Schmid et al. reviewed these designs and analyses two decades ago (2002, chapter 6 in Loreau et al. 2002 OUP book) and the overwhelming consensus in recent decades has been to use a replacement series and test the corresponding null hypothesis.

Some wrong impressions. We are not trying to supplant “replacement series” with “additive series”; we use “additive series” designs to supplement “replacement series” design for partitioning changes in ecosystem productivity by mechanisms of species interactions, which would not be possible with “replacement series” design alone, as suggested by many including reviewers.   

It is unclear to me whether rejecting the 'new' null hypothesis presented in the manuscript would be of interest to ecologists, agronomists, conservationists, or others. Most biodiversity experiments and additive partitions have tested and quantified diversity effects against the null hypothesis that there is no difference between intraspecific and interspecific interactions. If there was no less competition and no more facilitation in mixtures than in monocultures, then there would be no positive diversity effects. Rejecting this null hypothesis is relevant when considering coexistence in ecology, overyielding in agronomy, and the consequences of biodiversity loss in conservation (e.g., Vandermeer 1981 Bioscience, Loreau 2010 Princeton Monograph). This manuscript proposes a different null hypothesis and it is not yet clear to me how it would be relevant to any of these ongoing discussions of changes in biodiversity.

Our method begins with the null expectation: that intraspecific and interspecific interactions are equivalent. We then propose the competitive hypothesis as a second non-exclusive hypothesis which tests the dominance of positive or negative specific interactions. As shown by its name, the additive partitioning model has been advocated for partitioning biodiversity effects by some ecological mechanisms (CE and SE). The ecological meaning of deviation from the two hypotheses are outlined at lines 201-221 for both individual species and community level assessments.   

The claim that all previous methods 'are not capable of quantifying changes in ecosystem productivity by species interactions and species or community level' is incorrect. As noted above, all approaches that compare mixtures, where there are interspecific interactions, to monocultures, where there are no species interactions, do this to some extent. By overstating the limitations of previous approaches, the manuscript fails to clearly identify what unique contribution it is offering, and how this builds on and goes beyond previous work.

The reviewer implies that a partial truth equals the whole truth. The same argument can also be applied to the additive partitioning if relative yield total or response ratio provides a kind of comparison between mixture and monocultures. Our statement is correct in the way that previous approaches are not designed to separate changes in ecosystem productivity by species interactions, as indicated by other reviewers. The additive partitioning is built on Price equation (covariance equation) that has never been biologically demonstrated for relevance in biodiversity partitioning (Bourrat et al., 2023).  

We made clear that our work is built on and beyond the null expectation with addition of competitive expectation.

The manuscript relies on simulations because it claims that current experiments are unable to test this, given that they have replacement series designs (lines 128-131). There are, however, dozens of experiments where the replacement series was repeated at multiple densities, which would allow a direct test of these ideas. In fact, these ideas have already been tested in these experiments and density effects were found to be nonsignificant (e.g., Finn et al. 2013).

Out of point. Again, we are not testing density effect. Partial density is used to determine competitive growth responses that species may achieve in mixture based on their relative competitive ability. We used simulations, as partial density monocultures are used only in one experimental study that has been included in our study.  

It seems that the authors are primarily interested in trees planted at a fixed density, with no opportunity for changes in density, and thus only changes in the size of individuals (e.g., Fig. 1). In natural and experimental systems, realized density differs from the initial planted density, and survivorship of seedlings can depend on both intraspecific and interspecific interactions. Thus, the constrained conditions under which these ideas are explored in this manuscript seem narrow and far from the more complex reality where density is not fixed.

We use fixed density only for convenience. In biodiversity experiments, density can increase or decrease over time from initial levels. However, initial density is generally used in evaluation of species interactions. If interest is community productivity, density change does not need to be considered. Again, we are not testing density effects.    

Additional detailed comments:

It is unclear to me which 'effects' are referred to on line 36. For example, are these diversity effects or just effects of competition? What is the response variable?

It means the effect of competitive interactions on productivity and should be clear based on previous sentences.

The usefulness of the approach is overstated on line 52. All partitioning approaches, including the new one proposed here, give the net result of many types of species interactions and thus cannot 'disentangle underlying mechanisms of species interactions.'

Not sure how many types of species interactions the reviewer referred to. If mechanisms of species interactions are grouped in three categories (positive, negative, and competitive) as has been in biodiversity research, our approach can tease them apart.   

The weaknesses of previous approaches are overstated throughout the manuscript, including in lines 60-61. All approaches provide some, but not all insights. Sweeping statements that previous approaches are not effective, without clarifying what they can and can't do, is unhelpful and incorrect. Also, these statements imply that the approach proposed here addresses the limitations of these previous approaches. I don't yet see how it does so.

The weaknesses of previous approaches are not overstated in terms of separating changes in ecosystem productivity by species interactions. As pointed by other reviewers, none of the previous approaches are designed for quantifying changes in ecosystem productivity by species interactions.   

The definitions given for the CE and SE on line 71 are incorrect. Competition affects both terms and CE can be negative or have nothing to do with positive interactions, as noted in many of the papers cited.

We are not trying to define CE and SE but only point out how CE and SE have been generally used in biodiversity research (see recent publication by Feng et al., 2022).

The proposed approach does not address the limitations noted on lines 73 and 74.

It does in terms of sources of net biodiversity effect, whether from positive, negative or competitive interactions.

The definition of positive interactions in lines 77 and 78 seems inconsistent with much of the literature, which instead focuses on facilitation or mutualism, rather than competition when describing positive interactions.

Much of the literature supports our definition (see Loreau and Hector, 2001). In biodiversity research, positive interactions include resource partitioning and facilitation. What we are trying to point out is that competition affects species and community level assessments based on the null expectation and needs to be separated.

Throughout the manuscript, competition is often used interchangeably with resource competition (e.g., line 82) and complementarity is often attributed to resource partitioning (e.g., line 77). This ignores apparent competition and partitioning enemy-free niche space, which has been found to contribute to biodiversity effects in many studies.

If apparent competition refers to interference competition, it is included in negative interaction. Changes in species-specific pests and pathogens in mixture will be captured in positive or negative effects through facilitation or interference.  

In what sense are competitive interactions positive for competitive species (lines 82-83)? By definition, competition is an interaction that has a negative effect. Do you mean that interspecific competition is less than intraspecific competition? I am having a very difficult time following the logic.

I am glad the reviewer raised this question that may confuse many others and has never been clearly discussed. It all depends on how comparison is made. If species performance in mixture are compared with that in partial density monocultures, as is in competition research, competition effect is negative for all species. If comparison is made between mixture and full density monocultures, as is done in biodiversity research, competition effect should be positive for more competitive species and negative for less competitive species, with resources flowing from less to more competitive species in mixture relative to full density monocultures.   

Therefore, the definitions of competitive interactions based on additive series design in competition research cannot be used to describe competitive interactions based on replacement series design in biodiversity research. In biodiversity research, the effects of competitive interactions are never clearly defined at species or community level and mixed up with those of other species interactions.      

Results are asserted on lines 93-95, but I cannot find the methods that produced these results. I am unable to evaluate the work without a repeatable description of the methods.

We have added references on sources of these data.

The description of the null hypothesis in the common additive partitioning approach on lines 145-146 is incorrect. In the null case, it does not assume that there are no interspecific interactions, but rather that interspecific and intraspecific interactions are equivalent.

Correct, changes have been made as suggested.

Recommendations for the authors:

Reviewer #2 (Recommendations For The Authors):

I recommend to:

- re-organize the presentation of the material (see my concerns in the public review section). The manuscript is very difficult to read.

Changes have been made to help with understanding of our approach. Figure 1 was modified to show the variations of competitive growth response with relative competitive ability from minimum (null expectation) to maximum (competitive exclusion).

- explore the mathematical form the the remainder term. It seems important to understand that the remainder capture terms unrelated to competition as defined in the present scope.

The remainder measures deviations from the null expectation, due to species differences in growth and competitive ability or competition effect. The term has clear meaning, positive for more competitive species and negative for less competitive species (lines 202-204), and does not need to be further explored or partitioned. The deviations of observed yields from competitive expectations are outlined in lines 205-221.  

Reviewer #4 (Recommendations For The Authors):

The authors should be sure to include reproducible methods and share any data and code.

Both simulation and experimental data are shared through supplementary tables. Calculations are included in excel spreadsheets and do not require program coding.

Author response:

The following is the authors’ response to the original reviews.

Public Reviews:

Reviewer 1 (Public Review):

Summary:

The authors present a mean-field model that describes the interplay between (protein) aggregation and phase separation. Different classes of interaction complexity and aggregate dimensionality are considered, both in calculations concerning (equilibrium) phase behavior and kinetics of assembly formation.

Strengths:

The present work is, although purely theoretical, of high interest to understanding biological processes that occur as a result of a coupling between protein aggregation and phase separation. Of course, such processes are abundant, in the living cell as well as in in-vitro experiments. I appreciate the consideration of aggregates with various dimensionality, as well as the categorization into different ”interaction classes”, together with the mentioning of experimental observations from biology. The model is convincing and underlines the complexity associated with the distribution of proteins across phases and aggregates in the living cell.

Weaknesses:

There are a few minor weaknesses.

Reviewer 2 (Public Review):

This work deals with a very difficult physical problem: relating the assembly of building blocks on a molecular scale to the appearance of large, macroscopic assemblies. This problem is particularly difficult to treat, because of the large number of units involved, and of the complex way in which these units-monomers-interact with each other and with the solvent. In order to make the problem treatable, the authors recur to a number of approximations: Among these, there is the assumption that the system is spatially homogeneous, i.e., its features are the same in all regions of space. In particular, the homogeneity assumption may not hold in biologically relevant systems such as cells, where the behavior close to the cell membrane may strongly differ from the one in the bulk. As a result, this hypothesis calls for a cautious consideration and interpretation of the results of this work. Another notable simplification introduced by the authors is the assumption that the system can only follow two possible behaviors: In the first, each monomer interacts equally with the solvent; no matter the size of the cluster of which it is part. In the second case, monomers in the bulk of a cluster and monomers at the assembly boundary interact with the solvent in a different way. These two cases are considered not only because they simplify the problem, but also because they are inspired by biologically relevant proteins.

With these simplifications, the authors trace the phase diagram of the system, characterizing its phases for different fractions of the volume occupied by the monomers and solvent, and for different values of the temperature. The results qualitatively reproduce some features observed in recent experiments, such as an anomalous distribution of cluster sizes below the system saturation threshold, and the gelation of condensed phases above such threshold.

Reviewer 3 (Public Review):

Summary:

The authors combine classical theories of phase separation and self-assembly to establish a framework for explaining the coupling between the two phenomena in the context of protein assemblies and condensates. By starting from a mean-field free energy for monomers and assemblies immersed in solvent and imposing conditions of equilibrium, the authors derive phase diagrams indicating how assemblies partition into different condensed phases as temperature and the total volume fraction of proteins are varied. They find that phase separation can promote assembly within the protein-rich phase, providing a potential mechanism for spatial control of assembly. They extend their theory to account for the possibility of gelation. They also create a theory for the kinetics of self-assembly within phase separated systems, predicting how assembly size distributions change with time within the different phases as well as how the volumes of the different phases change with time.

Strengths:

The theoretical framework that the authors present is an interesting marriage of classic theories of phase separation and self-assembly. Its simplicity should make it a powerful general tool for understanding the thermodynamics of assembly coupled to phase separation, and it should provide a useful framework for analyzing experiments on assembly within biomolecular condensates.

The key advance over previous work is that the authors now account for how self-assembly can change the boundaries of the phase diagram.

A second interesting point is the explicit theoretical consideration for the possibility that gelation (i.e. self-assembly into a macroscopic aggregate) could account for widely observed solidification of condensates. While this concept has been broadly discussed, to date I have yet to see a rigorous theoretical analysis of the possibility.

The kinetic theory in sections 5 and 6 is also interesting as it extends on previous work by considering the kinetics of phase separation as well as those of self-assembly.

Weaknesses:

A key point the authors make about their theory is that it allows, as opposed to previous research, to study non-dilute limits. It is true that they consider gelation when the 3D assemblies become macroscopic. However, dilute solution theory assumptions seem to be embedded in many aspects of their theory, and it is not always clear where else the non-dilute limits are considered. Is it in the inter-species interaction χij? Why then do they never explore cases for which χij is nonzero in their analysis?

We explicitly consider that monomers and aggregates are non-dilute with respect to solvent. This is evident in accounting for the mixing entropy of all components, including the solvent. Moreover, we account for interactions among the monomers and the different aggregates with the solvent. We consider the case where each monomeric unit, independent in aggregate it is part of, interacts the same way with the solvent. Please note that this case corresponds to a non-dilute scenario where interactions indeed drive phase separation.

The connection between this theory and biological systems is described in the introduction but lost along the main text. It would be very helpful to point out, for instance, that the presence of phase separation might induce aggregation of proteins. This point is described formally at the end of Section 3, but a more qualitative connection to biological systems would be very useful here.

We thank the referee for the useful comment, we now mention this in the introduction (line 80) and point out the biological relevance of assembly formation and localization via the presence of phase separation (lines 268 and 283).

Building on the previous point, it would be helpful to give an intuitive sense of where the equations derived in the Appendices and presented in the main text come from and to spell out clear physical interpretations of the results. For example, it would be helpful to point out that Eq. 4 is a form of the law of mass action, familiar from introductory chemistry. It would be useful to better explain how the current work extends on existing previous work from these authors as well as others. Along these lines, closely related work by W. Jacobs and B. Rogers [O. Hedge et al. 2023, https://arxiv.org/abs/2301.06134; T. Li et al. 2023, https://arxiv.org/abs/2306.13198] should be cited in the introduction. The results discussed in the first paragraph of Section 3 on assembly size distributions in a homogeneous system are well-known from classic theories of self-assembly. This should be acknowledged and appropriate references should be added; see for instance, Rev. Mod. Phys. 93, 025008 and Statistical Thermodynamics Of Surfaces, Interfaces, And Membranes by Sam Safran. Equation 14 for the kinetic of volume fractions is given with reference to Bauermann et al. 2022, but it should be accompanied by a better intuitive interpretation of its terms in the main text. In particular, how should one understand the third term in this equation? Why does the change in volume impact the change of volume fraction in this way?

We thank the referee for the suggestions. We have included the missing references, with a particular emphasis on DNA nanostars that inhibit phase separation in DNA liquids in the definition of class II. We added intuitive explanations of the main equations, such as Eqs. (4),(8),(14), (17), and (18). Notice that, according to Mysels, Karol J., J. Chem. Educ., 33, 178 (1956) (https://pubs-acs-org.sire.ub.edu/doi/epdf/10.1021/ed033p178) we refer to (18) as the law of mass action.

The discussion in the last paragraph of Section 6 should be clarified. How can the total amount of protein in both phases decrease? This would necessarily violate either mass or volume conservation. Also, the discussion of why the volume is non-monotonic in time is not clear.

A decrease in the total amount of protein in both phases does not violate mass conservation, if the volume of the phases varies accordingly. In particular, the volume of the denser phase should grow. This given, in the case presented the total protein amount in the dense phase decreases, while in the dilute phase increases. For this reason, we revised the paragraph and now explain the results in more detail (see lines starting from 407). The nonmonotonic volume change is indeed a puzzling finding that, as we now state in the manuscript, requires further investigation. Given the lack of analytical approaches available to tackle the complex kinetics in the presence of coexisting phases, we believe that this analysis goes beyond the scope of the present paper.

Recommendations for the authors

Reviewer 1 (Recommendations For The Authors):

Line 96: I feel a mentioning/definition/explanation and perhaps some discussion on the parameter M (limiting aggregate size) would have been in place in the introduction of Equation (1). Furthermore, in the usual interpretation, Flory interaction parameters (symbolized χ) are dimensionless, as, classically, they represent an exchange energy (normalized by kT), defined on a monomeric basis. Here they seem to carry the dimension of energy.

We thank the reviewer for the observation. We have included a brief comment on M and mentioned that we use χ parameters that carry the dimension of energy such that, varying kBT, we scale at the same time the term containing interaction propensities (χ) and the one containing internal energies (_e_int). See the comment on line 127

Line 150: The choice of ρi \= i physically implies that a single protein is assumed to have the same as a solvent molecule. This may be a bit of a stretch. This assumption leads to an overestimation of the translational entropy of the aggregates (first term in Equation (1)). Acknowledging that ρ_1 >> ρs_ would give a pronounced desymmetrization of the phase diagram (I suspect).

Indeed, in the case of monomers only, the assumption leads to a symmetric phase diagram which may be unrealistic. Once assemblies form, however, the phase diagram becomes asymmetric and for this reason we decided to assume ρi \= i, simplifying the theoretical analysis. We have added a clarifying sentence in the manuscript, see line 163

Furthermore, the pictures in Figure 1a-c suggest the presence of a disordered residue, the degree of swelling of which might affect binding strength (see for instance: https://doi.org/10.3389/fnmol.2022.962526).

We added a comment on the possible coupling between internal free energies and interaction propensities, such as the swelling mechanism that affects binding sites, and included the reference above (line 215).

Line 154-156: It’s unclear what is meant with ”an internal bond that keeps each assembly together”. How should this be interpreted on an intuitive physical level?

We apologise for being unclear. We meant the internal bonds that lead to the formation of assemblies. We have now rephrased this sentence in the main text (lines starting from 169).

Line 254: The fact that ϕsg is defined below does not mean it does not fall out of the air here. The same holds for the consideration of the limit M →∞. Ideally, the main text should stand on its own, in particular with respect to physical intuitiveness, as well as the necessity and interest of discussion topics. Technical details, derivations and additional information can be in an appendix.

We agree with the referee and added some physical insights about the limit. We now also state clearly in the main text (line 298) that _ϕ_sg is affected by temperature and the free energy of internal bonds.

Line 257: ”Since we do not explicitly include the solvent in assembly formation we will consider the gel as a phase without solvent and thus ϕtot \= 1”. I’m not sure if I can agree with this. I would say, a gel, certainly in biological context, almost per definition contains a large fraction of solvent, i.e. here water. The situation ”ϕtot \= 1” would rather be a solid precipitate. Is gelation properly captured by this model?

We thank the referee for this very relevant observation. We now state in the main text that the model predicts a macroscopic assembly which we call ’the gel phase’, in agreement with previous literature. Then, to clarify, we added the sentence ”Please note that, since we do not explicitly include the solvent in assembly formation (see reaction scheme in Fig.1a), in our model the gel corresponds to a phase without solvent, _ϕ_tot \= 1. To account for biological gels that can be rich in water, our theory can be straightforwardly extended by incorporating the solvent into the reaction scheme.”, see main text line 300.

Line 268: Shouldn’t ”solvent” be ”solution”? If fsol is given by Equation (1), surely not only the solvent is considered.

Indeed, this is a typo, and we now use the term ’solution’ instead of ’solvent.’

Line 273: At this stage, the only information provided in the main text is that ω∞ is ”a constant that does not affect chemical nor phase equilibrium, except in the limit M →∞” (see lines 153-154). This is a little bit too abstract for me. Again, the main text should stand on its own, meaning the reader should not have to rely on an appendix to at least have an intuitive physical understanding of any modeling or input parameter discussed in the main text.

We thank the reviewer for pointing this out. We now comment on the physical interpretation of ω∞ in the main text, see lines from 320 on.

Figure 4. appears in Equation (39) but it is not defined.

We thank the reviewer for pointing this out. We have reshaped appendix 6A, making use of chemical activities and clarified the origin of the rate .

Line 317. I don’t fully understand the intention of the remark on the model being adaptable for ”primary and secondary nucleation”. How/in what way is this different from association and dissociation? For instance, classical nucleation theory is based on association and dissociation of monomeric units to and from clusters.

We agree that the kinetic rate coefficients kij (appearing in the association and dissociation rates ∆rij, Eq. 17) in our manuscript already depend on assembly length, see Appendix 6 B, where we now clarified their definition. Please note that, however, that secondary nucleation is a special kind of association, for which the kinetic rate coefficients corresponding to associations of small assemblies, i.e. kij with_i,j_ ≪ M, explicitly depend on the presence of large assemblies with sizes l ≫ 1. In our manuscript, we have not accounted for such a dependence. We now make this aspect clear in the manuscript, see Appendix 6 B.

Line 321. Why is ∆rij called the ”monomer exchange rate”? In line 318 the same parameter is defined as the ”reaction rate for the formation of a (i+j)-mer”. Why should these be the same?

We thank the reviewer for spotting this typo.

Line 323. Why do these calculations use M = 15?

The exploration of a 15-dimensional phase space is already numerically challenging. We are currently working on a generalization of the numerical scheme to work with larger values of M but, to discuss the fundamental physical principles, we kept M \= 15.

Reviewer 2 (Recommendations For The Authors):

The manuscript presents several issues, on both the scientific and presentational level, which need to be carefully addressed. Please find below a list of the points that need to be addressed by the authors, divided into major and minor points. Major issues:

• A general, major concern about the results in the paper is the homogeneity assumption. I do understand that repeating the whole analysis presented in the manuscript by allowing for spatial inhomogeneities partially goes beyond the scope of this paper. However, the authors should at least discuss how such inhomogeneities may alter the results in a qualitative way, and treat explicitly the presence of inhomogeneity in one prototypical case treated in the manuscript. Namely, what happens if the volume fractions and relative molecular volumes in the free energy (1) depend on space, e.g., ϕi → ϕi(x)?

We would like to stress that, in the present paper, we do account for spatial inhomogeneities. Indeed, in the case of phase separation, we consider systems which are divided into two phases, characterized by different values of the assemblies’ volume fractions ϕi. We do, however, consider the system to be homogeneous inside the phases, implying a jump in the value of the volume fraction at the interface between the two phases. In this sense, the analysis we carry out is valid in the thermodynamic limit, where gradients of the volume fractions ϕi(x) within the phases, can be neglected. On the other hand, considering the full spatial problem, i.e. solving the equations for M \= 15 spatially varying fields, would be numerically extremely challenging.

• The authors’ results relate molecular assembly- a phenomenon at the molecular scale-to phase separation-a mesoscopic or macroscopic phenomenon. The authors should stress the conceptual importance of this connection between scales, and present their results from the perspective of a multi-scale model.

We thank the reviewer for pointing this out. We now emphasize the multi-scale feature of our model in the introduction (line 80).

• Starting from Section 1, the reader is not well guided through the sections that follow. The authors should provide an outline of the line of though that they are going to follow in the following sections, and logically connect each section to the next one with a short paragraph at the end of each section. This paragraph should resume what has been addressed in the current section, and the connection with the topic that will be addressed in the next one.

We agree with the reviewer and have added a transitioning sentence at the end of each paragraph.

• ’We focus on linear assemblies (d = 1)’: Given the striking differences of the results between d = 1 and d > 1 shown above, the authors should discuss what happens for d > 1 as well.

• ’In figure Fig. 5a, we show the initial and final equilibrium binodals (black and coloured curve, respectively), for the case of linear assemblies (d = 1) belonging to class 1’: Again, show what happens for d > 1.

We agree with the reviewer, the kinetics in d > 1 would be definitely interesting. However, in this case, one assembly can become macroscopic (i.e. M must be set to ∞). This requires some substantial modification in the kinetic scheme, like introducing an absorbing boundary condition for monomers ’sucked in’ the gel. We prefer to leave this for future work, and now state it explicitly in the manuscript (line 383).

• ’This difference arises because, within class 2, monomers in the bulk of an assembly have reduced interaction propensity with respect to the boundary ones. As a consequence, the formation of large clusters shifts the onset of phase separation to higher ϕtot values.’: To prove this argument, the authors should show Fig. 2g and h for d > 1. In fact, by varying d, the effect of the boundary vs. bulk also varies.

We prefer to discuss the thermodynamics of d > 1 in section 4 on gelation. There we present only a single phase diagram so as not to blow up the discussion on equilibrium too much.

• ’referring for simplicity to systems belonging to Class 1’: The authors should do the same analysis for Class 2.

We agree with the reviewer. However, again not to blow up the discussion on equilibrium, we leave it for future work.

• ’other, implying that the corresponding Flory-Huggins parameter χij vanishes’: Why?

The explanation based on a lattice model is reported in Appendix 2, and is now more clearly referenced (line 185).

Minor issues:

• Eq. (10): Here the authors should explain in the main text, possibly in a simple and intuitive way, why the number of monomers i and the space dimension d enter the righthand side of this equation in this particular way.

We thank the reviewer for pointing this out. We added the physical origin of the scaling with dimension in Eq. (10) and in Eq. (8), as pointed out by reviewer 3.

• ’The second and fifth terms of fsol characterize the internal free energies’: What do you mean by ’characterize the internal free energies’? Please clarify.

As we now state more clearly (lines 114-120), these two contributions include the internal free energies ω_s and _ωi, stemming from the free energy of internal bonds that lead to assembly formation.

• ’depend on the scaling form of the’: Scaling with respect to what ? Please clarify.

We have now clarified that the scaling is with respect to the assembly size i.

• Figure 2 is way too dense: it should be split into two figures, and the legend of each of the two figures should be expanded to properly guide the reader to understand the figures.

We understand the reviewer’s point of view. To avoid altering the present flow, we decided not to split the figure, but we have included shaded boxes to better guide the reader.

• ’this is a consequence of the gelation transition’: Please clarify

• ’and this limitation can be dealt with by introducing explicitly the infinite-sized gel in the free energy’: Why? Please clarify.

We have now rephrased these sentences, hopefully in a clearer way. We now state: ’We know that this divergence is physical, and is caused by the gelation transition. This limitation can be dealt with by introducing explicitly a term in the free energy that accounts for an infinite-sized assembly (the gel)’, see lines 320-322.

• Figure 4: Add plots of panels d, e, h and i with log scale on the y axis to make explicit an eventual exponential behavior, and revise the text accordingly

Not to further complicate Figure 4, we preferred to display the logarithmic plots of the equilibrium distribution in the appendix, see Figure A3-1.

• ’... an equilibrium distribution which monotonously decreases with assembly size’: It is not the distributions that decreases but the cluster volume fraction, please rephrase.

We thank the reviewer for pointing this out and have now rephrased this sentence (line 394).

Reviewer 3 (Recommendations For The Authors):

I could not obtain the exact form of Eq 29 in App 3, can the authors elaborate on this calculation. App 3: What does it mean binodal agrees well with ϕsg? And doesn’t ϕsg depend on temperature through phi tilde? What temperature is this result for?

We apologise for the unclear explanation. We now state in detail that Eq. (29) is obtained by plugging the expression of ϕi given in Eq. (24) into Eq. (1), in the main text. The dependence of ϕ₁ on ϕ_tot is expressed in Eq. (26), and we have omitted linear terms in ϕ_tot, since they do not affect phase equilibrium (see lines 802-809). Moreover, ϕsg depends indeed on k_BT. We refer to the comparison between the full curve ϕsg in the k_BT−ϕ_tot plane, and the branch of the binodal between the triple point (indicated now with a cross) and ϕ_tot \= 1. The two curves are close, as expected since both correspond to the boundary between homogeneous mixtures and the gel state, obtained with different methods.

The references to Figures in the appendices are confusing. Please make it clear whether Figures in the main text or the appendices are being referenced. On a related note, the Appendix figures seem to be placed in appendices whose text describes something else - Appendix 2, Figure 1 should be moved to Appendix 3; Appendix 3, Figure 1 should be moved to Appendix 4; etc.

We revised the appendix, corrected the figure positions and clarified their references.

Author response:

The following is the authors’ response to the original reviews.

Public Reviews:

Reviewer #1 (Public Review):

Summary:

This research offers an in-depth exploration and quantification of social vocalization within three families of Mongolian gerbils. In an enlarged, semi-natural environment, the study continuously monitored two parent gerbils and their four pups from P14 to P34. Through dimensionality reduction and clustering, a diverse range of gerbil call types was identified. Interestingly, distinct sets of vocalizations were used by different families in their daily interactions, with unique transition structures exhibited across these families. The primary results of this study are compelling, although some elements could benefit from clarification

Strengths:

Three elements of this study warrant emphasis. Firstly, it bridges the gap between laboratory and natural environments. This approach offers the opportunity to examine natural social behavior within a controlled setting (such as specified family composition, diet, and life stages), maintaining the social relevance of the behavior. Secondly, it seeks to understand short-timescale behaviors, like vocalizations, within the broader context of daily and life-stage timescales. Lastly, the use of unsupervised learning precludes the injection of human bias, such as pre-defined call categories, allowing the discovery of the diversity of vocal outputs.

Weaknesses:

(1) While the notable differences in vocal clusters across families are convincing, the drivers of these differences remain unclear. Are they attributable to "dialect," call usage, or specific vocalizing individuals (e.g., adults vs. pups)? Further investigation, via a literature review or additional observation, into acoustic differences between adult and pup calls is recommended. Moreover, a consistent post-weaning decrease in the bottom-left cluster (Fig. S3) invites interpretation: could this reflect drops in pup vocalization?

Thank you for bringing up this point of clarification. Without knowledge of individual vocalizers, we are unable to rigorously assess pronunciation differences between individuals, however we can get a clear proxy for dialect through observing usage differences between families. We’ve added the following text (blue) in the Discussion to help clarify:

“To address whether gerbils also exhibit family specific vocal features, we compared GMM-labeled vocal cluster usages across the three recorded families and showed differences in vocal type usage (Figure 3). The differences in this study align with the definition of human vocal dialect, which is a regional or social variety of language that can differ in pronunciation, grammatical, semantic and/or language use differences (Henry et al., 2015). This definition of dialect is inclusive of both pronunciation differences (e.g. a Bostonian’s characteristic pronunciation of “car” as “cah”) and usage differences (e.g. a Bostonian’s preferential usage of the words “Go Red Sox” vs. a New Yorker’s preferential usage of the words “Go Yankees”). In our case, vocal clusters can be rarely observed in some families yet highly over-expressed in others (e.g. analogous to language usage differences in humans), or highly expressed in both families, but contain subtle spectrotemporal variations (Figure 3D, Family 1 cluster 11 vs. Family 3 clusters 2, 18, 30; e.g. analogous to pronunciation differences in humans).”

Indeed, our recordings obtained after pup removal could suggest that adults may use fewer low frequency calls (bottom left cluster in UMAP). However, this dataset does not permit a proper assessment of post-weaning pup calls. In fact, our results and the literature shows that adults are likely to use low frequency calls, but only during social interactions with pups or other adults. For example, Furuyama et al. 2022 describe a number of low frequency call types used by adults in agonistic social interactions, which look similar to a low frequency call type used by pups described in Silberstein et al. 2023. Similarly, Ter-Mikaelian et al. 2012 (their Figure 6) recorded several types of sonic vocalizations during adult social interaction. To our knowledge, it has not been shown whether gerbil pups and adults produce distinct call types. It is a challenging problem to solve, as animals placed in isolation (i.e. an experimental condition for which the identity of the vocalizer is known) vocalize infrequently and of the limited number they might emit, they do not use the full range of vocalizations described in the literature (RP personal observations). To properly address this question, one would need to elicit full use of the vocal repertoire through free social interaction, then attribute calls to individual vocalizers via sound source localization and/or head-mounted microphones — we are currently pursuing both of these technical challenges, but this is outside the scope of this manuscript.

Although the literature reflects the limitations discussed above, we have added a brief paragraph to the Discussion (limitations section) that addresses the reviewer’s question about the development of vocalizations:

“Although we were not able to attribute vocalizations to individual family members, we did seek to determine the importance of family structure by comparing audio recordings before and after removal of the pups at P30. The results show a clear effect of family integrity, and the sudden reduction of sonic calls following pup removal (Figure S3) could suggest that these vocalizations are produced selectively by pups.

However, there is ample evidence that adult gerbils also produce sonic vocalizations. For example, a number of low frequency call types are used by adults during a range of social interactions (Ter-Mikaelian et al., 2012; Furuyama et al., 2022), some of which are similar to a low frequency call type used by pups (Silberstein et al., 2023). Vocalization patterns of developing gerbils depend on isolation or staged interactions. Thus, when gerbil pups are recorded during isolation, ultrasonic vocalization rate declines and sonic vocalizations increase for animals that are in a high arousal state (De Ghett 1974, Silberstein et al., 2023). As gerbils progress from juvenile to adolescent development (P17-55) a significant increase in ultrasonic vocalization rate is observed during dyadic social encounters, with a distinct change in usage pattern that depends upon the sex of each animal (Holman & Seale 1991, Holman et al. 1995). The development of vocalization types has been assessed in another member of the Gerbillinae subfamily, called fat-tailed gerbils (Pachyuromys duprasi), during isolation and handling. Here, the number of ultrasonic vocalization syllable types increase from neonatal to adult animals (Zaytseva et al. 2019), while some very low frequency sonic call types were rarely observed after P20 (Zaytseva et al. 2020). By comparison, mouse syllable usage changes during development, but pups produced 10 of the 11 syllable types produced by adults (Grimsley et al. 2011). In summary, our understanding of the maturation of vocalization usage remains limited by our inability to obtain longitudinal data from individual animals within their natural social setting. For example, when recorded in their natural environment, chimpanzees display a prolonged maturation of vocalization complexity, such as the probability of a unique utterance in a sequence, with the greatest changes occuring when animals begin to experience non-kin social interactions (Bortolato et al. 2023).”

(2) Developmental progression, particularly during pre-weaning periods when pup vocal output remains unstable, might be another factor influencing cross-family vocal differences. Representing data from this non-stationary process as an overall density map could result in the loss of time-dependent information. For instance, were dominating call types consistently present throughout the recording period, or were they prominent only at specific times? Displaying the evolution of the density map would enhance understanding of this aspect.

This is a great suggestion. Thank you for bringing it up. To address this, we have added an additional figure (Figure 4) to the main text (Note that the former Figure 4 is now Figure 5). New text associated with this new figure was added to the Results and Discussion sections:

Results

“Vocal usage differences remain stable across days of development It is possible that the observed vocal usage differences could result from varying developmental progression of vocal behavior or overexpression of certain vocal types during specific periods within the recording. To assess the potential effect of daily variation on family specific vocal usage, we visualized density maps of vocal usage across days for each of the families (Figure 4A). There are two noteworthy trends: 1.) the density map remains coarsely stable across days (rows) and 2.) the maps look distinct across families on any given day (columns). This is a qualitative approximation for the repertoire’s stability, but does not take into account variation of call type usage (as defined by GMM clustering of the latent space). Figure 4B, shows the normalized usage of each cluster type over development for each family. Cluster usages during the period of “full family, shared recording days” (postnatal days beneath the purple bars) are stable across days within families – as is apparent by the horizontal striations in the plot – though each family maintains this stability through using a unique set of call types. This is addressed empirically in Figure 4C, which shows clearly separable PCA projections of the cluster usages shown in Figure 4B (purple days). Finally, we computed the pairwise Mean Max Discrepancy (MMD) between latent distributions of vocalizations from individual recording days for each of the families (Figure 4D). This shows that across-family repertoire differences are substantially larger than within-family differences. This is visualized in a multidimensional scaling projection of the MMD matrix in Figure 4E.”

Discussion

“The described family differences collapse data from multiple days into a single comparison, however it’s possible that factors such as vocal development and/or high usage of particular vocal types during specific periods of the recording could explain family differences. Therefore, we took advantage of the longitudinal nature of our dataset to assess whether repertoire differences remain stable across time. First, we visualized vocal repertoire usage across days as either UMAP probability density maps (Figure 4A) or daily GMM cluster usages (Figure 4B). Though qualitative, one can appreciate that family repertoire usage remains stable across days and appears to differ on a consistent daily basis across families. To formally quantify this, we first projected GMM cluster usages from Figure 4B into PC space and show that family GMM cluster usage patterns are highly separable, regardless of postnatal day (Figure 4C). If families had used a more overlapping set of call types, then the projections would have appeared intermixed. Next, we performed a cluster-free analysis by computing the pairwise MMD distance between VAE latent distributions of vocalizations from each family and day (Figure 4D). This analysis shows very low MMD values across days within a family (i.e. the repertoire is highly consistent with itself), and high MMD values across families/days (greater than would be expected by chance; see shuffle control in Figure S2D). The relative differences in this matrix are made clear in Figure 4E, which provides additional evidence that family vocal repertoires remain stable across days and are consistently different from other families. Taken together, we believe that this is compelling evidence that differences in vocal repertoires between families are not driven by dominating call types during specific phases in the recording period; rather, families consistently emit characteristic sets of call types across days. This opens up the possibility to assess repertoire differences over much shorter time periods (e.g. 24 hours) in future studies.”

(3) Family-specific vocalizations were credited to the transition structure, a finding that may seem obvious if the 1-gram (i.e., the proportion of call types) already differs. This result lacks depth unless it can be demonstrated that, firstly, the transition matrix provides a robust description of the data, and secondly, different families arrange the same set of syllables into unique sequences.

Thank you for these important suggestions. We agree that it is true that the 2-gram transition structure must vary based on the 1-gram structure. To determine whether this influences the interpretation of the finding, we have added Figure S5 and the following text in the Results section:

“To determine whether differences in 1-gram structure contribute to differences in the transition (2-gram) structure, we performed a number of controls. Although subtle, vertical streaks are clearly present in shuffled transition matrices that correspond to 1-gram usages (Figure S5A-B). Given the shuffled data structure, we sought to determine whether the observed transition probabilities differed significantly from chance levels. We randomly shuffled label sequences 1000 times independently for each family to generate a null transition matrix distribution. Using these null distributions and the observed transition probabilities, we computed a p-value for each transition using a one-sample t-test and created a binary transition matrix indicating which transitions happen above chance levels (Figure S5C, black pixels, p <= 0.05 after post hoc Benjamini-Hochberg multiple comparisons correction). As is made clear in Figure S5C, most transitions for each family occur significantly above chance levels, despite the inherent 1-gram structure. Moreover, by looking at transitions from a highly usage cluster type used roughly the same proportion across families (cluster 12), we show that families arrange the same sets of vocal clusters into unique sequences (Figure S5D). We believe that this provides compelling evidence that the 1-gram structure does not change the interpretation of the main claim that transition structure varies by family. “””

To address your second point, we inspected frequent transitions from individual syllables to all other syllables using bigram transition probability graphs. This revealed a common trend that across all families, many shared and unshared transitions existed, suggesting that families use the same sets of syllables to make unique transition patterns. Figure S5D shows a single syllable example of the phenomenon, with red lines indicating the shared transition types between families and black showing transition patterns not shared between families (i.e. unique family-specific transitions, or lack thereof).”

Reviewer #2 (Public Review):

Peterson et al., perform a series of behavioral experiments to study the repertoire and variance of Mongolian gerbil vocalizations across social groups (families). A key strength of the study is the use of a behavioral paradigm which allows for long term audio recordings under naturalistic conditions. This experimental set-up results in the identification of additional vocalization types. In combination with state of the art methods for vocalization analysis, the authors demonstrate that the distribution of sound types and the transitions between these sound types across three gerbil families is different. This is a highly compelling finding which suggests that individual families may develop distinct vocal repertoires. One potential limitation of the study lies in the cluster analysis used for identifying distinct vocalization types. The authors use a Gaussian Mixed Model (GMM) trained on variational auto Encoder derived latent representation of vocalizations to classify recorded sounds into clusters. Through the analysis the authors identify 70 distinct clusters and demonstrate a differential usage of these sound clusters across families. While the authors acknowledge the inherent challenges in cluster analysis and provide additional analyses (i.e. maximum mean discrepancy, MMD), additional analysis would increase the strength of the conclusions. In particular, analysis with different cluster sizes would be valuable. An additional limitation of the study is that due to the methodology that is used, the authors can not provide any information about the bioacoustic features that contribute to differences in sound types across families which limits interpretations about how the animals may perceive and react to these sounds in an ethologically relevant manner.

The conclusions of this paper are well supported by data, but certain parts of the data analysis should be expanded and more fully explained.

• Can the authors comment on the potential biological significance of the 70 sound clusters? Does each cluster represent a single sound type? How many vocal clusters can be attributed to a single individual? Similarly, can the authors comment on the intra-individual and inter-individual variability of the sound types within and across families?

Previous work documenting the Mongolian gerbil repertoire (Ter-Mikaelian 2012, Kobayasi 2012) has revealed ~12 vocalization types that vary with social context. Our thinking is that we are capturing these ~12 (plus a few more, as illustrated in Figure 2C) as well as individual or family-specific variations of some call types. Although the number of discrete call types is likely less than 70, it’s plausible that variation due to vocalizer identity pushes some calls into unique clusters. This idea is supported by the fact that both naked mole rats and Mongolian gerbils have been shown to exhibit individual-specific variation in vocalizations, though only in single call types (Barker 2021, Figure 1; Nishiyama 2011, Table I). The current study is not ideal to test this prediction, as we cannot attribute each vocalization to individual family members. Using our 4-mic array, we attempted to apply established sound source localization techniques to assign vocalizations to individuals (Neunuebel 2015), but the technique failed, presumably due to high amounts of reverberation in the arena. We are currently developing a custom deep learning based sound localization algorithm, and had hoped to extract individual animal vocalizations from our data set (part of the reason why this manuscript has taken longer than expected to return!), but the performance is not yet satisfactory for large groups of animals. We have added text to the Methods sections with the context outlined above to further justify the use of ~70 clusters.

• As a main conclusion of the paper rests on the different distribution of sound clusters across families, it is important to validate the robustness of these differences across different cluster parameters. Specifically, the authors state that "we selected 70 clusters as the most parsimonious fit". Could the authors provide more details about how this was fit? Specifically, could the authors expand upon what is meant by "prior domain knowledge about the number of vocal types...". If the authors chose a range of cluster values (i.e. 10, 30, 50, 90) does the significance of the results still hold?

Thank you for the suggestion, this is an important point that we have addressed with new analyses in the revision (see GMM clustering methods and new Figure S4). The prior domain knowledge referenced is with respect to the information known about the Mongolian gerbil vocal types provided in the response above. We have made this more clear in the discussion.

We mainly based our selection of the number of clusters using the elbow method on GMM held-out log likelihood (Figure S2C). Around 70 clusters is when the likelihood begins to plateau, though it’s clear that there are a number of reasonable cluster sizes. To assess whether cluster size has an effect on interpretation of the family differences result, we added Figure S5, where we varied the number of GMM clusters used and compared cluster usage differences across families (Figure S4A). We quantified pairwise family differences in cluster usage by computing the sum of the absolute value of differential cluster usages, for each GMM cluster value (Figure S4B). We find that relative usage differences remain unchanged across the range of cluster values used, indicating that GMM cluster size does bias the finding.

• While VAEs are powerful tools for analyzing complex datasets in this case they are restricted to analysis of spectrogram images. Have the authors identified any acoustic differences (i.e. in pitch, frequency, and other sound components) across families?

Though it’s true that this VAE is limited to spectrograms, the VAE latent space has been shown to correspond to real acoustic features such as frequency and duration, and contain a higher representational capacity than traditional acoustic features (Goffinet 2021, Figure 2). Therefore, clustering of the latent space necessarily means that vocalizations with similar acoustic features are clustered together regardless of their family identity.

Despite this, your point is well taken that there could be systematic differences in certain acoustic features for specific call types. We are not able to ascertain this with the current dataset. This is addressed in Barker 2021 by recording a single call type (soft chirp) from individuals within and across families. Mongolian gerbils have been shown to exhibit individual differences in the initial, terminal, minimum, and maximum frequency of the ultrasonic up-frequency modulated call type (Figure 2, top right green; Nishiyama 2011, Figure 1A ). Therefore it’s possible that family-specific differences exist for that particular call type. To assess whether other call types show family or individual differences, it’s necessary to either 1.) elicit all call types from an animal in isolation or 2.) determine vocalizer identity in social-vocal interactions. The problem with the former idea is that gerbils only produce up-frequency modulated USVs in isolation and there is no known way to elicit the full vocal repertoire in single animals. The latter idea would allow for full use of the vocal repertoire, but requires invasive techniques (e.g., skull-implanted microphones, or awake-behaving laryngeal nerve recordings) that permit assignment of vocalizations to individuals during a natural social interaction. We are actively exploring solutions to both problems.

It’s likely that future studies will look deeper into acoustic differences between individuals and families. Therefore, we have added acoustic feature quantification of vocalizations in each of the GMM clusters as a reference (Figure S6).

Reviewer #3 (Public Review):

Summary:

In this study, Peterson et al. longitudinally record and document the vocal repertoires of three Mongolian gerbil families. Using unsupervised learning techniques, they map the variability across these groups, finding that while overall statistics of, e.g., vocal emission rates and bout lengths are similar, families differed markedly in their distributions of syllable types and the transitions between these types within bouts. In addition, the large and rich data are likely to be valuable to others in the field.

Strengths:

- Extensive data collection across multiple days in multiple family groups.

-  Thoughtful application of modern analysis techniques for analyzing vocal repertoires. - Careful examination of the statistical structure of vocal behavior, with indications that these gerbils, like naked mole rats, may differ in repertoire across families.

Weaknesses:

- The work is largely descriptive, documenting behavior rather than testing a specific hypothesis.

- The number of families (N=3) is somewhat limited.

We agree that the number of families is relatively small. However, our new analysis of vocal repertoire by postnatal day (Figure 4) demonstrates that the finding is quite robust. A high sample-size study was outside the scope of this initial observational study given the difficulty of obtaining and processing longitudinal data of this scale. In light of new analyses in Figure 4, we are confident that future studies will not need so much data to characterize family-specific differences. A single 24-hour recording should be sufficient, making comparison of many more families relatively straightforward.

Recommendations for the authors:

Reviewer #1 (Recommendations For The Authors):

Several minor concerns:

(1) The three thresholds used for vocalization segmentation lack explanation.

Figure 1C's first vocal event appears to define the first gap via the gray threshold (th_2, as the trace does not cross the black line) and the second gap via the black threshold (th_1 or th_3). And this is not addressed in the Methods section.

Thank you for bringing this to our attention. We agree, this is presented in an unnecessarily complicated way. We have updated the methods section describing the thresholding procedure.

“Sound onsets are detected when the amplitude exceeds 'th_3' (black dashed line, Figure 1C), and sound offset occurs when there is a subsequent local minimum e.g., amplitude less than 'th_2' (gray dashed line, Figure 1C), or 'th_1' (black dashed line, Figure 1C), whichever comes first. In this specific use case, th_2 (5) will always come before th_1 (2), therefore the gray dashed line will always be the offset. A subsequent onset will be marked if the sound amplitude crosses th_2 or th_3, whichever comes first. For example, the first sound event detected in Figure 1C shows the sound amplitude rising above the black dashed line (th_3) and marks an onset. Subsequently, the amplitude trace falls below the gray dashed line (th_2) and an offset is marked. Finally, the amplitude rises above th_2 without dipping below th_3 and an onset for a new sound event is marked. Had the amplitude dipped below th_3, a new sound event onset would be marked when the amplitude trace subsequently exceeded th_3 (e.g. between sound event 2 and 3, Figure 1C). The maximum and minimum syllable durations were selected based on published duration ranges of gerbil vocalizations (Ter-Mikaelian et al. 2012, Kobayasi & Riquimaroux, 2012).”

(2) The determination of multi-syllabic calls could be explained further. In Figure 1C, for instance, do syllables separated by short gaps (e.g., the first syllable and the rest of the first group, and the third group in this example) belong to the same call or different calls?

We have added an operational definition of mono vs. multisyllabic calls in the Results section:

“Vocalizations occur as either single syllables bounded by silence (monosyllabic) or consist of combinations of single syllables without a silent interval (multisyllabic).”

Under this definition, the examples you mentioned in Figure 1C are considered monosyllabic. One could reasonably expand the definition to include calls separated by less than X ms of silence for example, however we choose not to do that in this study. A deeper understanding of the phonation mechanisms for different gerbil vocalization types would be helpful to more rigorously determine the distinction between mono vs. multisyllabic vocalizations.

(3) Labeling the calls shown in Fig. 3D in the latent feature space would help highlight within-family diversity and between-family similarities.

Great suggestion. We have updated Figure 3 to include where in UMAP space each family’s preferred clusters are.

(4) In the introduction, the statement, "Therefore, our study considers the possibility that there is a diversity of vocalizations within the gerbil family social group" doesn't naturally follow from the previous example. This could be rephrased.

Agreed, thank you. We revised this section of the introduction to flow better.

Reviewer #2 (Recommendations For The Authors):

While outside the scope of the current study the authors may consider the following experiments and analysis for future studies:

• Do vocal repertories retain their family signatures across subsequent generations of pups? (i.e. if vocalizations are continually monitored during second or third litters of the same parents).

• Do the authors observe any long-term changes in family repertoires related to the developmental trajectory of the pups? Are there changes in individual pup vocal features or sound type usage throughout development?

Thank you for these great suggestions. Given that naked mole rats learn vocalizations through cultural transmission, it would be interesting to see whether other subterranean species with complex social structures (gerbils, voles, rats) have similar abilities. A straightforward way to assess this possibility could be as you suggest — are latent distributions of vocalizations from multi-generational families closer together than cross-family differences? If true, this would provide compelling evidence to investigate further.

We partially address your second suggestion in our response to Reviewer 1 and in Figure S4, which shows that the family repertoire remains stable throughout this particular period of development. This doesn’t rule out the possibility that there could be other phases of development that undergo more vocal change. Your final suggestion is an area that we are actively researching and eager to know the answer to. A follow-up question: could differences in pup vocal features contribute to differential care by parents?

Reviewer #3 (Recommendations For The Authors):

In all, I found the paper clearly written and the figures easy to follow. One small suggestion:

Figure 1: I can't see the black and gray thresholds described in the caption very well. Perhaps a zoom-in to the first 0.15s or so of the normalized amplitude plot would better display these.

Agreed, thank you. We added a zoom-in to Figure 1.

Author response:

The following is the authors’ response to the original reviews.

Public Reviews:

Reviewer #1 (Public Review):

Summary:

In this manuscript, Unckless and colleagues address the issue of the maintenance of genetic diversity of the gene diptericin A, which encodes an antimicrobial peptide in the model organism Drosophila melanogaster.

Strengths:

The data indicate that flies homozygous for the dptA S69 allele are better protected against some bacteria. By contrast, male flies homozygous for the R69 allele better resist starvation than flies homozygous for the S69 allele.

Weaknesses:

-I am surprised by the inconsistency between the data presented in Fig. 1A and Fig. S2A for the survival of male flies after infection with P. rettgeri. I am not convinced that the data presented support the claim that females have lower survival rates than males when infected with P. rettgeri (lines 176-182).

The two figures are pasted above (1A left, S2A right). The reviewer is correct that the two experiments look different in terms of overall outcomes for males, though qualitatively similar. These two experiments were performed by different researchers, and as much as we attempt to infect consistently from researcher to researcher, some have heavier hands than others. It is true that the genotype that has the largest sex effect is the arginine line (blue) where females (in this experiment) are as bad as the null allele, and males are more intermediate. Also note that the experiments in S2A (male and female) were done in the same block so they are the better comparison. We’ve reflected this in the manuscript.

- The data in Fig. 2 do not seem to support the claim that female flies with either the dptA S69 or the R69 alleles have a longer lifespan than males (lines 211-215). A comment on the [delta] dpt line, which is one of the CRISPR edited lines, would be welcome.

We’ve reworded this section based on these comments.

- The data in Fig. 2B show that male flies with the dptA S69 or R69 alleles have the same lifespan when poly-associated with L. plantarum and A. tropicalis, which contradicts the claim of the authors (lines 256-260).

This is correct – the effect is only in females. It has been corrected.

Reviewer #2 (Public Review):

Summary: In this study, the authors delve into the mechanisms responsible for the maintenance of two diptericin alleles within Drosophila populations. Diptericin is a significant antimicrobial peptide that plays a dual role in fly defense against systemic bacterial infections and in shaping the gut bacterial community, contributing to gut homeostasis.

Strengths: The study unquestionably demonstrates the distinct functions of these two diptericin alleles in responding to systemic infections caused by specific bacteria and in regulating gut homeostasis and fly physiology. Notably, these effects vary between male and female flies.

Weaknesses: Although the findings are highly intriguing and shed light on crucial mechanisms contributing to the preservation of both diptericin alleles in fly populations, a more comprehensive investigation is warranted to dissect the selection mechanisms at play, particularly concerning diptericin's roles in systemic infection and gut homeostasis. Unfortunately, the results from the association study conducted on wild-caught flies lack conclusive evidence.

This is true that the wild fly association study is mostly a negative result. We’ve backed off the claim about the Morganella association.

Major Concerns:

Lines 120-134: The second hypothesis is not adequately defined or articulated. Please revise it to provide more clarity. Additionally, it should be explicitly stated that the first part of the first hypothesis (pathogen specificity), i.e., the superior survival of the S allele in Providencia infections compared to the R allele, has been previously investigated and supported by the results in the Unckless et al. 2016 paper. The current study aims to additionally investigate the opposite scenario: whether the R allele exhibits better survival in a different infection. Please consider revising to emphasize this point.

We’ve reworded this section and added references to both the Unckless et al. 2016 and Hanson et al. 2023 papers.

Figures and statistical analyses: It is essential to present the results of significant differences from the statistical analyses within Figures 1B, 2B, and 3. Additionally, please include detailed descriptions of the statistical analysis methods in the figure legends. Specify whether the error bars represent standard error or standard deviation, particularly in Figure 3, where assays were conducted with as few as 3 flies.

We have added statistical details as requested.

Lines 317-318 (as well as 320-328): The data related to P. rettgeri appear somewhat incomplete, and the authors acknowledge that bacterial load varies significantly, and this bacterium establishes poorly in the gut. These data may introduce more noise than clarity to the study. Please consider revising these sections by either providing more data, refining the presentation, or possibly removing them altogether.

The fact that P. rettgeri establishes poorly in the gut in wildtype flies is the result of several unpublished experiments in the Lazzaro and Unckless labs. We don’t have this as a figure because it was not directly tested in these experiments. We’ve added a note that it is personal observation and we’ve reworked the discussion in the second section.

Lines 335-387 and Figure 4: Although these results are intriguing and suggest interactions between functional diptericin and fly physiology, some mediated by the gut microbiome, they remain descriptive and do not significantly contribute to our understanding of the mechanism that maintains the diptericin alleles.

While the reviewer is correct that these experiments do not elucidate mechanism, they do strongly suggest (based on the controlled nature of the experiments) that the physiological tradeoffs are due to Diptericin genotype. The disagreement is the level of “mechanism”. At the evolutionary level, the demonstration of a physiological cost of a protective immune allele is sufficient to explain the maintenance of alleles. However, we have not determined (and did not attempt to determine) why Diptericin genotype influences these traits. That will have to wait for future experiments.

Lines 399-400: The contrast between this result and statement and the highly reproducible data presented in Figures 2-4 should be discussed.

We’ve added some discussion to this section including a reference to the “inconstancy” of the Drosophila gut microbiome.

Lines 422-429 and Figure 5D: The conclusion regarding an association between diptericin alleles and Morganellaceae bacteria is not clearly supported by Figure 5D and lacks statistical evidence.

We’ve changed this to just be suggestive.

Reviewer #3 (Public Review):

Summary:

This paper investigates the evolutionary aspects around a single amino acid polymorphism in an immune peptide (the antimicrobial peptide Diptericin A) of Drosophila melanogaster. This polymorphism was shown in an earlier population genetic study to be under long-term balancing selection. Using flies with different AA at this immune peptide it was found that one allelic form provides better survival of systemic infections by a bacterial pathogen, but that the alternative allele provides its carriers a longer lifespan under certain conditions (depending on the microbiota). It is suggested that these contrasting fitness effects of the two alleles contribute to balance their long-term evolutionary fate.

Strengths:

The approach taken and the results presented are interesting and show the way forward for studying such polymorphisms experimentally.

Weaknesses:

(1) A clear demonstration (in one experiment) that the antagonistic effect of the two selection pressures isolated is not provided.

The study is overwhelming with many experiments and countless statistical tests. The overall conclusion of the many experiments and tests suggests that "dptS69 flies survive systemic infection better, while dptS69R flies survive some opportunistic gut infections better." (line 444-446). Given the number of results, different experiments, and hundreds of tests conducted, how can we make sure that the result is not just one of many possible combinations? I suggest experimentally testing this conclusion in one experiment (one may call this the "killer-experiment") with the relevant treatments being conducted at the same time, side by side, and the appropriate statistical test being conducted by a statistical test for a treatment x genotype interaction effect.

This is a nice idea but would not work in practice since the fly lines used are different (gnotobiotic vs conventional) and gnotobiotics have to be derived from axenic lines that need a few generations to recover from the bleaching treatment.

(2) The implication that the two forms of selection acting on the immune peptide are maintained by balancing selection is not supported.

The picture presented about how balancing selection is working is rather simplistic and not convincing. In particular, it is not distinguished between fluctuating selection (FL) and balancing selection (BL). BL is the result of negative frequency-dependent selection. It may act within populations (e.g. Red Queen type processes, mating types) or between populations (local adaptation). FL is a process that is sometimes suggested to produce BL, but this is only the case when selection is negative frequency dependent. In most cases, FL does not lead to BL.

The presented study is introduced with a framework of BL, but the aspects investigated are all better described as FL (as the title says: "A suite of selective pressures ..."). The two models presented in the introduction (lines 62 to 69; two pathogens, cost of resistance) are both examples for FL, not for BL.

We’ve added a discussion of how fluctuating selection and balancing selection relate at the end of the discussion.

Finally, no evidence is presented that the different selection pressures suggested to select on the different allelic forms of the immune peptide are acting to produce a pattern of negative frequency dependence.

We are not arguing for negative frequency dependent selection. We assume throughout that Dpt allele does not drive overall frequency of P. rettgeri in populations since it is a ubiquitous microbe. So evolution within D. melanogaster therefore has little to no effect on density of the pathogen.

Recommendations for the authors:

Reviewer #2 (Recommendations For The Authors):

Minor Comments:

Line 31: Rewrite the sentence mentioning "homozygous serine" for improved clarity, especially since the S/R polymorphism of Diptericin has not been introduced yet.

This has been changed to be vague in terms of specific alleles and just refers to “one allele” vs the other.

Lines 87-94: Consider reorganizing this paragraph to maintain a logical flow of the discussion on the Drosophila immune system and the IMD pathway.

We explored other orders, but we think that as is (IMD to AMPs in general to AMPs in Drosophila) makes the most sense here.

Line 99: Provide an explanation of balancing selection for a broader readership, differentiating it from other modes of selection.

We added a brief discussion but note that the intro has significant discussion of balancing selection.

Lines 105-106: Please provide a proper reference. Additionally, ensure that the Unkless et al. 2016 paper is correctly referenced, both in lines 111 and 138-141.

This has been added.

Lines 138-141: It would be beneficial to state that the previous study by Unkless et al. 2016 did not control for genetic background, which is why the assay was redone with gene editing.

This has been added.

Lines 296-303: Clarify the source of the survival observations and consider incorporating this data into Figure 2 for improved visualization.

We’ve clarified that this is Figure 2.

Lines 390-394: Explain the distinctions between vials and cages, particularly in terms of food consumption, exposure to bacteria, etc., which can be relevant to gut homeostasis.

We’ve added a discussion of why these two approaches are complementary.

Reviewer #3 (Recommendations For The Authors):

Statistics

Statistical results are limited to the presentation of p-values (several hundred of them!). For a proper assessment of the statistical analyses, one would also want to see the models used and the test statistics obtained.

The statistical tests done are often unclear. For example, in several experiments, pools of 3 trials (blocs) of multiple animals were tested. The blocs need to be included in the model. Likewise, it seems that multiple delta-dpt fly genotypes were produced. Apparently, they were not distinguished later. Were they considered in the statistical analyses? By contrast, two lines of dptS69R flies were reported to show differences. What concept was applied to test for line difference in some cases and not in others?

In the same dataset (i.e. data resulting from one experiment), it seems that mostly multiple tests were done. For example, in one case each treatment was contrasted to the dptS69 flies. It is generally not acceptable to break down one dataset in multiple subsets and conduct tests with each subtest. One single model for each experiment should be done. This may then be followed by post-hoc tests to see which treatments differ from each other.

We’ve attempted to clarify these statistical approaches throughout.

Minor points

In the legend of Figure 3 it says: "A) monoassociations where each plot represents a different experiment,". This is unclear to me. First, how many plots are there: 3 or 12? Second, what means "experiment"? Are these treatments, or entirely different experiments? How was this statistically taken into account?

We’ve changed this to “different condition” which is clearer. We performed statistical analysis independently for each condition and we’ve now discussed that.

Fig. 5D. It is suggested in the text ("Most intriguing", line 426) and the figure legend that the abundance of Morganellaceae in wild-caught flies differs among genotypes. This is not visible in the figure and not convincingly shown in the text. No stats are given.

We’ve now added that these differences are not significant.

Line 458-461: This sentence is unclear.

We’ve attempted to clarify.

What is a "a traditional adaptive immune system"?

We’ve reworded to “an adaptive immune system”.

There are several typos in the manuscript. Please correct.

We’ve attempted to fix typos throughout.

Bold statements are often without references.

We’ve attempted to add appropriate references throughout.

Author response:

The following is the authors’ response to the original reviews.

Reviewer #1:

R1-01 - Does ank-G-GFP label all isoforms (190, 270 and 480kDa) of ankG? From the images of the AIS and noR it appears that the large forms (270 and 480 kDa) are probably tagged with GFP. Did the authors check for puncta along dendrites and in dendritic spines, which are thought to be formed by the small (190 kDa) isoform? Perhaps a western blot to show that Ank-G-GFP labels all isoforms would be a useful addition to this study.

We believe that AnkG-GFP indeed labels the major Ank3 transcripts in the brain, including the 190, 270, and 480 kDa isoforms, based both on known mRNA exon usage and on Western blot analysis (data not shown). Thus, theoretically, this model would be useful for examining the localization of 190 kD ankyrin-G to dendritic spines. While we attempted to examine this in sections from tissue, it was difficult to separate punctate ankyrinG-GFP labeling from the background. However, these experiments were done in genetic crosses that would label most pyramidal neurons in a given area (i.e. CaMKIIa-Cre). Given the Cre-dependence of this model, future experiments could utilize sparse transduction with a Cre virus that also fills neurons with soluble fluorophores (i.e. mCherry or tdTomato) to mark isolated neurons and identify dendritic spines, as exemplified in Fig. 2D. This would allow examination of subcellular localization of ankyrin-G within single pyramidal cells before and after induction of synaptic plasticity.

R1-02 - In Figure 2, does all the native Ank-G get replaced by Ank-G-GFP? In Fig. 2E the GFP signal along the AIS of CamKII +ve neurons does not appear to be very homogeneous compared to the BIV-spectrin label. Have the authors carried out more experiments like those in 2F, using antibodies that label AnkG together with the GFP fluorescence of the labeled AnkG? It would also be informative to know if, as one might expect, the total levels of ankG-GFP correlate with the levels of ankG at the AIS.

We agree that this is an important point and conducted additional experiments to address your concerns. Of course, we cannot exclude that some unmodified ankyrin-G remains in the AIS or other structures. We expect the turnover of the protein to be rather slow, and native ankyrin-G likely remains to some degree. However, our quantification demonstrates that the ankyrin-G-GFP labeling is sufficiently homogeneous to accurately represent AIS size, indicating proportional levels of GFP to native ankyrin-G. Animals were crossed with a CaMKIIa-Cre driver line and ex vivo slices were imaged live and after immunolabeling. We found a strong correlation between live ankyrin-G-GFP (patch clamp chamber), postfix ankyrin-G-GFP, postfix ankyrin-G, and βIV-spectrin immunosignals of the same AIS. Furthermore, our measurements of AIS length using the intrinsic GFP signal in combination with ankyrin-G, or βIV-spectrin antibodies showed significant overlap (see R103). We now included these graphs as supplemental Fig. S2 in the manuscript (pp. 8-9, ll. 173-177).

R1-03 - Does the length and position of the AIS change when Ank-G is tagged with GFP? This seems like important information that is needed to make sure that there are no structural differences in AIS morphology when compared to native Ank-G.

This is a very important point. We used the βIV-spectrin signal to compare the length of AIS with and without GFP modification in acute slices after patch-clamp recordings (N= 3 animals, 27 GFP+ and 48 GFP- AIS). As secondary control, we plotted the measurements of 160 AIS from a Thy1-GFP mouse line (N = 3 animals, 160 AIS). We found no significant difference in the length and position of the βIV-spectrin signal between GFP positive and negative AIS (p=0.3364 unpaired t-test, p=0.6138 non-parametric Mann-Whitney test, respectively). We have now included this analysis as Supplemental Fig. S2A in the manuscript (pp. 8-9, ll. 173-177). 

R1-04 - How was node length measured in Figure 3? Was this done using the endogenous ank-G signal? In this figure, it would be informative to also quantify the number of noRs with a Nav1.6 stain. Perhaps even check if there are correlations between Ank-G-GFP and Nav1.6 levels. In this figure, it appears that comparisons are carried out between Ank-G-GFP +ve and -ve neurons in the same cryosections, from Ank-G-GFP mice crossed with CamKIIa-Cre. I worry that this may not be comparing the same types of axons. What cells do the CamKIIa -ve axons belong to? Also, the labels on the bar graph are confusing - perhaps GFP+ve and GFP-ve would be clearer?

The reviewer raises an important point. We forgot to declare the signal which was used to measure node length in the manuscript. We have corrected this error and clearly state now in the Fig.3C legend that we used the ankyrin-G signal to quantify node length. Furthermore, using CaMKIIa-Cre mediated expression triggers ankyrin-G-GFP only in a genetically defined subset of neurons. Nodes that do not belong to this subgroup might very well have different node properties. Yet, we cannot assign potential differences in node length to the presence or absence of the GFP label, since we do not have an independent labeling technique for the very same subset of neurons. Since node lengths were similar and showed the same spread of lengths in our sample (Fig. 3C), we assume that the GFP length does probably not affect node length to a significant degree. We have now discussed this limitation in the result (p. 7, ll. 159-165) and method section (p. 30, ll. 644-645) and provide Supplementary Fig. S1 for more clarity. As suggested by the reviewer, we have measured mean fluorescence intensities between 91 GFP+ and 141 GFP- nodes using automated image processing in Imaris. The nodes were again defined by the ankyrin-G signal. We found no difference in length and ellipticity between the groups. We repeated this analysis and compared fluorescence intensities of Nav1.6 and ankyrin-G antibodies and again found no statistical differences between both groups. As suggested by the reviewer, we investigated whether ankyrin-G-GFP interferes with the fluorescence intensities of sodium channels (Nav1.6) and ankyrin-G in general. While the GFP signal showed a strong correlation with ankyrin-G, we found no interdependence with the Nav1.6 signal, indicating that the GFP label does not interfere with the general molecular composition of the nodes. We included these new analyses in Supplemental Fig. S1 (p. 7, ll. 159-165).

R1-05 - In Figure 4 it would also be important to show the distribution of AIS molecules along the AIS, compared to the GFP signal, to establish whether this spatial arrangement of AIS-specific molecules remains intact. For example, Nav1.6 has been described as a more distally-located channel. As the authors point out, the example in A appears to show precisely this feature, but there is no quantification. The same applies to Kv1.2. This would also allow the authors to provide some quantification across multiple AISs, rather than just example images.

We agree that quantifying and comparing AIS-associated proteins would be informative. We measured the intensity profiles of Nav1.6 and Kv2.1 in neighboring AIS and found no preferences for either end of the AIS, neither of GFP-positive nor GFP-negative AIS. We want to note that not all neurons exhibit a distal localization of Nav1.6 and hypothesize that our samples (neocortex layer II) also fall into this group. We included this new graph as Supplemental Fig. S2D and E in the manuscript (p. 9, ll. 180-184).

R1-08 - In Figure 4, did the +Cre condition result in all cells showing a GFP-labelled AIS? If not, were the autocorrelations for +Cre-treated neurons done specifically on cells that expressed AnkG-GFP?

We assume the reviewer refers to the autocorrelation in Figure 6. In this in vitro paradigm, we used virus-induced Cre expression which triggered ankyrin-G-GFP in almost all neurons. The orange boxplots describe the autocorrelation of all ankyrin-G, using a C-terminal antibody as in Fig.6C, but in neurons that also express ankyrin-G-GFP. The green samples use the GFP signal of ankyrin-GFP. We clarified this in the graph and legend of Fig. 6C (pages 14-15).

R1-09 - As mentioned above in Figure 3, the comparisons in Figure 5 (GFP +ve and -ve neurons) may not be comparing like-for-like neurons. I imagine that many of the CamKII+ve cells in the cortex and hippocampus will be GABAergic interneurons, whereas presumably all of the CamKII+ve neurons will be pyramidal cells. Have the authors made sure that they are comparing across the same cell types? The fact that the number of axo-axonic synapses is similar across the two populations (Fig. 5B) does suggest that similar neuron types (presumably pyramidal cells) were compared in the hippocampus, but some other way of making sure would be a nice addition.

We agree with the reviewer that the grey and green boxes are not sampled from the same subset of neurons, since only CaMKIIa-positive principal cells will express ankyrin-G-GFP. However, we are confident that the selected AIS belong to pyramidal neurons in both cases. Principal neurons can be well distinguished from interneurons not only by the size, shape, and position of their somas but also by the length and thickness of their AIS. We have performed previous studies on the AIS of interneurons using genetic GAD and parvalbumin markers. Thus, we are confident that the plots in 5A and 5B are sampled from pyramidal neurons, though certainly from genetically different subsets. We now highlight and discuss this limitation in the result section (p. 11, ll. 215-217) and modified the graph in Fig. 5A and 5B for clarity.

R1-10 - In Figure 6, what was the promoter for the DCre and Cre+ lentivirus? Was this also driven by CamKIIa? In culture it is not always easy to be sure of neuronal identity - did the authors try to bias their analysis to specific neuronal types?

Indeed, the nature of the promotor was not stated in the legend or method section, which we now corrected. We used lentiviral FUW-nGFP-Cre and FUW-nGFP-ΔCre constructs to trigger ankyrin-G-GFP expression. Both viruses use the CMV (Cytomegalovirus) promoter, which drives constitutively high levels of gene expression in a wide range of cell types, including neuronal cells. The majority of neurons in dissociated hippocampal cultures are excitatory, especially larger cells with larger AIS, which were preferably used in the analysis. Thus, we cannot claim that AIS nanostructure is intact in cultured interneurons, but this is also true for in vivo conditions in general. Since mice did not show any obvious behavioral phenotypes, we are positive that interneuron functionality is preserved. We also note that the parallel expression of nuclear GFP in the infected neurons was undesired, but did not impact STED imaging due to that technique’s high resolution. 

R1-11 - The ability to visualize the plasticity of the AIS in real-time is an important advance in the field. The loss of proximal Ank-G-GFP signal upon local application of 15 mM KCl is particularly interesting. The fact that neighboring AISs are not affected is surprising - do the authors know how local their KCl application was? Also, although the neighboring AISs are a nice control, the one control lacking here is the local application of normal solution (preferably 15 mM NaCl to account for osmolarity changes) to make sure that this does not affect the properties of the AIS.

We used KCl puffs in previous, unrelated experiments where we observed that only cells directly in front of the pipette are visibly depolarized by an acute KCl puff (measured by patch-clamp). Due to technical limitations, patched and live imaged neurons were generally in the first 2-5 cell layers of the brain slice, which is well perfused by the constant flow of oxygenated ACSF. KCl is thus quickly diluted and carried away. We have visualized the concentration gradients via puff application by puffing the fluorescent marker fluorescein in the same recording condition. The cone of fluorescence was only visible in front of the pipette and vanished in less than a second post-pressure application. To verify that it is indeed KCl and not the mechanical stress that lead to the loss of proximal Ank-G-GFP, one would indeed need an ACSF puff control, which we did for other studies. However, this is not the point we wanted to make. Instead of studying live single-cell AIS plasticity, we want to demonstrate that such investigations are generally possible using the ankyrin-G-GFP line.

Author response image 1.

R1-12 - The ability to be able to image AISs in vivo is another important finding. Were the authors able to image noRs as well?

We believe that this is indeed the case. The panels in Figure 9C contain densely labeled puncta that also remain in position from week 1 to week 2. These are likely nodes of Ranvier, although we do not have the means to verify their presence at this time.

Reviewer #2:

R2-01 - Are there indeed different Ank-G-GFP isoforms expressed in this model and could they correspond to classical neuronal Ank-G isoforms?

This is an important issue that was also raised by reviewer #1. Please consult the respective section R1-01 above for our response.

R2-02 - What is the rationale of doing Ank-G co-labelling in the case of Ank-G-GFP expression, rather than Pan-Nav staining for example? The co-staining with Nav1.6 antibody, when present, is however convincing.

We used the co-labeling to emphasize that the ankyrin-G-GFP construct allows reliable investigation of the whole AIS. This is why we wanted to demonstrate that the ankyrin-G-GFP signal overlaps with other AIS markers, as well as all ankyrin-G in general (including potentially remaining native and unlabeled ankyrin-G). This was also a point raised by Reviewer 1, which is why we provided some additional graphs (see response R1-02). However, we agree that staining with another independent marker, such as Nav1.6 or βIVspectrin was necessary. 

R2-03 - Figure 2D and F: what is the rationale for not using betaIV-Spectrin staining as in the other panels of this figure? Furthermore, could betaIV-Spectrin localization be affected by Ank-GGFP expression, as betaIV-Spectrin is known to depend on Ank-G for its AIS targeting? Are there any other AIS markers, which localization is known to be independent of Ank-G, that could have been used?

We have compiled this figure from a multitude of different experimental setups from different labs to showcase the reliability and robustness of the ankyrin-G-GFP label. This is why the type of staining is not consistent among panels. However, we provide some quantification on the possible impact of ankyrin-G-GFP expression on the βIV-spectrin signal and the composition of the AIS in general. The STED image verifies that the basic subcellular arrangement of the cytoskeleton, including βIV-spectrin, remains intact (Fig. 6). Most AIS markers are at least in some way dependent on ankyrin-G expression, but FGF14 and neurofascin may be the most independent candidates (Fig. 4).

R2-04 - Did the authors measure the mean AIS length and distance from cell soma in Ank-G-GFPexpressing neurons versus non-expressing ones (considering the same neuronal subtypes) to assess whether these were unaffected by Ank-G-GFP expression?

This is an important point that was also raised by Reviewer 1 (see also our comments to R1-03). We have included this analysis now in the manuscript as Supplemental Fig. S2A (pp. 8-9, ll. 173-177).

R2-05 - Figure 5C: the microglial staining and 3D reconstruction could have been clearer.

We have modified the image and 3D rendering to make Figure 5C clearer to the reader. We hope that our changes suffice.

R2-06 - Figure 8: do hippocampal neurons retain their electrophysiological properties after 20 DIV? It could strengthen this part of the work to have access to the electrophysiological data mentioned in the text. 

This is an important issue. We did not perform any electrophysiological recordings in OTCs in the course of this study. Panel E uses acute hippocampal slices like in Fig. 7. We have performed patch-clamp experiments up to DIV 10 for an unrelated study (see graph for action potential firing, Author response image 2). There are not many studies performing electrophysiology in slice cultures due to the formation of a glial scar on top of the slices. However, multielectrode array (MEA) recordings demonstrated that hippocampal organotypic slice cultures remain viable and show electric activity past DIV 20 (though with decreased viability and activity). We kindly refer to the following publications on that matter:

Author response image 2.

Sample traces of action potentials triggered by cuttrent injections

Gong W, Senčar J, Bakkum DJ, Jäckel D, Obien ME, Radivojevic M, Hierlemann AR. Multiple SingleUnit Long-Term Tracking on Organotypic Hippocampal Slices Using High-Density Microelectrode Arrays. Front Neurosci. 2016 Nov 22;10:537. doi: 10.3389/fnins.2016.00537. PMID: 27920665; PMCID: PMC5118563.

Mohajerani MH, Cherubini E. Spontaneous recurrent network activity in organotypic rat hippocampal slices. Eur J Neurosci. 2005 Jul;22(1):107-18. doi: 10.1111/j.1460-9568.2005.04198.x. PMID: 16029200.

Author response:

The following is the authors’ response to the original reviews.

Public Reviews:

Reviewer #1 (Public Review):

In the manuscript, the authors explore the mechanism by which Taenia solium larvae may contribute to human epilepsy. This is extremely important question to address because T. solium is a significant cause of epilepsy and is extremely understudied. Advances in determining how T. solium may contribute to epilepsy could have significant impact on this form of epilepsy. Excitingly, the authors convincingly show that Taenia larvae contain and release glutamate sufficient to depolarize neurons and induce recurrent excitation reminiscent of seizures. They use a combination of cutting-edge tools including electrophysiology, calcium and glutamate imaging, and biochemical approaches to demonstrate this important advance. They also show that this occurs in neurons from both mice and humans. This is relevant for pathophysiology of chronic epilepsy development. This study does not rule out other aspects of T. solium that may also contribute to epilepsy, including immunological aspects, but demonstrates a clear potential role for glutamate.

Strengths:

- The authors examine not only T. solium homogenate, but also excretory/secretory products which suggests glutamate may play a role in multiple aspects of disease progression.

- The authors confirm that the human relevant pathogen also causes neuronal depolarization in human brain tissue

- There is very high clinical relevance. Preventing epileptogenesis/seizures possibly with Glu-R antagonists or by more actively removing glutamate as a second possible treatment approach in addition to/replacing post-infection immune response.

- Effects are consistent across multiple species (rat, mouse, human) and methodological assays (GluSnFR AND current clamp recordings AND Ca imaging)

- High K content (comparable levels to high-K seizure models) of larvae could have also caused depolarization. Adequate experiments to exclude K and other suspected larvae contents (i.e. Substance P).

Weaknesses:

- Acute study is limited to studying depolarization in slices and it is unclear what is necessary/sufficient for in vivo seizure generation or epileptogenesis for chronic epilepsy. - There is likely a significant role of the immune system that is not explored here. This issue is adequately addressed in the discussion, however, and the glutamate data is considered in this context.

Discuss impact:

- Interfering with peri-larval glutamate signaling may hold promise to prevent ictogenesis and chronic epileptogenesis as this is a very understudied cause of epilepsy with unknown mechanistic etiology.

Additional context for interpreting significance:

- High medical need as most common adult onset epilepsy in many parts of the world

We thank Reviewer 1 for their positive and thorough assessment of our manuscript. We have elected to respond to and address the following aspects from their “Recommendations For The Authors” below:

Reviewer #1 (Recommendations For The Authors):

Additional experiments/analysis:

-   Fig 4a-c: Larva on a slice and not next to it? Negative results maybe because its E/S products are just washed away (assuming submerged recording chamber/conditions)? Experiments and negative results described here do not seem conclusive. Should be discussed at least?

We agree with the reviewer and have added the following sentence to the relevant section of the Results: ‘Our submerged recording setup might have led to swift diffusion or washout of released glutamate, possibly explaining the lack of observable changes.’

Writing & presentation:

- Data is not always reported consistently in text and figures, examples:

- Results in text are reported varyingly without explanation:

- Mean and/or median? SEM or SD and/or IQR? Stat info included in text or not? i.e. lines 130/131 vs. 160/161

Results and data are now presented in a more uniform fashion. We report medians and IQRs, sample size, statistical test result, statistical test used in that order.

- Larval release data interrupts reading flow, lines 246-252 double up results presented in Fig 5F.

This section has now been significantly abbreviated and reads as follows: ‘T. crassiceps larvae released a relatively constant median daily amount of glutamate, ranging from 41.59 – 60.15 ug/20 larvae, which showed no statistically significant difference across days one to six. Similarly, T. crassiceps larvae released a relatively constant median daily amount of aspartate, ranging from 9.431 – 14.18 ug/20 larvae, which showed no statistically significant difference across days one to six.’

- Results in figures are reported in different styles:

Results have now been made uniform, reporting medians and IQRs and: sample size, p test result, statistical test used, figure # reported in that order.

- Fig 6: E/S glu concentration seems to be significantly higher in solium vs crassiceps (about 6fold higher in solium). Should be discussed at least.

Given the small sample size from T. solium (see response below), we do not draw attention to this difference and instead simply make the point that T. solium larvae contain and release glutamate.

- In this context - N=1 may be sufficient for proof of principle (release) but seems too small of a cohort to describe non-constant release of glu over days (Fig 6D). Is initial release on day 1, no release and recovery in the following days reproducible? Is very high glu content of E/S content (15-fold higher in comparison to solium homogenate AND 6-fold higher in comparison to crassiceps homogenate and E/S content). Not sure if Fig 6D is adding relevant information, especially since it is based on n = 1

We agree that a N=1 is only sufficient for proof of principle. However it is worth noting that the measurements still reflect the cumulative release from 20 larvae. Nonetheless, the statement in text has been simplified to say: ‘These results demonstrate that T. solium larvae continually release glutamate and aspartate into their immediate surroundings.’ As this focusses on the point that the larvae release glutamate and aspartate continuously and that we can’t draw conclusions about the variability over days.

Methods:

- Human slices, mention cortex - what part, patient data would be interesting. I.e. etiology of epilepsy, epilepsy duration 

In the Materials and Methods section “Brain slice preparation” we have now added a table with the requested information.

- For Taenia solium: How were they acquired and used in these experiments?

In the Materials and Methods section “Taenia maintenance and preparation of whole cyst homogenates and E/S products” we describe how Taenia solium larvae were acquired and used.

- Was access resistance monitored? Add exclusion criteria for patch experiments

Figure supplement tables containing the basic properties for each cell recording have been added for each figure and the following statements were added to the electrophysiology section of the Methods: ‘Basic properties of each cell were recorded (supplementary files 1, 2, 3, 4, 6).’ and ‘Cells were excluded from analyses if the Ra was greater than 80 Ω or if the resting membrane potential was above –40 mV.’  

- Cannot see any reference to mouse slices in methods? Also, mouse organotypic cultures (for AAV?)? Or only acute slices from mice and organotypic hip cultures from rats? Seems to have been mouse and rat organotypic cultures? But not clear with further clarification in methods.

We have now added the following clarification to the methods: ‘For experiments using calcium and glutamate imaging mouse hippocampal organotypic brain slices were used. For all other experiments rat hippocampal organotypic brain slices were used. A subset of experiments used acute human cortical brain slices and are specified.’

- How long after the wash-in phase was the wash-out phase data collected?

For wash-in recordings drugs were washed in for 8 mins before recordings were made. Drugs were washed out for at least 8 mins before wash-out recordings were made. This information has been added to the Materials and Methods section.

- In general, the M&M section seems to have been written hastily - author's internal remarks "supplier?" are still present.

The M&M section has been thoroughly proofread for errors and internal remarks removed or corrected.

- A little more information on the clinical subjects would be appreciated. I.e. duration of epilepsy? Localization? What cortex? Usual temporal lobe or other regions?

We have now added a table with this information to the Materials and Methods section “Brain slice preparation”.

Minor corrections text/figures:

- i.e. 3D,F,H,J show individual data points, thats great, but maybe add mean/median marker (as results are reported like this in text)  like in fig 4G,I and others

Figures 3D,F,H & J have been revised to include median and IQR.

- Only one patient mentioned in acknowledgements, but 2 in methods and text

We apologize for this oversight and now acknowledge both patients in the acknowledgements.

- Fig 1 B-F individual puffs are described as increasing - consistent with cellular effects (1st puff depolarizes, 2nd puff elicits 1 AP, 3rd puff elicits AP burst)  However, dilution ratio of homogenate or puff concentrations are not mentioned (or potentially longer than 20 ms puffs for 2nd and 3rd stimulus?) in text or figures. Seems to be enough space to indicate in figure as well (i.e. multiple or thicker arrows for subsequent puffs or label with homogenate dilution/concentration in figure).

We state in the results section associated with Fig. 1 that increasing the amount of homogenate delivered was achieved by increasing the pressure applied to the ejection system. We now include this information in the figure legend.

- Figure legend describes 30 ms puff for Ca imaging whereas ephys data (from text) is 20 ms puff. Was Ca imaging performed in acute mouse hippocampal slices (as figure text suggests) or were those organotypic hippocampal cultures from mice?

Ca2+  imaging was performed in mouse hippocampal organotypic brain slice cultures. The figure text for Fig. 1 E) states “widefield fluorescence image of neurons in the dentate gyrus of a mouse hippocampal organotypic brain slice culture expressing the genetically encoded Ca2+ reporter GCAMP6s...”

- 11.4 mM K is reported for homogenate in text only. How variable is that? How many n? No SD reported in text and no individual data points reported since this experiment is not represented as a figure.

This has been clarified in the text by adding (N = 1, homogenate prepared from >100 larvae).

- Same results (effect of 11.4 mM K on Vm) described twice in one paragraph, compare lines 126-131 with 131-136.

The repetition has been removed.

- Line 182 - example for consistency: decide IQR or SD/SEM

To improve consistency, we have changed to median and IQR throughout.

- Neuronal recordings are reported as hippocampal pyramidal neurons (i.e. line 222) but some recordings were made from dentate granule cells - please clarify which neurons were recorded in ephys, ca imaging, GluSnFr imaging

For each experiment we describe which type of neurons were recorded from. For rodent recordings these were hippocampal pyramidal neurons except in the case of the Ca2+ imaging example where the widefield recording was over the dentate gyrus subfield.

- Line 309: "should" seems to be an extra word

We have removed the word ‘should’ and made the sentence shorter and clearer. It now reads: ‘Given our finding that cestode larvae contain and release significant quantities of glutamate, it is possible that homeostatic mechanisms for taking up and metabolizing glutamate fail to compensate for larvalderived glutamate in the extracellular space. Therefore, similar glutamate-dependent excitotoxic and epileptogenic processes that occur in stroke, traumatic brain injury and CNS tumors are likely to also occur in NCC.’

Reviewer #2 (Public Review):

Since neurocysticercosis is associated with epilepsy, the authors wish to establish how cestode larvae affect neurons. The underlying hypothesis is that the larvae may directly excite neurons and thus favor seizure genesis.

To test this hypothesis, the authors collected biological materials from larvae (from either homogenates or excretory/secretory products), and applied them to hippocampal neurons (rats and mice) and human cortical neurons.

This constitutes a major strength of the paper, providing a direct reading of larvae's biological effects. Another strength is the combination of methods, including patch clamp, Ca, and glutamate imaging.

We thank the Reviewer 2 for their review of the strength and weaknesses of our manuscript. We respond to the identified weaknesses below.

There are some weaknesses:

(1) The main one relates to the statement: "Together, these results indicate that T. crassiceps larvae homogenate results not just in a transient depolarization of cells in the immediate vicinity of application, but can also trigger a wave of excitation that propagates through the brain slice in both space and time. This demonstrates that T. crassiceps homogenate can initiate seizurelike activity under suitable conditions."

The only "evidence" of propagation is an image at two time points. It is one experiment, and there is no quantification. Either increase n's and perform a quantification, or remove such a statement.

We acknowledge that the data is from one experiment, with the intention of demonstrating that it is plausible for intense depolarization of a subset of neurons to result in the initiation and propagation of seizure-like activity to nearby neurons under suitable conditions. However, we agree that it is prudent to remove this statement and have done so.

Likewise, there is no evidence of seizure genesis. A single cell recording is shown. The presence of a seizure-like event should be evaluated with field recordings.

In this experiment the Ca2+ imaging demonstrates activity spreading from the site of the restricted homogenate puff to all surrounding neurons. Furthermore, the whole-cell recoding is typical of a slice wide seizure-like event.  

(2) Control puff experiments are lacking for Fig 1. Would puffing ACSF also produce a depolarization, and even firing, as suggested in Fig. 2D? This is needed for at least one species.

We agree and have added this data for the rat and mouse neuron in a new Figure 1-figure supplement 1.

(3) What is the rationale to use a Cs-based solution? Even in the presence of TTX and with blocking K channels, the depolarization may be sufficient to activate Ca channels (LVGs), which would further contribute to the depolarization. Why not perform voltage clamp recordings to directly the current?

The intention of the Cs-based solution was to block K+ channels and reduce the effect of moderately raised K+ in the homogenate to isolate the contribution of other causative agents of depolarization (i.e. glutamate / aspartate). We agree that performing voltage clamp recordings would have been useful for directly recording the currents responsible for depolarization. 

(4) Why did you use organotypic slices? Since you wish to model adult epilepsy, it would have been more relevant to use fresh slices from adult rats/mice. At least, discuss the caveat of using a network still in development in vitro.

Recordings were performed 6–14 days post culture, which is equivalent to postnatal Days (P) 12 to 22. Previous work has shown that neurons in the organotypic hippocampal brain slice are relatively mature (Gähwiler et al., 1997). For example they possess mature Cl- homeostasis mechanisms at this point, as evidenced by their hyperpolarizing EGABA (Raimondo et al., 2012).  

(5) Please include both the number of slices and number of cells recorded in each condition. This is the standard (the number of cells is not enough).

This has now been added to all relevant sections of the results text.  

(6) Please provide a table with the basic properties of cells (Rin, Rs, etc.). This is standard to assess the quality of the recordings.

Tables containing the basic properties for each cell recording have been created for each figure (as Figure supplements) and the following statement was added to the electrophysiology section of the Methods: ‘Basic properties of each cell were recorded (see Figure supplements).’

(7) Please provide a table on patient's profile. This is standard when using human material. Were these TLE cases (and "control" cortex) or epileptogenic cortex?

We have now added a basic table on the patient’s profiles to the Materials and Methods section.

Globally, the authors achieved their aims. They show convincingly that larvae material can depolarize neurons, with glutamate (and aspartate) as the most likely candidates.

This is important not only because it provides mechanistic insight but also potential therapeutic targets. The result is impactful, as the authors use quasi-naturalistic conditions, to assess what might happen in the human brain. The experimental design is appropriate to address the question. It can be replicated by any interested person.

We thank the Reviewer 2 for their enthusiastic and constructive assessment of our manuscript. We have elected to respond to and address the following aspects from their “Recommendations For The Authors” below:

Reviewer #2 (Recommendations For The Authors):

lines 132 and following are a repetition of those above

These have been removed.

line 151 Fig "2" missing

This has been added.

187, 190 should be E, F not C, D

This has been changed in the text.  

481, 482 supplier?

This has been corrected and the correct suppliers described.

Reviewer #3 (Public Review):

This paper has high significance because it addresses a prevalent parasitic infection of the nervous system, Neurocysticercosis (NCC). The infection is caused by larvae of the parasitic cestode Taenia solium It is a leading cause of epilepsy in adults worldwide

To address the effects of cestode larvae, homogenates and excretory/secretory products of larvae were added to organotypic brain slice cultures of rodents or layer 2/3 of human cortical brain slices from patients with refractory epilepsy.

We thank Reviewer 3 for their helpful comments and suggestions for improvement which we address below.

A self-made pressure ejection system was used to puff larvae homogenate (20 ms puff) onto the soma of patched neurons. The mechanical force could have caused depolarizaton so a vehicle control is critical. On line 150 they appear to have used saline in this regard, and clarification would be good. Were the controls here (and aCSF elsewhere) done with the low Mg2+o aCSF like the larvae homogenates?

We agree and have added examples where aCSF alone was pressure ejected onto the same rat and mouse neurons in a new Figure 1-figure supplement 1. In Figure 1, the same aCSF as that was used to bathe the slices was used. In Figure 2D-G, either PBS (which larval homogenates were prepared in) or growth medium (which contain larval E/S products) were used as comparative controls.

They found that neurons depolarized after larvae homogenate exposure and the effect was mediated by glutamate but not nicotinic receptors for acetylcholine (nAChRs), acid-sensing channels or substance P. To address nAChRs, they used 10uM mecamyline, and for ASICs 2mM amiloride which seems like a high concentration. Could the concentrations be confirmed for their selectivity? 

We did not independently verify the selectivity of the antagonist concentrations used in our study. However, the persistence of depolarizations despite the use of high concentrations of mecamylamine (10 μM) and amiloride (2 mM) provides strong evidence that neither nAChRs nor ASICs are primarily responsible for mediating these responses. The high concentrations used, while potentially raising concerns about specificity, actually strengthen our conclusion that these receptor types are not involved in the observed effect.

Glutamate receptor antagonists, used in combination, were 10uM CNQX, 50uM DAP5, and 2mM kynurenic acid. These concentrations are twice what most use. Please discuss. 

We intentionally used higher-than-typical concentrations of glutamate receptor antagonists in our experimental design. Our rationale for this approach was to ensure maximal blockade of glutamate receptors, thereby minimizing the possibility of residual receptor activity confounding our results.

Also, it would be very interesting to know if the glutamate receptor is AMPA, Kainic acid, or NMDA. Were metabotropic antagonists ever tested? That would be logical because CNQX/DAPR/Kynurenic acid did not block all of the depolarization.

We appreciate the reviewer's interest in the specific glutamate receptor subtypes involved in our study. Our research primarily focused on ionotropic glutamate receptors as a group, without differentiating the individual contributions of AMPA, Kainate, and NMDA receptors. This approach, while broad, allowed us to establish the involvement of glutamatergic signalling in the observed effects. We acknowledge that we did not investigate metabotropic glutamate receptors in this study. Importantly, we demonstrate later in our manuscript that the larval products contain both glutamate and aspartate. Therefore the precise nature of the glutamate-dependent depolarization observed using a particular experimental preparation would depend on the specific types of neurons exposed to the homogenate and the expression profile of different glutamate receptor subtypes on these neurons.

They also showed the elevated K+ in the homogenate (~11 mM) could not account for the depolarization. However, the experiment with K+ was not done in a low Mg2+o buffer (Or was it -please clarify). 

The experiment where 11.39 mM K+ as well as the experiment with T. crass. Homogenate with a cesium internal and added TTX were all done in standard 2 mM Mg2+ containing aCSF.

They also confirmed that only small molecules led to the depolarization after filtering out very large molecules. That supports the conclusion that glutamate - which is quite small - could be responsible. It is logical to test substance P because the Intro points out prior work links the larvae and seizures by inflammation and implicates substance P. However, why focus on nAChRs and ASIC?

These were chosen as they are ionotropic receptors which mediate depolarization and hence could conceivably be responsible for the homogenate-induced depolarization we observed.

The depolarizations caused seizure-like events in slices. The slices were exposed to a proconvulant buffer though- low Mg2+o. This buffer can cause spontaneous seizure-like events so it is important to know what the buffer did alone.

We agree that a low M2+ buffer solution can elicit seizure-like events in organotypic slices alone. However, the timing of the onset of the seizure-like event in the example presented in Figure 1 strongly suggests that it was triggered by the T. crass homogenate puff. Nonetheless, on the suggestion of the other reviewers we have reduced emphasis on our experimental evidence for the ability of T. crass. homogenate to illicit seizure-like events.  

They suggest the effects could underlie seizure generation in NCC. However, there is only one event that is seizure-like in the paper and it is just an inset. Were others similar? How frequency were they? How long?

Please see the response above as well as our response to Reviewer 1 who raised a similar concern.

Using Glutamate-sensing fluorescent reporters they found the larvae contain glutamate and can release it, a strength of the paper.

Fig. 4. Could an inset be added to show the effects are very fast? That would support an effect of glutamate.

We have not added an inset. However, given the scale bar (500 ms) for the trace provided, the response is very fast.  

Why is aspartate relatively weak and glutamate relatively effective as an agonist?

Glutamate generally has a higher affinity for glutamate receptors compared to aspartate. This is particularly true for AMPA and kainate receptors, where glutamate is the primary endogenous agonist. Similarly iGluSnFR has a higher sensitivity for glutamate over aspartate (Marvin et al., 2013).

Could some of the variability in Fig 4G be due to choice of different cell types? That would be consistent with Fig 5B where only a fraction of cells in the culture showed a response to the larvae nearby. 

Whilst differences in cell types could contribute to the variability in Fig 4G, all the responses were recorded from hippocampal pyramidal neurons and hence it is more likely that the variability is a function of other sources of variation including differences in iGluSnFR expression, depth of the cell imaged, the proximity of the puffer pipette etc. In Fig. 5B we think the lack of response may be due to the fact that any released glutamate by the live larvae was not able reach the iGluSnFR neurons at sufficient concentrations due to the nature of our submerged recording setup. We have added the following sentence to the results. ‘Our submerged recording setup might have led to swift diffusion or washout of released glutamate, possibly explaining the lack of observable changes.’

On what basis was the ROI drawn in Fig. 5B.

The ROI drawn in Fig. 5B was selected to include all iGluSnFR expressing neurons in the brain slice. which were captured in the field of view.

Also in 5B, I don't see anything in the transmitted image. What should be seen exactly?

We agree that it is difficult to resolve much in the transmitted image. However, both the brain slice on the left as well as a T. crass. larva on the right is visible and outlined with a green or orange dashed line respectively.

Human brain slices were from temporal cortex of patients with refractory epilepsy. Was the temporal cortex devoid of pathology and EEG abnormalities? This area may be quite involved in the epilepsy because refractory epilepsy that goes to surgery is often temporal lobe epilepsy. Please discuss the limitations of studying the temporal cortex of humans with epilepsy since it may be more susceptible to depolarizations of many kinds, not just larvae.

We acknowledge the important limitations of using temporal cortex tissue from patients with refractory epilepsy. While we aimed to use visually normal tissue, we recognize that the tissue may have underlying pathology or functional abnormalities not visible to the naked eye. It may also be more susceptible to induced depolarizations due to epilepsy-related changes in neuronal excitability. Despite these limitations, we believe our human tissue data still provides valuable data that the larval homogenates can induce depolarization in human as well as rodent neurons.  

Please discuss the limitations of the cultures - they are from very young animals and cultured for 6-14 days.

We acknowledge the potential limitations of our experimental model using organotypic hippocampal slice cultures from young animals. The use of relatively immature tissue may not fully represent the adult nervous system due to developmental differences in receptor expression, synaptic connections, and network properties. The 6-14 day culture period, while allowing some maturation, may induce changes that differ from the in vivo environment, including alterations in cellular physiology and network reorganization. Despite these limitations, this model provides a valuable balance between preserved local circuitry and experimental accessibility. Future studies comparing results with acute adult slices and in vivo models would be beneficial to validate and extend our findings.

References:

Gähwiler, B.H. et al. (1997) ‘Organotypic slice cultures: a technique has come of age.’, Trends in neurosciences, 20(10), pp. 471–7.

Marvin, J.S. et al. (2013) ‘An optimized fluorescent probe for visualizing glutamate neurotransmission.’, Nature methods, 10(2), pp. 162–70. Available at: https://doi.org/10.1038/nmeth.2333.

Raimondo, J.V. et al. (2012) ‘Optogenetic silencing strategies differ in their effects on inhibitory synaptic transmission.’, Nat. Neurosci., 15(8), pp. 1102–4. Available at: https://doi.org/10.1038/nn.3143.

Author response:

Public Reviews:

Reviewer #1 (Public review):

Summary:

The authors describe a method to probe both the proteins associated with genomic elements in cells, as well as 3D contacts between sites in chromatin. The approach is interesting and promising, and it is great to see a proximity labeling method like this that can make both proteins and 3D contacts. It utilizes DNA oligomers, which will likely make it a widely adopted method. However, the manuscript over-interprets its successes, which are likely due to the limited appropriate controls, and of any validation experiments. I think the study requires better proteomic controls, and some validation experiments of the "new" proteins and 3D contacts described. In addition, toning down the claims made in the paper would assist those looking to implement one of the various available proximity labeling methods and would make this manuscript more reliable to non-experts.

Strengths:

(1) The mapping of 3D contacts for 20 kb regions using proximity labeling is beautiful.

(2) The use of in situ hybridization will probably improve background and specificity.

(3) The use of fixed cells should prove enabling and is a strong alternative to similar, living cell methods.

Weaknesses:

(1) A major drawback to the experimental approach of this study is the "multiplexed comparisons". Using the mtDNA as a comparator is not a great comparison - there is no reason to think the telomeres/centrosomes would look like mtDNA as a whole. The mito proteome is much less complex. It is going to provide a large number of false positives. The centromere/telomere comparison is ok, if one is interested in what's different between those two repetitive elements. But the more realistic use case of this method would be "what is at a specific genomic element"? A purely nuclear-localized control would be needed for that. Or a genomic element that has nothing interesting at it (I do not know of one). You can see this in the label-free work: non-specific, nuclear GO terms are enriched likely due to the random plus non-random labeling in the nucleus. What would a Telo vs general nucleus GSEA look like? (GSEA should be used for quantitative data, no GO). That would provide some specificity. Figures 2G and S4A are encouraging, but a) these proteins are largely sequestered in their respective locations, and b) no validation by an orthogonal method like ChIP or Cut and Run/Tag is used.

You can also see this in the enormous number of "enriched" proteins in the supplemental volcano plots. The hypothesis-supporting ones are labeled, but do the authors really believe all of those proteins are specific to the loci being looked at? Maybe compared to mitochondria, but it's hard to believe there are not a lot of false positives in those blue clouds. I believe the authors are more seeing mito vs nucleus + Telo than the stated comparison. For example, if you have no labeling in the nucleus in the control (Figures 1C and 2C) you cannot separate background labeling from specific labeling. Same with mito vs. nuc+Telo. It is not the proper control to say what is specifically at the Telo.

I would like to see a Telo vs nuclear control and a Centromere vs nuc control. One could then subtract the background from both experiments, then contrast Telo vs Cent for a proper, rigorous comparison. However, I realize that is a lot of work, so rewriting the manuscript to better and more accurately reflect what was accomplished here, and its limitations, would suffice.

(2) A second major drawback is the lack of validation experiments. References to literature are helpful but do not make up for the lack of validation of a new method claiming new protein-DNA or DNA-DNA interactions. At least a handful of newly described proximal proteins need to be validated by an orthogonal method, like ChIP qPCR, other genomic methods, or gel shifts if they are likely to directly bind DNA. It is ok to have false positives in a challenging assay like this. But it needs to be well and clearly estimated and communicated.

(3) The mapping of 3D contacts for 20 kb regions is beautiful. Some added discussion on this method's benefits over HiC-variants would be welcomed.

(4) The study claims this method circumvents the need for transfectable cells. However, the authors go on to describe how they needed tons of cells, now in solution, to get it to work. The intro should be more in line with what was actually accomplished.

(5) Comments like "Compared to other repetitive elements in the human genome...." appear to circumvent the fact that this method is still (apparently) largely limited to repetitive elements. Other than Glopro, which did analyze non-repetitive promoter elements, most comparable methods looked at telomeres. So, this isn't quite the advancement you are implying. Plus, the overlap with telomeric proteins and other studies should be addressed. However, that will be challenging due to the controls used here, discussed above.

We thank the Reviewer for their careful reading of manuscript and constructive suggestions. We plan to substantially revise the framing and presentation of manuscript to address the concerns raised by all three reviewers.

Reviewer #2 (Public review):

Summary

Liu and MacGann et al. introduce the method DNA O-MAP that uses oligo-based ISH probes to recruit horseradish peroxidase for targeted proximity biotinylation at specific DNA loci. The method's specificity was tested by profiling the proteomic composition at repetitive DNA loci such as telomeres and pericentromeric alpha satellite repeats. In addition, the authors provide proof-of-principle for the capture and mapping of contact frequencies between individual DNA loop anchors.

Strengths

Identifying locus-specific proteomes still represents a major technical challenge and remains an outstanding issue (1). Theoretically, this method could benefit from the specificity of ISH probes and be applied to identify proteomes at non-repetitive DNA loci. This method also requires significantly fewer cells than other ISH- or dCas9-based locus-enrichment methods. Another potential advantage to be tested is the lack of cell line engineering that allows its application to primary cell lines or tissue.

Weaknesses

The authors indicate that DNA O-MAP is superior to other methods for identifying locus-specific proteomes. Still, no proof exists that this method could uncover proteomes at non-repetitive DNA loci. Also, there is very little validation of novel factors to confirm the superiority of the technique regarding specificity.

The authors first tested their method's specificity at repetitive telomeric regions, and like other approaches, expected low-abundant telomere-specific proteins were absent (for example, all subunits of the telomerase holoenzyme complex). Detecting known proteins while identifying noncanonical and unexpected protein factors with high confidence could indicate that DNA O-MAP does not fully capture biologically crucial proteins due to insufficient enrichment of locus-specific factors. The newly identified proteins in Figure 1E might still be relevant, but independent validation is missing entirely. In my opinion, the current data cannot be interpreted as successfully describing local protein composition.

Finally, the authors could have discussed the limitations of DNA O-MAP and made a fair comparison to other existing methods (2-5). Unlike targeted proximity biotinylation methods, DNA O-MAP requires paraformaldehyde crosslinking, which has several disadvantages. For instance, transient protein-protein interactions may not be efficiently retained on crosslinked chromatin. Similarly, some proteins may not be crosslinked by formaldehyde and thus will be lost during preparation (6).

(1) Gauchier M, van Mierlo G, Vermeulen M, Dejardin J. Purification and enrichment of specific chromatin loci. Nat Methods. 2020;17(4):380-9.

(2) Dejardin J, Kingston RE. Purification of proteins associated with specific genomic Loci. Cell. 2009;136(1):175-86.

(3) Liu X, Zhang Y, Chen Y, Li M, Zhou F, Li K, et al. In Situ Capture of Chromatin Interactions by Biotinylated dCas9. Cell. 2017;170(5):1028-43 e19.

(4) Villasenor R, Pfaendler R, Ambrosi C, Butz S, Giuliani S, Bryan E, et al. ChromID identifies the protein interactome at chromatin marks. Nat Biotechnol. 2020;38(6):728-36.

(5) Santos-Barriopedro I, van Mierlo G, Vermeulen M. Off-the-shelf proximity biotinylation for interaction proteomics. Nat Commun. 2021;12(1):5015.

(6) Schmiedeberg L, Skene P, Deaton A, Bird A. A temporal threshold for formaldehyde crosslinking and fixation. PLoS One. 2009;4(2):e4636.

We thank the Reviewer for their constructive feedback on our work. As noted above, we plan to substantially revise the framing and presentation of manuscript to address the concerns raised by all three reviewers.

Reviewer #3 (Public review):

Significance of the Findings:

The study by Liu et al. presents a novel method, DNA-O-MAP, which combines locus-specific hybridisation with proximity biotinylation to isolate specific genomic regions and their associated proteins. The potential significance of this approach lies in its purported ability to target genomic loci with heightened specificity by enabling extensive washing prior to the biotinylation reaction, theoretically improving the signal-to-noise ratio when compared with other methods such as dCas9-based techniques. Should the method prove successful, it could represent a notable advancement in the field of chromatin biology, particularly in establishing the proteomes of individual chromatin regions - an extremely challenging objective that has not yet been comprehensively addressed by existing methodologies.

Strength of the Evidence:

The evidence presented by the authors is somewhat mixed, and the robustness of the findings appears to be preliminary at this stage. While certain data indicate that DNA-O-MAP may function effectively for repetitive DNA regions, a number of the claims made in the manuscript are either unsupported or require further substantiation. There are significant concerns about the resolution of the method, with substantial biotinylation signals extending well beyond the intended target regions (megabases around the target), suggesting a lack of specificity and poor resolution, particularly for smaller loci. Furthermore, comparisons with previous techniques are unfounded since the authors have not provided direct comparisons with the same mass spectrometry (MS) equipment and protocols. Additionally, although the authors assert an advantage in multiplexing, this claim appears overstated, as previous methods could achieve similar outcomes through TMT multiplexing. Therefore, while the method has potential, the evidence requires more rigorous support, comprehensive benchmarking, and further experimental validation to demonstrate the claimed improvements in specificity and practical applicability.

We thank the Reviewer for providing detailed critiques of our manuscript. As noted above, we plan to substantially revise the framing and presentation of manuscript to address the concerns raised by all three reviewers.

Author response:

Reviewer #1 (Public review):

Summary:

The crystal structure of the Sld3CBD-Cdc45 complex presented by Li et al. is a novel contribution that significantly advances our understanding of CMG formation during the rate-limiting step of DNA replication initiation. This structure provides insights into the intermediate steps of CMG formation. The study builds upon previously known structures of Sld3 and Cdc45 and offers new perspectives into how Cdc45 is loaded onto MCM DH through Sld3-Sld7. The most notable finding is the structural difference in Sld3CBD when bound to Cdc45, particularly the arrangement of the α8-helix, which is essential for Cdc45 binding and may also pertain to its metazoan counterpart, Treslin. Additionally, the conformational shift in the DHHA1 domain of Cdc45 suggests a possible mechanism for its binding to MCM2NTD.

Strengths:

The manuscript is generally well-written, with a precise structural analysis and a solid methodological section that will significantly advance future studies in the field. The predictions based on structural alignments are intriguing and provide a new direction for exploring CMG formation, potentially shaping the future of DNA replication research.

Weaknesses:

The main weakness of the manuscript lies in the lack of experimental validation for the proposed Sld3-Sld7-Cdc45 model. Specifically, the claim that Sld3 binding to Cdc45-MCM does not inhibit GINS binding, a finding that contradicts previous research, is not sufficiently substantiated with experimental evidence. To strengthen their model, the authors must provide additional experimental data to support this mechanism. Also, the authors have not compared the recently published Cryo-EM structures of the metazoan CMG helicases with their predicted models to see if Sld3/Treslin does not cause any clash with the GINS when bound to the CMG. Still, the work holds great potential in its current form but requires further experiments to confirm the authors' conclusions.

We appreciate the reviewers’ careful reading and the comments.

The structure of Sld3CBD-Cdc45 showed that the binding site of Cdc45 to Sld3CBD was distinct from the binding ranges of Cdc45 to GINS and MCM, indicating that the Sld3CBD, MCM, and GINS bind to separate sites of Cdc45 on the CMG complex. The SCMG-DNA model confirmed such a binding situation but did not show whether the binding of Sld3 to Cdc45 affects the recruitment of GINS (by GINS-Dbp11-Sld2) for CMG formation. We will modify our manuscript and discuss this point. Also, we will check the recently published Cryo-EM structures of the metazoan CMG helicases with their predicted models to confirm our conclusions. We will try to conduct the experiments as suggested.

Reviewer #2 (Public review):

Summary

The manuscript presents valuable findings, particularly in the crystal structure of the Sld3CBD-Cdc45 interaction and the identification of additional sequences involved in their binding. The modeling of the Sld7-Sld3CBD-CDC45 subcomplex is novel, and the results provide insights into potential conformational changes that occur upon interaction. However, the work remains incomplete as several main claims are only partially supported by experimental data, particularly the proposed model for Sld3 interaction with GINS on the CMG. Additionally, the single-stranded DNA binding data from different species do not convincingly advance the manuscript's central arguments.

Strengths

(1) The Sld3CBD-Cdc45 structure is a novel contribution, revealing critical residues involved in the interaction.

(2) The model structures generated from the crystal data are well presented and provide valuable insights into the interaction sequences between Sld3 and Cdc45.

(3) The experiments testing the requirements for interaction sequences are thorough and conducted well, with clear figures supporting the conclusions.

(4) The conformational changes observed in Sld3 and Cdc45 upon binding are interesting and enhance our understanding of the interaction.

(5) The modeling of the Sld7-Sld3CBD-CDC45 subcomplex is a new and valuable addition to the field.

Weaknesses

(1) The proposed model for Sld3 interacting with GINS on the CMG needs more experimental validation and conflicts with published findings. These discrepancies need more detailed discussion and exploration.

(2) The section on the binding of Sld3 complexes to origin single-stranded DNA needs significant improvement. The comparisons between Sld3-CBD, Sld3CBD-Cdc45, and Sld7-Sld3CBD-Cdc45 involve complexes from different species, limiting the comparisons' value.

(3) The authors' model proposing the release of Sld3 from CMG based on its binding to single-stranded DNA is unclear and needs more elaboration.

We appreciate your positive comments. As suggested, we will try to improve the experiments and manuscript and discuss in more detail, including the interaction between Sld3 and GINS on the CMG, ssDNA-binding section, and the explanations of why we use different species for comparison and more elaboration on the Sld3-release proposal.

Reviewer #3 (Public review):

Summary:

The paper by Li et al. describes the crystal structure of a complex of Sld3-Cdc45-binding domain (CBD) with Cdc45 and a model of the dimer of an Sld3-binding protein, Sld7, with two Sld3-CBD-Cdc45 for the tethering. In addition, the authors showed the genetic analysis of the amino acid substitution of residues of Sld3 in the interface with Cdc45 and biochemical analysis of the protein interaction between Sld3 and Cdc45 as well as DNA binding activity of Sld3 to the single-strand DNAs of the ARS sequence.

Strengths:

The authors provided a nice model of an intermediate step in the assembly of an active Cdc45-MCM-GINS (CMG) double hexamers at the replication origin, which is mediated by the Sld3-Sld7 complex. The dimer of the Sld3-Sld7 complexes tethers two MCM hexamers together for the recruitment of GINS-Pol epsilon on the replication origin.

Weaknesses:

The biochemical analysis should be carefully evaluated with more quantitative ways to strengthen the authors' conclusion.

We thank your positive assessment. We will provide more quantitative information and try to quantify the experiments as suggested.

Author response:

Reviewer 1:

(1) I think the article is a little too immature in its current form. I'd recommend that the authors work on their writing. For example, the objectives of the article are not completely clear to me after reading the manuscript, composed of parts where the authors seem to focus on SGCs, and others where they study "engram" neurons without differentiating the neuronal type (Figure 5). The next version of the manuscript should clearly establish the objectives and sub-aims.

Our overarching focus was to identify whether intrinsic physiology and circuit connectivity of SGCs contribute to their unique overrepresentation in neurons labeled as part of a behaviorally relevant dentate engram. Since our systematic analysis of “engram SGCs” did not support the proposal that engram SGCs drive robust feedforward excitation of engram GCs or feedback inhibition of non-engram GCs, we examined an alternative hypothesis that inputs drive recruitment of neurons, regardless of subtype (in figure 5). These are sparsely labeled neurons, with mixed populations of GCs and SGCs undergoing paired recordings. Since the focus of the experiment was input correlation between two simultaneously recorded neurons, we did not report the individual cell types. We regret that this caused confusion and will clarify this issue in the revised manuscript.

(2) In addition, some results are not entirely novel (e.g., the disproportionate recruitment as well as the distinctive physiological properties of SGCs), and/or based on correlations that do not fully support the conclusions of the article. In addition to re-writing, I believe that the article would benefit from being enriched with further analyses or even additional experiments before being resubmitted in a more definitive form.

We would like to note that while we and others have previously reported the distinctive SGC physiology, this study is the first to compare physiological properties of SGCs labeled as part of an engram to unlabeled SGCs. That was the thrust of the data presented which may have been missed and will be emphasized in the revision. Similarly, while others have shown higher SGC recruitment in dentate engrams, we had to validate this in the dentate dependent behaviors that we adopted in this study. We also note that the proportional SGC recruitment in our study, based on morphometric classification, differs from what was reported previously. These aspects of study, which were considered confirmatory, represent the necessary validation needed to proceed with the novel cell-type specific paired recordings and optogenetic analyses of engram neurons presented in subsequent sections of the manuscript. We will emphasize these considerations in the revised manuscript.

Reviewer 2:

(1) The authors conclude that SGCs are disproportionately recruited into cfos assemblies during the enriched environment and Barnes maze task given that their classifier identifies about 30% of labelled cells as SGCs in both cases and that another study using a different method (Save et al., 2019) identified less than 5% of an unbiased sample of granule cells as SGCs. To make matters worse, the classifier deployed here was itself established on a biased sample of GCs patched in the molecular layer and granule cell layer, respectively, at even numbers (Gupta et al., 2020). The first thing the authors would need to show to make the claim that SGCs are disproportionately recruited into memory ensembles is that the fraction of GCs identified as SGCs with their own classifier is significantly lower than 30% using their own method on a random sample of GCs (e.g. through sparse viral labelling). As the authors correctly state in their discussion, morphological samples from patch-clamp studies are problematic for this purpose because of inherent technical issues (i.e. easier access to scattered GCs in the molecular layer).

We regret that there seems to be some confusion about use of a classifier. We did NOT use any automated classifier in this study. All cell type classifications in the study were conducted by experienced investigators examining cell morphology and classifying cells based on established morphometric criteria. In our prior study (Gupta et al., 2020) we had conducted an automated cluster analysis that was able to classify GCs and SGCs as different cell types. The principal components underlying the automated clustering in Gupta et al 2020 were consistent with the major criteria identified in prior morphology-based analyses by us and others (including Williams et al 2010 and Save et al., 2019). To date, in the absence of a validated molecular marker, morphometry from recorded and filled cells or sparsely labeled neurons is the only established method to classify SGCs. This was the approach we adopted, and this will be further clarified in the revisions.

(2) The authors claim that recurrent excitation from SGCs onto GCs or other SGCs is irrelevant because they did not find any connections in 32 simultaneous recordings (plus 63 in the next experiment). Without a demonstration that other connections from SGCs (e.g. onto mossy cells or interneurons) are preserved in their preparation and if so at what rates, it is unclear whether this experiment is indicative of the underlying biology or the quality of the preparation. The argument that spontaneous EPSCs are observed is not very convincing as these could equally well arise from severed axons (in fact we would expect that the vast majority of inputs are not from local excitatory cells). The argument on line 418 that SGCs have compact axons isn't particularly convincing either given that the morphologies from which they were derived were also obtained in slice preparations and would be subject to the same likelihood of severing the axon. Finally, even in paired slice recordings from CA3 pyramidal cells the experimentally detected connectivity rates are only around 1% (Guzman et al., 2016). The authors would need to record from a lot more than 32 pairs (and show convincing positive controls regarding other connections) to make the claim that connectivity is too low to be relevant.

As noted in our discussion, we are fully cognizant that potential SGC to GC connections may have been missed by the nature of slice physiology experiments and made every effort to limit this possibility. As noted in the manuscript, we only analyzed GC/SGC pairs where hilar axon collaterals of the neurons were recovered. We do not claim that SGC to GC/SGC connections are irrelevant, rather, we indicate that these connections, if present, are sparse and unlikely to drive engram refinement. Interestingly, wide field optical stimulation, designed to activate multiple labeled engram neurons and axon terminals including those of SGCs whose somata were outside the slice, did not lead to EPSCs in other unlabeled GCs or SGCs suggesting the lack of robust SGC to GC/SGC synaptic connectivity. While we have previously published paired recordings from interneurons to GCs (Proddutur  et al 2023) , we agree that recordings demonstrating the presence of SGC/GC to hilar neuron synapses would serve as an added control in the revised manuscript.

(3) Another troubling sign is the fact that optogenetic GC stimulation rarely ever evokes feedback inhibition onto other cells which contrasts with both other in vitro (e.g. Braganza et al., 2020) and in vivo studies (Stefanelli et al., 2016) studies. Without a convincing demonstration that monosynaptic connections between SGCs/GCs and interneurons in both directions is preserved at least at the rates previously described in other slice studies (e.g. Geiger et al., 1997, Neuron, Hainmueller et al., 2014, PNAS, Savanthrapadian et al., 2014, J. Neurosci), the notion that this setting could be closer to naturalistic memory processing than the in vivo experiments in Stefanelli et al. (e.g. lines 443-444) strikes me as odd. In any case, the discussion should clearly state that compromised connectivity in the slice preparation is likely a significant confound when comparing these results.

We would like to note that our data are consistent with Braganza 2020 study, as we explain below. Moreover, we would like to point out that the demonstration of “feedback inhibition” in the Stefanelli study was NOT in engram or behaviorally labeled neurons nor was it in vivo. As we explain below, the physiological assay in Stefanelli was in slices and in a cohort of GCs with virally driven ChR2 expression. Thus, we are fully confident that our experimental paradigm better reflects a behavioral engram. As noted in response (2, we have previously published paired monosynaptic connections from interneurons to GCs (Proddutur  et al 2023) and find the connectivity consistent with published data. However, we agree that recordings demonstrating the presence of SGC/GC to hilar neuron synapses  or recruitment of feedback inhibition by focal activation of GCs would serve to allay concerns regarding slice preparation. We also submit that we already discuss the potential concerns regarding compromised connectivity in slice preparations.

Regarding the lack of optically evoked feedback inhibition, we would like to point out that the Braganza 2020 study examined focal optogenetic activation of GCs, where a high density of GCs was labeled using a Prox-cre line. They reported that about 2-4% of these densely labeled cells need to be recruited to evoke feedback IPSCs. Our experimental condition, where ChR2 was expressed in behaviorally labeled neurons, leads to sparse labeling much less than the focal 4% needed to evoke IPSCs in the Braganza study. We do not claim that feedback inhibition cannot be activated by focal activation of a cohort of GCs and even show an example of paired recording with feedback GC inhibition of an SGC. Our conclusion is that the few sparsely labeled neurons during a behavioral episode do not support robust feedback inhibition proposed to mediate engram refinement. We submit that our findings are fully consistent with the sparse GC driven feedback inhibition, and the need to activate a cohort of focal GCs to recruit feedback inhibition, reported in Braganza 2020

Regarding the Stefanelli study, we maintain that our behaviorally relevant in vivo labeling approach is more naturalistic than the DREADD and Channelrhodopsin driven artificial “engrams” generated in the Stefanelli study. Of note, we used cFOS driven TRAP mice to label, in vivo, neurons active during a behavior and then undertook slice physiology studies in these mice a week later. In contrast, the slice physiology data demonstrating putative feedback inhibition in the Stefanelli study (Fig 5) used wildtype mice injected with AAV CAMKII-cre and AAV-DIO-ChR2. Thus, unlike our study, the physiological data demonstrating feedback inhibition in the Stefanelli study was not performed in a behaviorally labeled engram. Apart from the one set of histological experiments using AAV-SARE-GFP to demonstrate increased GFP labeling of SST neurons in behavior, all other data presented in the Stefanelli study are generated based on artificially generated engrams where optogenetic activation or silencing on granule cells was used to manipulate the numbers of neurons active during a task followed by histological analysis of cFOS staining or behaviors. Thus, the physiological experiments in the Stefanelli et al (2016) generated by wide field activation of a large cohort of GCs labeled by focal virally driven ChR2 expression, were similar to wide field optical stimulation studies in the Braganza 2020 study, and were NOT conducted in a behavioral engram. The strength of our study is in the use of a behaviorally tagged engram neurons for analysis and our findings in sparsely labeled neurons are consistent with the reports in Braganza 2020. We will further clarify in our discussion that the data presented in the Stefanelli study do NOT represent a natural behavior generated engram.

(4) Probably the most convincing finding in this study is the higher zero-time lag correlation of spontaneous EPSCs in labelled vs. unlabeled pairs. Unfortunately, the fact that the authors use spontaneous EPSCs to begin with, which likely represent a mixture of spontaneous release from severed axons, minis, and coordinated discharge from intact axon segments or entire neurons, makes it very hard to determine the meaning and relevance of this finding. At the bare minimum, the authors need to show if and how strongly differences in baseline spontaneous EPSC rates between different cells and slices are contributing to this phenomenon. I would encourage the authors to use low-intensity extracellular stimulation at multiple foci to determine whether labelled pairs really share higher numbers of input from common presynaptic axons or cells compared to unlabeled pairs as they claim. I would also suggest the authors use conventional Cross correlograms (CCG; see e.g. English et al., 2017, Neuron; Senzai and Buzsaki, 2017, Neuron) instead of their somewhat convoluted interval-selective correlation analysis to illustrate co-dependencies between the event time series. The references above also illustrate a more robust approach to determining whether peaks in the CCGs exceed chance levels.

We appreciate the comment can provide additional data on the EPSC frequency in individual labeled and unlabeled cells in the revised manuscript. As indicated in the manuscript, we constrained our analysis to cell pairs with comparable EPSC frequency in order to avoid additional confounds in analysis. We have additional experiments to show that over 50% of the sEPSCs represent action potential driven events which we will include in the revised manuscript. We thank the reviewer for the suggestion to explores alternative methods of analyses including CCGs to further strengthen our findings.

(5) Finally, one of the biggest caveats of the study is that the ensemble is labelled a full week before the slice experiment and thereby represents a latent state of a memory rather than encoding consolidation, or recall processes. The authors acknowledge that in the discussion but they should also be mindful of this when discussing other (especially in vivo) studies and comparing their results to these. For instance, Pignatelli et al 2018 show drastic changes in GC engram activity and features driven by behavioral memory recall, so the results of the current study may be very different if slices were cut immediately after memory acquisition (if that was possible with a different labelling strategy), or if animals were re-exposed to the enriched environment right before sacrificing the animal.

As noted by the reviewer, we fully acknowledge and are cognizant of the concern that slices prepared a week after labeling may not reflect ongoing encoding. Although our data show that labeled cells are reactivated in higher proportion during recall, we have discussed this caveat and will include alternative experimental strategies in the discussion.

Reviewer 3:

(1) Engram cells are (i) activated by a learning experience, (ii) physically or chemically modified by the learning experience, and (iii) reactivated by subsequent presentation of the stimuli present at the learning experience (or some portion thereof), resulting in memory retrieval. The authors show that exposure to Barnes Maze and the enriched environment-activated semilunar granule cells and granule cells preferentially in the superior blade of the dentate gyrus, and a significant fraction were reactivated on re-exposure. However, physical or chemical modification by experience was not tested. Experience modifies engram cells, and a common modification is the Hebbian, i.e., potentiation of excitatory synapses. The authors recorded EPSCs from labeled and unlabeled GCs and SGCs. Was there a difference in the amplitude or frequency of EPSCs recorded from labeled and unlabeled cells?

We agree that we did not examine the physical or chemical modifications by experience. Although we constrained our sEPSC analysis to cell pairs with comparable sEPSC frequency, we will include data on sEPSC parameters in labeled and unlabeled cells in the revised manuscript.

(2) The authors studied five sequential sections, each 250 μm apart across the septotemporal axis, which were immunostained for c-Fos and analyzed for quantification. Is this an adequate sample? Also, it would help to report the dorso-ventral gradient since more engram cells are in the dorsal hippocampus. Slices shown in the figures appear to be from the dorsal hippocampus.

We thank the reviewer for the comment. We analyzed sections along the dorso-ventral gradient. As explained in the methods, there is considerable animal to animal variability in the number of labeled cells which was why we had to use matched littermate pairs in our experiments This variability could render it difficult to tease apart dorsoventral differences.

(3) The authors investigated the role of surround inhibition in establishing memory engram SGCs and GCs. Surprisingly, they found no evidence of lateral inhibition in the slice preparation. Interneurons, e.g., PV interneurons, have large axonal arbors that may be cut during slicing. Similarly, the authors point out that some excitatory connections may be lost in slices. This is a limitation of slice electrophysiology.

We agree that slice physiology has limitations and discuss this caveat. As noted in response (2, we have previously published paired monosynaptic connections from interneurons to GCs (Proddutur  et al 2023) and find the connectivity consistent with published data. However, we agree that recordings demonstrating the presence of SGC/GC to hilar neuron synapses  or recruitment of feedback inhibition by focal activation of GCs would serve to allay concerns regarding slice preparation.

Author response:

The following is the authors’ response to the original reviews.

Reviewer #1 (Public Review):

The study by Chikermane and colleagues investigates the functional, structural, and dopaminergic network substrates of cortical beta oscillations (13-30 Hz). The major strength of the work lies in the methodology taken by the authors, namely a multimodal lesion network mapping. First, using invasive electrophysiological recordings from healthy cortical territories of epileptic patients they identify regions with the highest beta power. Next, they leverage open-access MRI data and PET atlases and use the identified high-beta regions as seeds to find (1) the whole-brain functional and structural maps of regions that form the putative underlying network of high-beta regions and (2) the spatial distribution of dopaminergic receptors that show correlation with nodal connectivity of the identified networks. These steps are achieved by generating aggregate functional, structural, and dopaminergic network maps using lead-DBS toolbox, and by contrasting the results with those obtained from high-alpha regions.

The main findings are:

(1) Beta power is strongest across frontal, cingulate, and insular regions in invasive electrophysiological data, and these regions map onto a shared functional and structural network. (2) The shared functional and structural networks show significant positive correlations with dopamine receptors across the cortex and basal ganglia (which is not the case for alpha, where correlations are found with GABA).

Nevertheless, a few clarifications regarding the choice of high-power electrodes and distributions of functional connectivity maps (i.e., strength and sign across cortex and sub-cortex) can help with understanding the results.

We thank the reviewer for this critical expert assessment. 

Reviewer #1 (Recommendations For The Authors):

To potentially enhance the quality of the manuscript in the current version, I kindly ask the authors to address the following points:

Major:

(A) Power analysis of electrophysiological data

(1) How were significant peaks identified exactly? I understand that the authors used FOOOF methodology to estimate periodic components of brain activity.

Thank you for pointing us to this lack of clarity. The application of FOOOF consists of the fitting of a one-over-f curve that delineates the aperiodic component followed by the definition of gaussians to fit periodic activity. This allows for extraction of periodic peak power estimates that are corrected for offset and exponent of the one-over-f or non-oscillatory aperiodic component in the spectrum (further information can be found here https://fooof-tools.github.io/fooof/auto_tutorials/plot_02-FOOOF.html). We included all peaks that could be fitted using the process.

How about aperiodic components (Figure 1, PSD plots)? 

We share the interest in aperiodic activity with the reviewer. However, given that the primary aim of this study was the description of beta oscillations and the methodology and results presentation is already very complex, we did not include the analysis of aperiodic activity in this manuscript. This could be done in the future and it would surely be interesting to visualize the whole brain connectomic fingerprints of aperiodic exponent and offset. With regard to the purely anatomical description of nonoscillatory aperiodic activity we would like to refer to Figure 8 in Frauscher et al. Brain 2018 (https://doi.org/10.1093/brain/awy035) where this is described. We have decided not to include additional information on this matter, because a) we felt that this would further convolute the results and discussion without directly addressing any of the hypotheses and aims that we set out to tackle and b) the interpretation of aperiodic activity is still a matter of intense research with conflicting results, which warrants very careful considerations of many aspects that again would go beyond the scope of this paper. 

In addition, to what degree would the results change if one identified the peaks relative to sites with no peak, similar to Frauscher et al. 

Beta activity, the oscillation of interest in our analysis is ubiquitous in the brain. In fact, of 1772 channels, only 21 channels did not exhibit a beta peak detectable with FOOOF. Thus, a comparison of 1751 against 21 would not yield meaningful results. We have therefore decided to focus on the channels in which beta activity is the strongest and dominant observable oscillation. 

If the FOOOF approach has some advantages, these should be pointed out or discussed.

FOOOF indeed has the advantage that it provides an objective and reproducible estimation of peak oscillatory activity that accounts for differences in aperiodic activity. To the best of our knowledge, there is no other approach that is nearly as well documented, validated and computationally reproducible. 

Changes in manuscript: We have now further clarified the definition of peak amplitudes in the results and methods section and have discussed the use of alternative measures in the limitations section of our manuscript.

Results: “The frequency band with the highest peak amplitude was identified using the extracted peak parameter (pw) for each channel and depicted as the dominant rhythm for the respective localisation (Figure 1).”

Methods: “Peak height was extracted using the pw parameter, which depicts peak amplitude after subtraction of any aperiodic activity.”

Discussion: “Alternative approaches could yield different results, e.g. reusing channels for each peak that is observable and contrasting them to channels where such peak was not present. However, in our study the majority of channels exhibited beta activity, even if peaks were of low amplitude, which we believe would have led to less interpretable results.”

(2) How exactly do the authors deal with channels with more than one peak? Some elaboration on this and how this could potentially impact the results would be appreciated. Sorry if I have missed it.

Indeed, a description of this was lacking so we are very thankful that the reviewer pointed this out. The maximum peak amplitude method was a winner-takes-all approach where in the case of multiple peaks, the peak with the higher amplitude was chosen. This method of course has drawbacks in the form of lost or disregarded peaks and remains a limitation to this study. 

Changes in manuscript: We have now clarified this in the methods and results sections, which now read: 

Methods: “In case of multiple peaks within the same region, we used only the highest peak amplitude.”

Results: “In case of multiple peaks within the same frequency band, we focused the analysis on the peak with the highest amplitude.”

And added the following to the Limitations section of the discussion: 

“Another limitation in our study is the fact that the statistical approach for the comparison of beta and alpha networks and even for multiple peaks within the same frequency band follows a winner takes all logic that is, by definition, a simplification, as most areas will contribute to more than one spatiospectrally distinct oscillatory network. Specifically, while multiple peaks within or across frequency bands could be present in each channel, we decided to allocate this channel to only the frequency band containing the highest peak amplitude.” 

(B) Network mapping

(1) Knowing that fMRI data are preprocessed by regressing the global signal, there are negative correlations across the functional networks. Unfortunately, the distribution, sign, and strength of the correlations are not quantitatively shown in any of the plots. Thus, it is unclear whether, e.g., corticocortical vs. subcortico-cortical correlations differ in strength and/or sign. I think this additional information is important for better understanding the up/down-regulation of beta, e.g., by DA signaling. Some discussion around this point in addition would be insightful, I think.

The referee is touching upon a very important and difficult point, which we have considered very carefully. Global signal regression is a controversial topic and the neurophysiological basis of negative correlations remains to be elucidated. We can justify our use of this approach based on an expert consensus described in Murphy & Fox 2017 (https://doi.org/10.1016%2Fj.neuroimage.2016.11.052), which highlights that global signal regression can improve the specificity of positive correlations, improve the correspondence to anatomical connectivity. The truth however is that, we relied on it, because it is the more commonly used and validated approach used in lesion network and DBS connectivity mapping and implemented in the Lead Mapper pipeline. Indeed all connectivity estimates are shown in Supplementary figure 3. We remain hesitant to raise the focus to these points, because of the uncertain underlying neural correlates. However, when looking at the values, it is interesting to note that most key regions of interest exhibit positive connectivity values. 

Changes in manuscript: We now point to the supplement containing all connectivity values in the results section more prominently: “All connectivity values including their sign are shown in figures as brain region averages parcellated with the automatic anatomical labelling atlas in supplementary figures 2&3.”

(2) I assume no thresholding is applied to the functional connectivity maps (in a graph-theoretical sense). Please clarify (this is also related to the comment above, in particular, the strength of correlations.

Indeed, we demonstrate SPM maps using family wise error corrected stats in figure 2, but all further analyses were performed on unthresholded maps as correctly pointed out by the referee. 

Changes in manuscript: 

Results: “Specifically, we analysed to what degree the spatial uptake patterns of dopamine, as measurable with fluorodopa (FDOPA; cohort average of 12 healthy subjects) and other dopamine signalling related tracers that bind D1/D2 receptors (average of N=17/44 respectively healthy subjects) or the dopamine transporter (DAT; cohort average of N=180 healthy subjects) were correlated with the unthresholded MRI connectivity maps.”

Methods: “This parcellation was applied to both PET and unthresholded structural and functional connectivity maps using SPM and custom code.”

Minor

(1) Methods, Connectivity analysis: The description of (mass-univariate) GLM analysis is confusing. The maps underwent preprocessing? Which preprocessing steps are meant here? What is the dependent variable and what are the predictors exactly?

We thank the reviewer for catching this error in our methods. We apologise for the confusion and mistake and thank the reviewer for catching it. Indeed, we have used t-tests without further preprocessing instead of a GLM. 

Changes in manuscript: The respective section has been removed from the methods section and intermediate steps have been clarified. The section now reads: “To investigate differences between beta dominant and alpha dominant functional connectivity networks, a two sample t-test was calculated for the condition where beta was greater than alpha and vice versa using SPM. Here, the connectivity maps from each dominant channel (1005 beta functional connectivity maps and 397 alpha connectivity maps) Estimation of model parameters yielded t-values for each voxel, indicating the strength and direction of differences between the two contrasts (beta > alpha, alpha > beta). To address the issue of multiple comparisons, we applied Family-Wise Error (FWE) correction, adjusting significance thresholds such that only voxels with p < 0.05 would be included.”

(2) I encourage the authors to find a better (visual) way of reporting Table 1, to make the main observations easier to grasp and compare (maybe a two-dimensional bar plot? Or color-coding the cells?)

Reply: Thank you for your suggestion to improve the table, the new table is adjusted to the recommended changes to make it more readable.

Reviewer #2 (Public Review):

Summary:

This is a very interesting paper that leveraged several publicly available datasets: invasive cortical recording in epilepsy patients, functional and structural connectomic data, and PET data related to dopaminergic and gaba-ergic synapses. These were combined to create a unified hypothesis of beta band oscillatory activity in the human brain. They show that beta frequency activity is ubiquitous, not just in sensorimotor areas, and cortical regions where beta predominated had high connectivity to regions high in dopamine re-uptake.

Strengths:

The authors leverage and integrate three publicly available human brain datasets in a creative way. While these public datasets are powerful tools for human neuroscience, it is innovative to combine these three types of data into a common brain space to generate novel findings and hypotheses. Findings are nicely controlled by separately examining cortical regions where alpha predominates (which have a different connectivity pattern). GABA uptake from PET studies is used as a control for the specificity of the relationship between beta activity and dopamine uptake. There is much interest in synchronized oscillatory activity as a mechanism of brain function and dysfunction, but the field is short on unifying hypotheses of why particular rhythms predominate in particular regions. This paper contributes nicely to that gap. It is ambitious in generating hypotheses, particularly that modulation of beta activity may be used as a "proxy" for modulating phasic dopamine release.

Weaknesses:

As the authors point out, the use of normative data is excellent for exploring hypotheses but does not address or explore individual variations which could lead to other insights. It is also biased to resting state activity; maps of task-related activity (if they were available) might show different findings.

The figures, results, introduction, and methods are admirably clear and succinct but the discussion could be both shorter and more convincing.

Reviewer #2 (Recommendations For The Authors):

The tone of the discussion is excessively lofty and abstract, and hard to follow in places. Specific examples in comments to authors below.

We thank the reviewer for their positive assessment and their constructive feedback on the discussion. Also in light of the other reviewers we have made a sincere effort to shorten, restructure and improve the discussion. Additionally, we have addressed all the specific comments the reviewer had below. We appended each change to the manuscript where appropriate below and have addressed all comments in the main text. Having that said, we see this paper and discussion to provide our most up-to-date and personal perspective on a correct concept on the interplay of beta oscillations and dopamine that is generalizable. Providing a concept that is so generalizable is very challenging and so far very few authors have even attempted this. One notable exception is the “status quo” concept by Fries & Engel. While we will do our very best to address the comments, we have decided not to deviate from our initial ambition to provide a discussion on a generalizable concept. Naturally such a concept must be very complex and therefore it will be hard to understand in parts. Through the revision, we hope that the readability and comprehensibility has improved, while it provides an in-depth perspective and hypothesis on how beta oscillations, dopamine and their brain circuits may facilitate brain function. Nevertheless, we want to express our honest gratitude for the thoroughness with which the reviewer has read and scrutinized our paper. The review clearly tells that the reviewer had the ambition to follow and understand what we were trying to convey, which can be rare nowadays. We are truly thankful for this.

The first sentence is not quite true, as invasive neurophysiology was not, and cannot be, done in healthy humans. "The present study combined three openly available datasets of invasive neurophysiology, MRI connectomics, and molecular neuroimaging in healthy humans to characterise the spatial distribution of brain regions exhibiting resting beta activity, their shared circuit architecture, and its correlation with molecular markers of dopamine signaling in the human brain."

Changes in manuscript: We have now removed the “healthy” from the respective sentence.

"Our results motivate to conceptualise the capacity to generate.... This is not clear.

Changes in manuscript: “Our results suggest that one common denominator of brain regions that generate beta activity, is their affiliation with beta oscillations as a feature that arises from a largescale global brain network that is modulated by dopamine.”

"Similarly, the robust beta modulation that is elicited by voluntary action in sensorimotor cortex and its correlation with motor symptoms of Parkinson's disease is long known" - the association between movement-related cortical beta desynchronization and Parkinson's motor signs is not well described - could the authors specify and reference this?

We thank the reviewer for pointing out this lack of clarity. We meant that independently beta is known for “movement” and for “movement disorders” and not “movement in movement disorders”. Having that said, there are some studies that suggest that beta ERD is altered in PD (e.g.https://doi.org/10.1093/cercor/bht121), but saying that this is “long known” would be an overstatement and was not our intention. We rephrased this sentence accordingly.

Changes in manuscript: The sentence now reads: “Moreover, the robust beta modulation that is elicited by voluntary action in sensorimotor cortex and its correlation with motor symptoms of Parkinson’s disease is long known.”

"...first fast-cyclic voltammetry experiments that allowed for combined measurement of dopamine release with invasive neurophysiology have provided first evidence that beta band oscillations in healthy non-human primates can differentially link dopamine release, beta oscillations and reward and motor control, depending on the contextual information and striatal domain" - This is not very clear - not sure what "differentially link" signifies.

I think the fact that this is not easy to understand signifies the complexity that we and the authors of the cited paper from Ann Graybiel’s lab aimed to communicate. In fact, we stayed very close to the phrasing used in their paper to try and avoid confusion (Title: Dopamine and beta-band oscillations differentially link to striatal value and motor control” - https://doi.org/10.1126/sciadv.abb9226). The specific results go beyond the scope of the discussion but are very interesting, so I would be happy if our paper would inspire readers to look it up. 

Changes in manuscript: We have now adapted the sentence to “In line with this more complex picture, direct measurement of dopamine concentration in non-human primates revealed specific interactions between dopamine release, beta oscillations, reward value and motor control, depending on contextual information and striatal domain. This shows that the relationship of dopamine and beta activity is not solely associated with either reward or movement and depends on where in the striatum beta activity is recorded.”

"In fact, one could argue that it can be contextualised in a recently described framework of neural reinforcement, that serves to orchestrate the re-entrance and refinement of neural population dynamics for the production of neural trajectories" - this is not clear - for example what is a neural trajectory? What is meant by "re-entrance and refinement"?

A neural trajectory refers to the path that the activity of a neural population takes through a high-dimensional space over time. It can be obtained through multivariate analysis of population activity with dimensionality reduction techniques, such as PCA. The concept of low-dimensional representations of high-dimensional neural activity has gained a lot of attention in computational neuroscience ever since high-channel count recordings of neural population activity have become available (an early and prominent example is Churchland et al., 2012 Nature https://doi.org/10.1038/nature11129 , while a more recent example is Safaie et al., Nature 2023 https://doi.org/10.1038/s41586-023-06714-0). The review we refer to by Rui Costa and colleagues (Athalye, V. R., Carmena, J. M. & Costa, R. M. Neural reinforcement: re-entering and refining neural dynamics leading to desirable outcomes. Curr Opin Neurobiol 60, 145–154 (2020) https://doi.org/10.1016/j.conb.2019.11.023) suggests that dopamine may serve to modulate the likelihood of a specific pattern to emerge and re-enter the cortex – basal ganglia loop, for the “reliable production of neural trajectories driving skillful behavior on-demand”. We believe that this concept could be revolutionary in our understanding of dopaminergic modulation and disoroders and together with colleague Alessia Cavallo have written an invited perspective on this topic (https://doi.org/10.1111/ejn.16222), which may help further clarify the topic. 

Changes in manuscript: We realize that this aspect may sound a bit unclear or far away from the data in this manuscript. However, given that we have spent more than a decade thinking about beta oscillations and how they can be conceptualized, we would prefer not to entirely change our points and rather bet on the possibility that the concepts become more widely accepted and well-known. Nevertheless, we have now adapted the text to make this a bit more clear:

“We hypothesise that, this “status quo” hypothesis could be equally or maybe even more adequately posed on the neural level. Namely, it could provide insights to what degree a certain activity pattern or synaptic connection is to be strengthened or weakened, in light of neural learning. We propose that this putative function can be contextualised in a recently described framework of neural reinforcement, that serves to orchestrate the re-entrance and refinement of neural population dynamics for the production of neural trajectories.”

"....after which it was quickly translated to first experimental studies using cortical or subcortical beta signals in human patients44." - reference 44 only deals with the use of subcortical beta, not cortical, in adaptive control.

The reviewer is right, in fact there is no study using motor cortex beta for adaptive DBS yet, but different studies have used different markers (especially gamma) since then. 

Changes in manuscript: We have rephrased and added citations accordingly: “This approach, also termed adaptive DBS, was first demonstrated based on cortical beta activity that was used to adapt pallidal DBS in the MPTP non-human primate model of PD43. It was quickly translated to first experimental studies using subcortical beta signals in human patients44, followed by further research using more complex cortical and subcortical sensing setups and biomarker combinations45,46.”

The paragraph headed " Implications for neurotechnology" is quite long and should be condensed and focused. It doesn't seem to support the last sentence, "....targeted interventions that can increase and decrease beta activity, as recently shown through phase specific modulation45 could be utilised to mimic phasic dopamine release as a neuroprosthetic approach to alter neural reinforcement38." - I don't quite follow the logic. The authors have clearly shown that beta-related circuits tend to be those linked to dopamine modulation, and may subserve tasks for which reinforcement learning is an important mechanism. However the logic of how modulation of beta activity can "substitute" for modulation of dopamine isn't clear. That would seem to require that the mechanism by which dopamine produces reinforcement, is via an effect on beta oscillation properties (phase, amplitude, frequency). Is there evidence for this? If so it should be better spelled out.

We realize that this is very speculative at this point. Indeed, we believe that subthalamic DBS can mimic dopaminergic control and in the future there may be new treatment avenues, e.g. using neurochemical using neurochemical interfaces for which beta could be informative to mimic dopamine release but ultimately explaining this would be very complex, so we have removed the sentence. With regard to the remaining text in the section, we considered shortening / condensing but felt that this paragraph is highly relevant for the ongoing development of neurotechnology and therefore decided to only remove the first and last sentences.

Changes in manuscript: We have removed the first and last sentences.

"While the abovementioned prospects are promising we should cautiously consider the limitations of our study." - an unnecessary sentence to start a "limitations" section, its clearly a paragraph about limitations. In general, authors should go thru discussion and reduce verbosity; it is not nearly as well edited as the rest of the paper.

Agreed. 

Changes in manuscript: We removed the sentence. 

Reviewer #3 (Public Review):

Summary:

In this paper, Chikermane et al. leverages a large open dataset of intracranial recordings (sEEG or ECoG) to analyze resting state (eyes closed) oscillatory activity from a variety of human brain areas. The authors identify a dominant proportion of channels in which beta band activity (12-30Hz) is most prominent and subsequently seek to relate this to anatomical connectivity data by using the sEEG/ECoG electrodes as seeds in a large set of MRI data from the human connectome project. This reveals separate regions and white matter tracts for alpha (primarily occipital) and beta (prefrontal cortex and basal ganglia) oscillations. Finally, using a third available dataset of PET imaging, the authors relate the parcellated signals to dopamine signaling as estimated by spatial uptake patterns of dopamine, and reveal a significant correlation between the functional connectivity maps and the dopamine reuptake maps, suggesting a functional relationship between the two.

Strengths:

Overall, I found the paper well justified, focused on an important topic, and interesting. The authors' use of 3 different open datasets was creative and informative, and it significantly adds to our understanding of different oscillatory networks in the human brain, and their more elusive relation with neuromodulator signaling networks by adding to our knowledge of the association between beta oscillations and dopamine signaling. Even my main comments about the lack of a theta network analysis and discussion points are relatively minor, and I believe this paper is valuable and informative.

Weaknesses:

The analyses were adequate, and the authors cleverly leveraged these different datasets to build an interesting story. The main aspect I found missing (in addition to some discussion items, see below) was an examination of the theta network. Theta oscillations have been involved in a number of cognitive processes including spatial navigation and memory, and have been proposed to have different potential originating brain regions, and it would be informative to see how their anatomical networks (e.g. as in Figure 2) look like under the author's analyses.

The authors devote a significant portion of the discussion to relating their findings to a popular hypothesis for the function of beta oscillations, the maintenance of the "status quo", mostly in the context of motor control. As the authors acknowledge, given the static nature of the data and lack of behavior, this interpretation remains largely speculative and I found it a bit too far-reaching given the data shown in the paper. In contrast, I missed a more detailed discussion on the growing literature indicating a role for beta in mood (e.g. in Kirkby et al. 2018), especially given the apparent lack of hippocampal and amygdala involvement in the paper, which was surprising.

We thank the reviewer for their insightful review of our manuscript. One of the aims of our paper was to provide the ground for a circuit-based conceptualization of beta activity, which does not primarily relate to behavior. Practically we have the ambition to provide a generalizable concept that can be applied to all behavioral domains including mood. The reason we focus on the “status quo” hypothesis, is that it is one of the very few if not only generalizable concept of the function of beta oscillations. Through our paper and the discussion, we have to redirect this concept towards a less cognitive/behavioral and more anatomical network based domain, while acknowledging principles that may overlap. We realize that this is very ambitious and this endeavour is necessarily very complex and not easy to communicate. In light of the reviewers comments, we have made an effort to improve the discussion as best we could without trailing too far away from what our initial aim was. We are thankful for the suggested reference, which we have now added to the discussion in the section where we have previously discussed beta as biomarker for mood, also noting the absence of beta dominant channels in amygdala and hippocampus. Here it should be clarified however, that a) only three channels were located in the amygdala of which one exhibited beta activity, we should be cautious to not overinterpret this result and b) most channels exhibited beta and just because beta wasn’t dominant, it doesn’t mean that beta is not present or important in these brain areas. Absence of evidence is not evidence for absence with the way we approached the analysis. We are thankful for the interesting reference, which we have now included our discussion. Notably the study used a complex network analysis, which we could not perform because we did not have parallel recordings from these areas in multiple patients. This is now noted in the limitations. 

Changes in manuscript: “For example, it was shown that beta is implicated in working memory28, utilisation of salient sensory cues29, language processing30, motivation31, sleep32, emotion recognition33, mood34 and may even serve as a biomarker for depressive symptom severity in the anterior cingulate cortex35” and “One impactful study reported that beta oscillatory sub-networks of Amygdala and hippocampus could reflect human variations in mood 34. This is interesting, but highlights another relevant limitation of our study, namely that recordings in different areas were stemming from different patients and thus, such sub-network analyses on the oscillatory level could not be conducted.” 

Major comment:

• Although the proportion of electrodes with theta-dominant oscillations was lower (~15%) than alpha (~22%) or beta (~57%), it would be very valuable to also see the same analyses the authors carried out in these frequency bands extended to theta oscillations.

We agree with the reviewer and appreciate the interest in other frequency bands; theta, alpha and gamma. Our primary interest was to provide a network concept of beta activity, but anticipated that interest would go beyond that frequency band. However, we also had to limit ourselves to what is communicable and comprehensible. The key aim for us was to provide a data-driven circuit description of beta activity that can lay ground for a generalizable concept of where beta oscillations emerge. Reproducing all analyses for every frequency band would clutter both the results and the discussion. Moreover, the honest truth is that funding and individual career plans of the researchers currently do not allow to allocate time for a reanalysis of all data which would be a significant effort. Therefore, we have decided to just add the topography of theta and gamma channels as a supplement. In case the reviewer is interested on a collaboration on extending this project to other frequency bands and circuits, we would like to invite them to get in touch and perhaps this could be a new collaborative project. Until then, we have extended our limitation that this would be important work for the future. 

Changes in manuscript: 

We have added and cited the new supplementary figure for the results from theta in the results section, which now reads: 

“Further information on the topography of theta channels are shown in supplementary figure 1.”

We would like to add that a sensible interpretation of results from gamma dominant channels is unlikely to be possible given the low count of channels with prominent resting activity in this frequency band. We have added the following text to the limitations section: “The aim of this study was to elucidate the circuit architecture of beta oscillations, which is why insights from this study for other frequency bands are limited. Future research investigating the specific circuits of theta, alpha and gamma oscillations and their relationship with neurotransmitter uptake could yield new important insights on the networks underlying human brain rhythms.“ 

Reviewer #3 (Recommendations For The Authors):

Minor comments:

• Results: "we performed non-parametric Spearman's correlations between the structural and functional connectivity maps of beta networks with neurotransmitter uptake". This is a significantly complex analysis that requires more detail for the reader to evaluate. There is more detail in the Figure 3 legend but still insufficient. The Methods offer more detail, but I found the description of the parcellation to be vague and I would appreciate a more detailed description.

We thank the reviewer for bringing the insufficient explanation of the methods used to calculate the correlations in analysis to our attention. We have now made an effort to provide more level of detail in the relevant paragraphs. 

Changes in manuscript: We have now made changes to both the Results and Methods sections and added the following explanations respectively:

Results: “Next, we resliced the beta network map and the PET images to allow for a meaningful comparison, using a combined parcellation with 476 brain regions that include cortex19, basal ganglia20, and cerebellum21. Here, each parcel – which was a collection of voxels belonging to a particular brain region – from the connectivity map was correlated with the same parcel containing average neurotransmitter uptake from the respective PET scan (see Figure 3A). In this way nonparametric Spearman’s correlations between PET intensity and structural and functional connectivity maps of beta networks were obtained, which indicate to what degree the spatial distribution of connectivity is similar to the distribution of neurotransmitter uptake.“

Methods: “A custom master parcellation in MNI space was created in Matlab using SPM functions by combining three existing parcellations to include cortical regions19, structures of the basal ganglia20 and cerebellar regions21. Regions that were (partially) overlapping between the atlases were only selected once. The final compound parcellation had 476 regions in total. This parcellation was applied to both PET and structural and functional connectivity maps using SPM and custom code. This allowed for the calculation of spatial correlations, providing a statistical measure of spatial similarity of the PET intensity and MRI connectivity distributions. For this, Spearman’s ranked correlations were used to calculate correlations between the PET images, such as the dopamine aggregate map and both functional and structural beta connectivity networks (Figure 3). The analysis was repeated for individual tracers showing similar results Supplementary figure 2. Finally, to validate these results, a control analysis was performed using a GABA PET scan from the same open dataset of neurotransmitter uptake following the same pipeline (Figure 2A, 2B).”

• All of the recordings were taken in an eyes-closed condition. This is likely to affect the power of alpha oscillations; the authors should comment on this.

We agree with the reviewer that this will likely have influenced the results. However, given that the key result of our paper is the abundance and circuit topography of beta oscillations, it is unlikely that increased alpha in some channels will have led to false positive results for beta. If anything, it may have increased the contrast leading to a more conservative estimate of which channels truly show strong beta dominance. On the other hand, we should acknowledge that this limitation can affect the interpretation of the alpha result. Another reason for us to primarily focus on beta in the discussion and results presentation. 

Changes in manuscript: We now comment on this in the results:

“It should be noted that that alpha recordings were performed in eyes closed which is known to increase alpha power, which may influence the generalizability of the alpha maps to an eyes open condition. However, given that our primary use of alpha was to act as a control, we believe that this should not affect the interpretability of the key findings of our study.” 

• Although the relative proportion of theta and gamma channels is lower, it would be interesting to see the distribution of channels in a SOM figure.

As described above, we have now added supplementary figure 1 that accommodates the topography but not the network analyses.

• Figure legend - typo - "Neither, alpha nor beta" - no comma needed.

Now fixed, thank you for pointing is to this lapse!

• Results: " ere, we aimed to investigate the whole brain circuit representation of beta activity, which is impossible with current neurophysiology approaches" not entirely accurate; suggest rephrasing it to "Here, we aimed to investigate the whole brain circuit representation of beta activity, which is impossible with non-invasive neurophysiology approaches "

Thank you for suggesting the alternative formulation. 

Changes in manuscript: The text has been modified as per the suggestion and now reads “Here, we aimed to investigate the whole brain circuit representation of beta activity, which is impossible with non-invasive neurophysiology approaches”.

• Results - typo - "cortical brain areas, that exhibit resting beta activity share a common brain network" - no comma needed.

Thank you for the suggestion, the comma has been removed to better the flow of the sentence structure as suggested.

Author response:

eLife Assessment

This useful study presents the first detailed and comprehensive description of brain sulcus anatomy of a range of carnivoran species based on a robust manual labeling model allowing species comparisons. Although the database is recognized and the method for reconstructing cortical surfaces is convincing, the evidence supporting the conclusions is incomplete due to the lack of appropriate quantitative measurements and analyses. Considering additional specimens to assess intraspecies variations, as well as exploring the functional correlates of interspecies differences would increase the scope of the study. Setting an instructive foundation for comparative anatomy, this study will be of interest to neuroscientists and neuroimaging researchers interested in that field, as well as in brain morphology and sulcal patterns, their phylogeny, and ontogeny in relation to functional development and behaviour. 

We are pleased that our primary objective of creating a comprehensive framework to navigate carnivoran brains is considered as successfully achieved and that our work is expected to be of broad interest to various disciplines, as it provides the foundation for future investigations into carnivoran brain organization.

As we will set out below, a description of the major sulci is an appropriate measure for large-scale comparative anatomy — it is stable enough in the population of each species to not require a large N, provides a suitable variability across species, and can be related to other aspects of between-species diversity. We will include a number of additional species to increase the scope of the study, as suggested. Although a quantitative assessment of functional correlates is, in principle, beyond the scope of this first foundational paper, we will provide a first start of this as well. We emphasize, however, that this was a secondary outcome, emerging after first application of the framework.

Public Reviews: 

Reviewer #1 (Public review): 

Summary: 

The paper by Boch and colleagues, entitled Comparative Neuroimaging of the Carnivore Brain: Neocortical Sulcal Anatomy, compares and describes the cortical sulci of eighteen carnivore species, and sets a benchmark for future work on comparative brains. 

Based on previous observations, electrophysiological, histological and neuroimaging studies and their own observations, the authors establish a correspondence between the cortical sulci and gyri of these species. The different folding patterns of all brain regions are detailed, put into perspective in relation to their phylogeny as well as their potential involvement in cortical area expansion and behavioral differences. 

Strengths: 

This is a pioneering article, very useful for comparative brain studies and conducted with great seriousness and based on many past studies. The article is well-written and very didactic. The different protocols for brain collection, perfusion, and scanning are very detailed. The images are self-explanatory and of high quality. The authors explain their choice of nomenclature and labels for sulci and gyri on all species, with many arguments. The opening on ecology and social behavior in the discussion is of great interest and helps to put into perspective the differences in folding found at the level of the different cortexes. In addition, the authors do not forget to put their results into the context of the laws of allometry. They explain, for example, that although the largest brains were the most folded and had the deepest folds in their dataset, they did not necessarily have unique sulci, unlike some of the smaller, smoother brains. 

Weaknesses: 

The article is aware of its limitations, not being able to take into account inter-individual variability within each species, inter-hemispheric asymmetries, or differences between males and females. However, this does not detract from their aim, which is to lay the foundations for a correspondence between the brains of carnivores so that navigation within the brains of these species can be simplified for future studies. This article does not include comparisons of morphometric data such as sulci depth, sulci wall surface, or thickness of the cortical ribbon around the sulci. 

We thank the reviewer for their overwhelmingly positive evaluation of our work. As noted by the reviewer, our primary aim was to establish a framework for navigating carnivoran brains to lay the foundation for future research. We are pleased that this objective is deemed as successfully achieved.

As the reviewer points out, we do not quantify within-species intraindividual differences. This is a conscious choice; we aimed to emphasize breadth of species over individuals, as is standard in large-scale comparative anatomy (cf. Heuer et al., 2023, eLife; Suarez et al., 2022, eLife). Following the logic of phylogenetic relationships, the presence of a particular sulcus in related species is also a measure of reliability. We felt safe in this choice, as previous work in both primates and carnivorans has shown that differences across major sulci across individuals are a matter of degree rather than a case of presence or absence (Connolly, 1950, External morphology of the primate brain, C.C. Thomas; Hecht et al., 2019 J Neurosci; Kawamuro 1971 Acta Anat., Kawamuro & Naito, 1977, Acta Anat.). In our revised manuscript, we aim to include some additional individuals of selected species as supplementary material, further illustrating this point.

We feel that measures such as sulci depth, sulci wall surface, or thickness of the cortical ribbon are measures that vary more across individuals and we have therefore not included them in the study. In addition, these are measures that are not generally used as between-species comparative measures, whereas sulcal patterning is (cf. Amiez et al., 2019, Nat Comms; Connolly, 1950; Miller et al., 2021, Brain Behav Evol; Radinsky 1975, J Mammal; Radinsky 1969, Ann N Y Acad Sci; Welker & Campos 1963 J. Comp Neurol).

Reviewer #2 (Public review): 

Summary: 

The authors have completed MRI-based descriptions of the sulcal anatomy of 18 carnivoran species that vary greatly in behaviour and ecology. In this descriptive study, different sulcal patterns are identified in relation to phylogeny and, to some extent, behaviour. The authors argue that the reported differences across families reflect behaviour and electrophysiology, but these correlations are not supported by any analyses. 

Strengths: 

A major strength of this paper is using very similar imaging methods across all specimens. Often papers like this rely on highly variable methods so that consistency reduces some of the variability that can arise due to methodology. 

The descriptive anatomy was accurate and precise. I could readily follow exactly where on the cortical surface the authors referring. This is not always the case for descriptive anatomy papers, so I appreciated the efforts the authors took to make the results understandable for a broader audience. 

I also greatly appreciate the authors making the images open access through their website. 

Weaknesses: 

Although I enjoyed many aspects of this manuscript, it is lacking in any quantitative analyses that would provide more insights into what these variations in sulcal anatomy might mean. The authors do discuss inter-clade differences in relation to behaviour and older electrophysiology papers by Welker, Campos, Johnson, and others, but it would be more biologically relevant to try to calculate surface areas or volumes of cortical fields defined by some of these sulci. For example, something like the endocast surface area measurements used by Sakai and colleagues would allow the authors to test for differences among clades, in relation to brain/body size, or behaviour. Quantitative measurements would also aid significantly in supporting some of the potential correlations hinted at in the Discussion. 

Although quantitative measurements would be helpful, there are also some significant concerns in relation to the specimens themselves. First, almost all of these are captive individuals. We know that environmental differences can alter neocortical development and humans and nonhuman animals and domestication affects neocortical volume and morphology. Whether captive breeding affects neocortical anatomy might not be known, but it can affect other brain regions and overall brain size and could affect sulcal patterns. Second, despite using similar imaging methods across specimens, fixation varied markedly across specimens. Fixation is unlikely to affect the ability to recognize deep sulci, but variations in shrinkage could nevertheless affect overall brain size and morphology, including the ability to recognize shallow sulci. Third, the sample size = 1 for every species examined. In humans and nonhuman animals, sulcal patterns can vary significantly among individuals. In domestic dogs, it can even vary greatly across breeds. It, therefore, remains unclear to what extent the pattern observed in one individual can be generalized for a species, let alone an entire genus or family. The lack of accounting for inter-individual variability makes it difficult to make any firm conclusions regarding the functional relevance of sulcal patterns. 

We thank the reviewer for their assessment of our work. The primary aim of this study was to establish a framework for navigating carnivoran brains by providing a comprehensive overview of all major neocortical sulci across eighteen different species. Given the inconsistent nomenclature in the literature and the lack of standardized criteria (“recipes”) for identifying the major sulci, we specifically focused on homogenizing the terminology and creating recipes for their identification. Moreover, we also generated digital surfaces of all brains and will also add sulcal masks to further facilitate future research building on our framework. We are pleased to hear that we succeeded in our primary objective.

We respectfully disagree with the reviewer on two accounts, where we believe the reviewer is not judging the scope of the current work.

The first is with respect to individual differences. To the best of our knowledge, differences between captive and wild animals, or indeed between individuals, do not affect the presence or absence of any major sulci. No differences in sulcal patterns were detected between captive and (semi-)wild macaques (cf. Sallet et al., 2011, Science; Testard et al., 2022, Sci Adv), different dog breeds (Hecht et al., 2019 J Neurosci) or foxes selectively bred to simulate domestication, compared to controls (Hecht et al., 2021 J. Neurosci). Indeed, we do not find major differences between wolf-like canid species, suggesting that a difference between individuals of the same species is even more unlikely. Nevertheless, we agree with the reviewer that building up a database like ours will benefit from providing as much information about the samples as possible to enable these issues to be tested. We, therefore, will update our table to include if the animals were from captive or wild populations. Moreover, we aim, where possible, to include both wild and captive animals of the same species if they are available in our revision.

The second is in the quantification of structure/function relationships. We believe the sulci atlases themselves are the main deliverables of this project. We felt it prudent to include some qualitative descriptions of the relationship between sulci as we observed them and behaviours as known from the literature as an illustration of the possibilities that this foundational work opens us. This approach also allowed us to confirm previous findings based on observations from a less diverse range of carnivoran species and families (Radinsky 1968 J Comp Neurol; Radinsky 1969, Ann N Y Acad Sci; Welker & Campos 1963 J Comp Neurol; Welker & Seidenstein, 1959 J Comp Neurol). However, a full statistical framework for analysis is beyond the scope of this paper. Our group has previously worked on methods to quantitatively compare brain organization across species — indeed, we have developed a full framework for doing so (Mars et al., 2021, Annu Rev Neurosci), based on the idea that brains that differ in size and morphology should be compared based on anatomical features in a common feature space. Previously, we have used white matter anatomy (Mars et al., 2018, eLife) and spatial transcriptomics (Beauchamp et al., 2021, eLife). The present work presents the foundation for this approach to be expanded to sulcal anatomy, but the full development of this approach will be the topic of future communications.

Nevertheless, we aim to include a first step quantitative analysis of the relationship between the presence and absence of particular sulci and the two behaviours of interest in our manuscript.

We also would like to emphasize that we strongly believe that looking at measures of brain organization at a more detailed level than brain size or relative brain size is informative. Indeed, studies looking at correlations between brain size and particular behavioural variables, although very prominent in the literature, have found it very difficult to distinguish between competing behavioural hypotheses (Healy, 2021, Adaptation and the brain, OUP). In contrast, connectivity has a much more direct relationship to behavioural differences across species (Bryant et al., 2024, bioRxiv), as does sulcal anatomy (Amiez et al., 2019, Nat Comms; Miller et al., 2021, Brain Behav Evol). Moreover, such measures are less sensitive to the effects of fixation since that will affect brain size but not the presence or absence of a sulcus.

Following the reviewer’s recommendations, we will endeavour to include an even broader range of species in the revised version.

Author response:

Public Reviews: 

Reviewer #1 (Public review): 

Summary: 

The authors address a fundamental question for cell and tissue biology using the skin epidermis as a paradigm and ask how stratifying self-renewing epithelia induce diCerentiation and upward migration in basal dividing progenitor cells to generate suprabasal barrier-forming cells that are essential for a functional barrier formed by such an epithelium. The authors show for the first time that an increase in intracellular actomyosin contractility, a hallmark of barrier-forming keratinocytes, is suCicient to trigger terminal diCerentiation. Hence the data provide in vivo evidence of the more general interdependency of cell mechanics and diCerentiation. The data appear to be of high quality and the evidences are strengthened through a combination of diCerent genetic mouse models, RNA sequencing, and immunofluorescence analysis. 

To generate and maintain the multilayered, barrier-forming epidermis, keratinocytes of the basal stem cell layer diCerentiate and move suprabasally accompanied by stepwise changes not only in gene expression but also in cell morphology, mechanics, and cell position. Whether any of these changes is instructive for diCerentiation itself and whether consecutive changes in diCerentiation are required remains unclear. Also, there are few comprehensive data sets on the exact changes in gene expression between diCerent states of keratinocyte diCerentiation. In this study, through genetic fluorescence labeling of cell states at diCerent developmental time points the authors were able to analyze gene expression of basal stem cells and suprabasal diCerentiated cells at two diCerent stages of maturation: E14 (embryonic day 14) when the epidermis comprises mostly two functional compartments (basal stem cells and suprabasal so-called intermediate cells) and E16 when the epidermis comprise three (living) compartments where the spinous layer separates basal stem cells from the barrier-forming granular layer, as is the case in adult epidermis. Using RNA bulk sequencing, the authors developed useful new markers for suprabasal stages of diCerentiation like MafB and Cox1. The transcription factor MafB was then shown to inhibit suprabasal proliferation in a MafB transgenic model. 

The data indicate that early in development at E14 the suprabasal intermediate cells resemble in terms of RNA expression, the barrier-forming granular layer at E16, suggesting that keratinocytes can undergo either stepwise (E16) or more direct (E14) terminal diCerentiation. 

Previous studies by several groups found an increased actomyosin contractility in the barrier-forming granular layer and showed that this increase in tension is important for epidermal barrier formation and function. However, it was not clear whether contractility itself serves as an instructive signal for diCerentiation. To address this question, the authors use a previously published model to induce premature hypercontractility in the spinous layer by using spastin overexpression (K10-Spastin) to disrupt microtubules (MT) thereby indirectly inducing actomyosin contractility. A second model activates myosin contractility more directly through overexpression of a constitutively active RhoA GEF (K10Arhgef11CA). Both models induce late diCerentiation of suprabasal keratinocytes regardless of the suprabasal position in either spinous or granular layer indicating that increased contractility is key to induce late diCerentiation of granular cells. A potential weakness of the K10-spastin model is the disruption of MT as the primary eCect which secondarily causes hypercontractility. However, their previous publications provided some evidence that the eCect on diCerentiation is driven by the increase in contractility (Ning et al. cell stem cell 2021). Moreover, the data are confirmed by the second model directly activating myosin through RhoA. These previous publications already indicated a role for contractility in diCerentiation but were focused on early diCerentiation. The data in this manuscript focus on the regulation of late diCerentiation in barrier-forming cells. These important data help to unravel the interdependencies of cell position, mechanical state, and diCerentiation in the epidermis, suggesting that an increase in cellular contractility in most apical positions within the epidermis can induce terminal diCerentiation. Importantly the authors show that despite contractility-induced nuclear localization of the mechanoresponsive transcription factor YAP in the barrier-forming granular layer, YAP nuclear localization is not suCicient to drive premature diCerentiation when forced to the nucleus in the spinous layer. 

Overall, this is a well-written manuscript and a comprehensive dataset. Only the RNA sequencing result should be presented more transparently providing the full lists of regulated genes instead of presenting just the GO analysis and selected target genes so that this analysis can serve as a useful repository. The authors themselves have profited from and used published datasets of gene expression of the granular cells. Moreover, some of the previous data should be better discussed though. The authors state that forced suprabasal contractility in their mouse models induces the expression of some genes of the epidermal diCerentiation complex (EDC). However, in their previous publication, the authors showed that major classical EDC genes are actually not regulated like filaggrin and loricrin (Muroyama and Lechler eLife 2017). This should be discussed better and necessitates including the full list of regulated genes to show what exactly is regulated. 

We thank all the reviewers for their suggestions and comments.

Thank you especially for the reminder to include gene lists. We had an excel document with all this data but neglected to upload it with the initial manuscript decision. This includes all the gene signatures for the diCerent cell compartments across development. We will also include a page that lists all EDC genes and whether they were up-regulated in intermediate cells and cells in which contractility was induced. Further, we note that all the RNA-Seq datasets are available for use on GEO. 

In our previous publication, we indeed included images showing a lack of change in loricrin and filaggrin in the embryos where spastin was expressed in the diCerentiated epidermis. Consistent with this, there is no change in Lor mRNA levels by RNA-Seq, (it is one of the rare EDC genes that is unchanged). In contrast, Flg mRNA was up in the RNASeq, though we didn’t see a dramatic change in protein levels. We have not further pursued whether this reflects translational regulation. That said, our data clearly show that other genes associated with granular fate were increased in the contractile skin.  

Reviewer #2 (Public review): 

Summary: 

The manuscript from Prado-Mantilla and co-workers addresses mechanisms of embryonic epidermis development, focusing on the intermediate layer cells, a transient population of suprabasal cells that contributes to the expansion of the epidermis through proliferation. Using bulk-RNA they show that these cells are transcriptionally distinct from the suprabasal spinous cells and identify specific marker genes for these populations. They then use transgenesis to demonstrate that one of these selected spinous layer-specific markers, the transcription factor MafB is capable of suppressing proliferation in the intermediate layers, providing a potential explanation for the shift of suprabasal cells into a non-proliferative state during development. Further, lineage tracing experiments show that the intermediate cells become granular cells without a spinous layer intermediate. Finally, the authors show that the intermediate layer cells express higher levels of contractilityrelated genes than spinous layers and overexpression of cytoskeletal regulators accelerates the diCerentiation of spinous layer cells into granular cells. 

Overall the manuscript presents a number of interesting observations on the developmental stage-specific identities of suprabasal cells and their diCerentiation trajectories and points to a potential role of contractility in promoting diCerentiation of suprabasal cells into granular cells. The precise mechanisms by which MafB suppresses proliferation, how the intermediate cells bypass the spinous layer stage to diCerentiate into granular cells, and how contractility feeds into these mechanisms remain open. Interestingly, while the mechanosensitive transcription factor YAP appears deferentially active in the two states, it is shown to be downstream rather than upstream of the observed diCerences in mechanics. 

Strengths: 

The authors use a nice combination of RNA sequencing, imaging, lineage tracing, and transgenesis to address the suprabasal to granular layer transition. The imaging is convincing and the biological eCects appear robust. The manuscript is clearly written and logical to follow. 

Weaknesses: 

While the data overall supports the authors' claims, there are a few minor weaknesses that pertain to the aspect of the role of contractility, The choice of spastin overexpression to modulate contractility is not ideal as spastin has multiple roles in regulating microtubule dynamics and membrane transport which could also be potential mechanisms explaining some of the phenotypes. Use of Arghap11 overexpression mitigates this eCect to some extent but overall it would have been more convincing to manipulate myosin activity directly. It would also be important to show that these manipulations increase the levels of F-actin and myosin II as shown for the intermediate layer. It would also be logical to address if further increasing contractility in the intermediate layer would enhance the diCerentiation of these cells. 

We agree with the reviewer that the development of additional tools to precisely control myosin activity will be of great use to the field. That said, our series of publications has clearly demonstrated that ablating microtubules results in increased contractility and that this phenocopies the eCects of Arhgef11 induced contractility (Ning et al, Cell Stem Cell 2021). Further, we showed that these phenotypes were rescued by myosin inhibition with blebbistatin. Our prior publications also showed a clear increase in junctional acto-myosin through expression of either spastin or Arhgef11, as well as increased staining for the tension sensitive epitope of alpha-catenin (alpha-18) (also in Ning et al, 2021).  We are not aware of tools that allow direct manipulation of myosin activity that currently exist in mouse models.  

The gene expression analyses are relatively superficial and rely heavily on GO term analyses which are of course informative but do not give the reader a good sense of what kind of genes and transcriptional programs are regulated. It would be useful to show volcano plots or heatmaps of actual gene expression changes as well as to perform additional analyses of for example gene set enrichment and/or transcription factor enrichment analyses to better describe the transcriptional programs 

We will include an excel document that lists all the gene signatures. Additionally, all of our data are deposited in GEO for others to perform their own analyses.  

Claims of changes in cell division/proliferation changes are made exclusively by quantifying EdU incorporation. It would be useful to more directly look at mitosis. At minimum Y-axis labels should be changed from "% Dividing cells" to % EdU+ cells to more accurately represent findings 

We will change the axis label to precisely match our analysis.  

Despite these minor weaknesses the manuscript is overall of high quality, sheds new light on the fundamental mechanisms of epidermal stratification during embryogenesis, and will likely be of interest to the skin research community. 

Reviewer #3 (Public review): 

Summary: 

This is an interesting paper by Lechler and colleagues describing the transcriptomic signature and fate of intermediate cells (ICs), a transient and poorly defined embryonic cell type in the skin. ICs are the first suprabasal cells in the stratifying skin and unlike laterdeveloping suprabasal cells, ICs continue to divide. Using bulk RNA seq to compare ICs to spinous and granular transcriptomes, the authors find that IC-specific gene signatures include hallmarks of granular cells, such as genes involved in lipid metabolism and skin barrier function that are not expressed in spinous cells. ICs were assumed to diCerentiate into spinous cells, but lineage tracing convincingly shows ICs diCerentiate directly into granular cells without passing through a spinous intermediate. Rather, basal cells give rise to the first spinous cells. They further show that transcripts associated with contractility are also shared signatures of ICs and granular cells, and overexpression of two contractility inducers (Spastin and ArhGEF-CA) can induce granular and repress spinous gene expression. This contractility-induced granular gene expression does not appear to be mediated by the mechanosensitive transcription factor, Yap. The paper also identifies new markers that distinguish IC and spinous layers and shows the spinous signature gene, MafB, is suCicient to repress proliferation when prematurely expressed in ICs. 

Strengths: 

Overall this is a well-executed study, and the data are clearly presented and the findings convincing. It provides an important contribution to the skin field by characterizing the features and fate of ICs, a much-understudied cell type, at high levels of spatial and transcriptomic detail. The conclusions challenge the assumption that ICs are spinous precursors through compelling lineage tracing data. The demonstration that diCerentiation can be induced by cell contractility is an intriguing finding and adds a growing list of examples where cell mechanics influence gene expression and diCerentiation. 

Weaknesses: 

A weakness of the study is an over-reliance on overexpression and suCiciency experiments to test the contributions of MafB, Yap, and contractility in diCerentiation. The inclusion of loss-of-function approaches would enable one to determine if, for example, contractility is required for the transition of ICs to granular fate, and whether MafB is required for spinous fate. Second, whether the induction of contractility-associated genes is accompanied by measurable changes in the physical properties or mechanics of the IC and granular layers is not directly shown. The inclusion of physical measurements would bolster the conclusion that mechanics lies upstream of diCerentiation. 

We agree that loss of function studies would be useful. For MafB, these have been performed in cultured human keratinocytes, where loss of MafB and its ortholog cMaf results in a phenotype consistent with loss of spinous diCerentiation (Lopes-Pajares, Dev Cell 2015). Due to the complex genetics involved, generating these double mutant mice is beyond the scope of this study. Loss of function studies of myosin are also complicated by genetic redundancy of the non-muscle type II myosin genes, as well as the role for these myosins in actin cross linking in addition to contractility. In addition, we have found that these myosins are quite stable in the embryonic intestine, with loss of protein delayed by several days from the induction of recombination. Therefore, elimination of myosins by embryonic day e14.5 with our current drivers is not likely possible. Thus, generation of inducible inhibitors of contractility is a valuable future goal. 

A number of recent papers have used AFM of skin sections to probe tissue rigidity. We have not attempted these studies and are unclear about the spatial resolution and whether, in the very thin epidermis at these stages we could spatially resolve diCerences. That said, we previously assessed the macro-contractility of tissues in which myosin activity was induced and demonstrated that there was a significant increase in this over a tissue-wide scale (Ning et al, Cell Stem Cell, 2021).  

Finally, whether the expression of granular-associated genes in ICs provides them with some sort of barrier function in the embryo is not addressed, so the role of ICs in epidermal development remains unclear. Although not essential to support the conclusions of this study, insights into the function of this transient cell layer would strengthen the overall impact.  

By traditional dye penetration assays, there is no epidermal barrier at the time that intermediate cells exist. One interpretation of the data is that cells are beginning to express mRNAs (and in some cases, proteins) so that they are able to rapidly generate a barrier as they become granular cells. We have attempted experiments to ablate intermediate cells with DTA expression - this resulted in ineCicient and delayed cell death and thus did not yield strong conclusions. Our findings that transcriptional regulators of granular diCerentiation (such as Grhl3 and Hopx) are also present in intermediate cells, should allow future analysis of the eCects of their ablation on the earliest stages of granular diCerentiation from intermediate cells.

Author response:

The following is the authors’ response to the original reviews.

Reviewer #1 (Public Review):

This paper aims to address the establishment and maintenance of neural circuitry in the case of a massive loss of neurons. The authors used genetic manipulations to ablate the principal projection neurons, the mitral/tufted cells, in the mouse olfactory bulb. Using diphtheria toxin (Tbx21-Cre:: loxP-DTA line) the authors ablated progressively large numbers of M/T cells postnatally. By injecting diphtheria toxin (DT) into the Tbx21-Cre:: loxP-iDTR line, the authors were able to control the timing of the ablation in the adult stage. Both methods led to the successful elimination of a majority of M/TCs by 4 months of age. The authors made a few interesting observations. First, they found that the initial pruning of the remaining M/T cell primary dendrite was unaffected. However, in adulthood, a significant portion of these cells extended primary dendrites to innervate multiple glomeruli. Moreover, the incoming olfactory sensory neuron (OSN) axons, as examined for those expressing the M72 receptor, showed a divergent innervation pattern as well. The authors conclude that M/T cell density is required to maintain the dendritic structures and the olfactory map. To address the functional consequences of eliminating a large portion of principal neurons, the authors conducted a series of behavioral assays. They found that learned odor discrimination was largely intact. On the other hand, mating and aggression were reduced. The authors concluded that learned behaviors are more resilient than innate ones.

The study is technically sound, and the results are clear-cut. The most striking result is the contrast between the normal dendritic pruning during early development and the expanded dendritic innervation in adulthood. It is a novel discovery that can lead to further investigation of how the single-glomerulus dendritic innervation is maintained. The authors conducted a

few experiments to address potential mechanisms, but it is inconclusive, as detailed below. It is also interesting to see that the massive neuronal loss did not severely impact learned odor discrimination. This result, together with previous studies showing nearly normal odor discrimination in the absence of large portions of the olfactory bulb or scrambled innervation patterns, attests to the redundancy and robustness of the sensory system. The discussion should take into account these other studies in a historical context.

Main comments:

(1) In previous studies, it has been concluded that dendritic pruning unfolds independently, regardless of the innervation pattern or activity of the OSNs. The new observation bolsters this conclusion by showing that a loss of neighboring M/T cells does not affect the developmental process. A more nuanced discussion comparing the results of these studies would strengthen the paper.

We thank the reviewer for the suggestion. We now include an extended discussion citing relevant previous works in the manuscript (Lines 351-374).

(2) The authors propose that a certain density of M/T is required to prevent the divergent innervation of primary dendrites, but the evidence is not sufficient to support this proposal. The experiment with low-dose DT injection to ablate a smaller portion of M/T cells did not change the percentage of cells innervating two or more glomeruli. The authors suggest that a threshold must be met, but this threshold is not determined.  

In our experiments using high-dose DT, we hypothesized that there may be many empty glomeruli (glomeruli not innervated by M/T cells), and as a result, that some of the remaining M/T cells could branch their apical dendrite tuft into multiple empty glomeruli. To test this hypothesis, we carried out another experiment using a lower dose of DT. In this experiment, the fraction of remaining M/T cells was 25% (~10,000 M/T cells), which was higher than with the high DT dose (5%, or around 2,000 M/T cells) , but still significantly lower than wild type mice (~40,000 cells M/T cells). With around 2,000 glomeruli and 10,000 M/T per bulb, it could be expected that each glomerulus would be innervated by ~5 M/T cells (on average). However, we found that the percentage of M/T cells projecting to multiple glomeruli (around 40%) was similar when either 10,000 or 2,000 of M/T remained in the bulb. In addition, it is important to emphasize that even in wt animals with a full set of M/T cells, a small percentage of M/T cells still innervate more than one glomerulus (Lin et al., 2000). Together, these observations suggest that the innervation of multiple glomeruli by M/T cells is not simply due to the presence of empty glomeruli, and that our hypothesis was not correct.

We have added a comment explaining this issue in the Results section (Lines 200-203).

(3) The authors suggest that neural activity is not required for this plasticity. The evidence was derived primarily from naris occlusion and neuronal silencing using Kir2.1. While the results are consistent with the notion, it is a rather narrow interpretation of how neural activity affects circuit configuration. Perturbation of neural activity also entails an increase in firing. Inducing the activity of the neurons may alter this plasticity. Silencing per se may induce a homeostatic response that expands the neurite innervation pattern to increase synaptic input to compensate for the loss of activity. Thus, further silencing the cells may not reduce multiglomerular innervation, but an increased activity may.

The experiments with Kir2.1 demonstrate that the structural plasticity observed after reducing the total number of M/T cells in an animal is not regulated by the firing action potentials in the remaining cells. Instead, this experiment indicates that the observed structural plasticity may be regulated by other types of mechanisms (including increased synaptic excitation as suggested by the reviewer) that do not require the firing of action potentials in M/T cells. 

We now have included a comment regarding this point (Lines 243-247).  

(4) There is a discrepancy between this study and the one by Fujimoto et al. (Developmental Cell; 2023), which shows that not only glutamatergic inputs to the primary dendrite can facilitate pruning of remaining dendrites but also Kir2.1 overexpression can significantly perturb dendritic pruning. This discrepancy is not discussed by the authors.

We agree that it would be useful to contrast these two works.

In our experiments, performed in adult animals, we blocked sensory input by performing naris occlusion before we induced ablation of M/T cells. In a separate experiment, also in adult animals, we expressed the Kir2.1 channel, to reduce the ability of neurons to fire action potentials. With both types of manipulations, we observed that the ablation of a large fraction of M/T cells still caused the remaining M/T cells to maintain a single apical dendrite that sprouts several new tufts towards multiple glomeruli. A recent paper (Fujimoto et al., 2023)) in which Kir2.1 was expressed in a large percentage of M/T starting during embryonic development showed that these “silent” M/T cells failed to prune their arbors to a single dendrite. In aggregate, these observations indicate that action potentials are necessary for the normal pruning that occurs during perinatal development (Fujimoto et al., 2023), but are not required for the expansion of dendritic trees caused by ablating a large fraction of M/T cells in adult animals (our current manuscript).

We have now explained the differences between both studies in the manuscript (Lines 427-439).

(5) An alternative interpretation of the discrepancy between the apparent normal pruning by p10 and expanded dendritic innervation in adulthood is that there are more cells before P10, when ~25% of M/T cells are present, but at a later date only 1-3% are present. 

The relationship between the number of M/T cells and single glomerulus innervation has not been explored during postnatal development. It would be important to test this hypothesis.

We agree with this comment, and in lines 375-381 we discuss the discrepancy between normal refinement during development, and dendritic sprouting in adults.

Cre is expressed in M/T cells and it induces DTA expression starting around P0. The elimination of M/T cells starts at this time, and continues until by P10, when more than 75% of M/T have been eliminated. At P21 more than 90% of M/T have been eliminated, and their number remains stable thereafter.

Pruning of the dendrites of M/T cells starts at P0 and it is mostly complete by P10. Therefore, it is possible that between P0 and P7, when dendrites are being pruned, the number of M/T cells remaining in the bulb is still over a threshold that does not interfere with the process of normal dendrite pruning. We agree that it would be very informative to perform additional experiments in the future where a large set of M/T cells could be ablated before pruning occurs (ideally before P0). 

(6) The authors attribute the change in the olfactory map to the loss of M/T cells. Another obvious possibility is that the diffused projection is a response to the change in the olfactory bulb size. With less space to occupy, the axons may be forced to innervate neighboring glomeruli. It is not known how the total number of glomeruli is affected. This question could be addressed by tracking developmental changes in bulb volume and glomerular numbers.

Certainly, this is a possibility, and we have now included a comment on this regard in the manuscript (Lines 473-480). 

We believe that there are three likely scenarios that could account for these observations:

(a) After ablating M/T cells, the tufts of the remaining M/T cells sprout into multiple glomeruli, and this causes the axons of OSNs to project into multiple glomeruli.

(b) Ablating M/T cells may cause changes in other OB cells that make synapses in the glomeruli (ETCs, PGCs, sAC, etc…), and the misrouting of OSN axons that we observed in our experiments may be a secondary effect caused by the elimination of M/T cells.

(c) After ablating the majority of M/T cells, the olfactory bulb gets reduced in size, and the axons of OSNs find it difficult to precisely converge on a target that now has become smaller. As a result, the axons of OSNs fail to converge on single glomeruli.

(7) The retained ability to discriminate odors upon reinforced training is not surprising in light of a number of earlier studies. For example, Slotnick and colleagues have shown that rats losing ~90% of the OB can retain odor discrimination. Weiss et al have shown that humans without an olfactory bulb can perform normal olfactory tasks. Gronowitz et al have used theoretical prediction and experimental results to demonstrate that perturbing the olfactory map does not have a major impact on olfactory discrimination. Fleischmann et al have shown that mice with a monoclonal nose can discriminate odors. The authors should discuss their results in these contexts.

We apologize for this important oversight - we now include a more elaborate discussion including the relevant references as suggested in the manuscript (Line 483-496).

(8) It should be noted that odor discrimination resulting from reinforcement training does not mean normal olfactory function. It is a highly artificial situation as the animals are overtrained. It should not be used as a measure of the robustness of the olfactory sense. Natural odor discrimination (without training), detection threshold, and innate appetitive/aversive response to certain odors may be affected. These experiments were not conducted.

We agree that the standard tests commonly used to measure olfactory function require substantial training, and thus, are quite artificial. However, these tests are used because they allow a more precise quantification of olfactory function than those relying on natural behaviors.  

We have now included a few sentences to address this point in the results (Lines 321322) and discussion sections (Lines 541-543).

(9) The social behaviors were conducted using relatively coarse measures (vaginal plug and display of aggression). Moreover, these behaviors are most likely affected by the disruption of the AOB mitral cells and have little to do with the dendritic pruning process described in the paper. It is misleading to lump social behaviors with innate responses to odors.

This point follows the same logic as the previous one. The olfactory tests that rely on natural behaviors are quite coarse and difficult to quantify. In contrast, the olfactory tests using apparatuses such as olfactometers can be quantified with precision, but they are artificial. We agree that some of the naturalistic behaviors that we studied such as mating or aggression may depend to a large extent on the AOB (although it is possible that the MOB may also be involved in these tasks to a degree). In our initial version of the manuscript, we commented on the anticipated relative involvement of the MOB and AOB in the studied tasks, but we have now added some additional sentences to make this point clearer. In addition, we now add a comment indicating that it is possible that the abnormal behaviors could simply be due to a reduction in the number of AOB M/T cells (~98.5% and ~ 85% elimination of M/T cells in the AOB in Tbx::DTA and Tbx::iDTR mice, respectively), regardless of the abnormal dendritic pruning of main OB M/T cells (Lines 530-534).

See Figure 5E - M/T cells in AOB (Lines 1238-1239). 

Reviewer #2 (Public Review):

The authors make the interesting observation that the developmental refinement of apical M/T cell dendrites into individual glomeruli proceeds normally even when the majority of neighboring M/T cells are ablated. At later stages, the remaining neurons develop additional dendrites that invade multiple glomeruli ectopically, and similarly, OSN inputs to glomeruli lose projection specificity as well. The authors conclude that the normal density of M/T neurons is not required for developmental refinement, but rather for maintaining specific connectivity in adults.

The observations are indeed quite striking; however, the authors' conclusions are not entirely supported by the data.

(1) It is unclear whether the expression of diphtheria toxin that eventually leads to the ablation of the large majority of M/T neurons compromises the cell biology of the remaining ones.

DT is an extremely potent toxin that kills cells by inhibiting proteins translation, and it has been demonstrated that the presence of a single DT molecule in a cell is sufficient to kill it, because of its highly efficient catalytic activity. Accordingly, previous experiments have shown that DT kills cells within a few hours after its appearance in the cytoplasm (Yamaizumi et al., 1978). In other words, all the published evidence suggests that if a cell is exposed to the action of DT, that cell will die shortly. There is no evidence that cells exposed to DT can survive and experience long-term effects. Finally, previous works have not observed any long-term changes in neurons directly caused by the actions of DT (Johnson et al., 2017).

(2) The authors interpret the growth of ectopic dendrites later in life as a lack of maintenance of dendrite structure; however, maybe the observed changes reflect actually adaptations that optimize wiring for extremely low numbers of M/T neurons. The finding that olfactory behavior was less affected than predicted supports this interpretation.

We do not know the cellular or molecular mechanisms that explain why reducing the density of M/T cells is followed by the growth of ectopic dendrites from the remaining M/T cells. We agree that the functional outcome of growing ectopic dendrites may result in an optimization of wiring in the bulb and could explain why olfactory function is relatively preserved. We now include a comment regarding this possibility (Lines 513-525).   

(3) The number of remaining M/T neurons is much higher at P10 than later. Can the relatively large number of remaining neurons (or their better health status) be the reason that dendrites refine normally at the early developmental stages rather than a (currently unknown) developmental capacity that preserves refinement?

We thank the reviewer for the suggestion, which was also raised by reviewer 1. 

We agree with this comment, and in lines 375-381 we discuss the discrepancy between normal refinement during development, and dendritic sprouting in adults.

Cre is expressed in M/T cells and it induces DTA expression starting around P0. The elimination of M/T cells starts at this time, and continues until by P10, when more than 75% of M/T have been eliminated. At P21 more than 90% of M/T have been eliminated, and their number remains stable thereafter.

Pruning of the dendrites of M/T cells starts at P0 and it is mostly complete by P10. Therefore, it is possible that between P0 and P7, when dendrites are being pruned, the number of M/T cells remaining in the bulb is still over a threshold that does not interfere with the process of normal dendrite pruning. We agree that it would be very informative to perform additional experiments in the future where a large set of M/T cells could be ablated before pruning occurs (ideally before P0). 

(4) While the effect of reduced M/T neuron density on both M/T dendrites and OSN axons is described well, the relationship between both needs to be characterized better: Is one effect preceding the other or do they occur simultaneously? Can one be the consequence of the other?

Previous works have demonstrated that disrupting the topographic projection of the OSN axons has no effect on the structure of the apical dendrite of M/T cells (Ma et al., 2014; Nishizumi et al., 2019). Our experiments ablating a large fraction of M/T cells suggest that they are necessary for the correct targeting of OSN axons into the bulb. However, our experiments do not allow us to tell apart these 2 scenarios: 

(a) the ablation of a large fraction of M/T cells directly causes the sprouting of the apical dendrite of M/T cells, and that this sprouting in turn causes the abnormal projection of OSN axons onto the bulb. 

(b) the ablation of a large fraction of M/T cells first causes the axons of OSN to project abnormally onto multiple glomeruli in the bulb, and this in turn causes the dendrite of remaining M/T cells to sprout onto multiple glomeruli. 

We now include a comment on the manuscript explaining this point. (Lines 473-492)

(5) Page 7: the observation that not all neurons develop additional dendrites is not a sign of differences between cell types, it may be purely stochastic.

This is correct, and we mention these 2 scenarios in the discussion (Line 407-408). 

(6) Page 8: the fact that activity blockade did not affect the formation of ectopic dendrites does not suggest that the process is not activity-dependent: both manipulations have the same effect and may just mask each other.

The experiments with Kir2.1 demonstrate that the structural plasticity observed after reducing the total number of M/T cells in an animal is not regulated by the firing action potentials in the remaining cells. Instead, this experiment indicates that the observed structural plasticity may be regulated by other types of mechanisms (including increased synaptic excitation as suggested by the reviewer) that do not require the firing of action potentials in M/T cells. 

We now have included a comment regarding this point (Lines 243-247).  

(7) It remains unclear how the observed structural changes can explain the behavioral effects.

We agree that the relationship between structural changes and behavior was not appropriately explained in our manuscript. Our manipulations cause two major changes in the olfactory system, one primary, and several secondary. The primary change is a large reduction in the number of M/T cells both in the MOB and AOB. This reduction in M/T cell number triggers significant secondary changes in the connectivity of the bulb, including an abnormal projection of OSNs onto the OB, and the growth of ectopic dendrites from the remaining M/T cells into multiple glomeruli.

The behavioral abnormalities displayed by these mice is ultimately caused by the reduction in the number of M/T cells, but it is likely that the secondary structural changes could regulate some of the behavioral phenomena that we observed. For example, in principle, it is possible that the ectopic dendrites innervating several glomeruli could help the bulb to perceive smells with a much reduced number of M/T cells. On the other hand, this promiscuous growth of dendrites into multiple glomeruli could make it more difficult for the animals to discriminate between smells. The same argument could be made about the fact that OSN axons project onto multiple glomeruli: we simply do not know if this change helps or makes it more difficult for the animal to detect smells.  

We now include a comment regarding this issue (Lines 513-525).   

Reviewer #1 (Recommendations For The Authors):

Additional experiments and a more thorough discussion of the results, as suggested in the public review, would significantly strengthen the paper. Below are some specific parts that need to be addressed.

There is a lack of information on how M/T cell numbers are quantified. Without the information, it is difficult to evaluate the claim. Using the tdTomato signal may miss cells that are not labeled due to the transgenic effect. 

Although we cannot conclude that we are identifying the complete set of M/T cells (because the transgenic lines may fail to label some M/T cells), the number of M/T cells that we observed is similar to that previously reported (Richard et al., 2010). This concern has been included in the Results section (Lines 121-124).

A more detailed description about M/T cells quantification has been added into the method section (Lines 627-632).

There is a lack of information on the timeline of treatment and how measurement of the olfactory bulb volume is conducted.

We now include a more detailed description of how the volume of the OB was measured in the methods (Lines 621-623).

The volume measurement is inconsistent with the pictures shown. In Figure 1, supplemental data 2 panels B and C, it appears that the bulbs in DTA and DTR mice are about half in length in each dimension. This would translate into ~1/8 of the volume of the control mice.

We measured the volume of the bulbs based on the Neurolucida reconstructions, and we observed that in both DTA and iDTR mice the volumes of their bulbs are roughly 50% compared to a wild type mouse. In Figure 1 - figure supplement 2 the sections that were shown for wild type, DTA and iDTR mice were not taken at the same position in the bulb, and this gave the impression that the bulbs from DTA and iDTR were much smaller than they really are. We now show sections for these three animals at equivalent positions in the bulb. 

Figure 1 E and F have no legend.

We apologize for this mistake - we have now added the legend for Figures 1E and F (Lines 1009-1013).

Figure 3, supplemental data 2, it is not clear what the readers should be looking at. The data is confusing even for experts in the field. The authors should describe the figures more clearly, pointing out what they are supposed to show.

We apologize for this, and we have now added a more detailed description of Figure3 – figure supplement 2 (Lines 1153-1167).

In several figures, it is not clearly written what the comparisons were for where there are indications of statistical significance above the bars.

We have now included a more detailed description of the statistics comparison in the figure legends.

AAV serotype should be specified.

The AAV serotype used to label M/T cells was the AAV-PHP.eB. We have added this information in the methods section of the manuscript. 

Reviewer #2 (Recommendations For The Authors):

Minor points

Page 5, para 2: "The decrease in neuronal plasticity with age": it is unclear what "the decrease" refers to.

We have changed this sentence in the text to make it clear:

“The decrease in structural plasticity of M/T cells after apical dendrite refinement (Mizrahi and Katz, 2003),….”

Line 146-148

Is there a quantification of the effect of Kir2.1 overexpression alone (example shown in Figure 3D)?

We did an experiment in IDTR animals in which a fraction of M/T cells expressed Kir2.1, and we split these animals in 2 groups: (a) animals that received an injection of DT, and (b) animals that did not receive any DT. We quantified the effect of Kir2.1 on M/T cells from animals that received DT injection (with an ablation of around of 90% of M/T cells) and we did not observe any clear statistically significant differences between cells expressing Kir2.1 or neurons that did not express Kir2.1 from other iDTR animals that also received DT injections. We did not quantify the possible effects of kir2.1 in the group of animals that did not receive DT because on a first inspection we did not observe any clear differences between Kir2.1 cells and neighboring wild type cells. 

References

Fujimoto S, Leiwe MN, Aihara S, Sakaguchi R, Muroyama Y, Kobayakawa R, Kobayakawa K, Saito T, Imai T. 2023. Activity-dependent local protection and lateral inhibition control synaptic competition in developing mitral cells in mice. Dev Cell S1534-5807(23)00237-X. doi:10.1016/j.devcel.2023.05.004

Johnson RE, Tien N-W, Shen N, Pearson JT, Soto F, Kerschensteiner D. 2017. Homeostatic plasticity shapes the visual system’s first synapse. Nat Commun 8:1220. doi:10.1038/s41467-017-01332-7

Lin DM, Wang F, Lowe G, Gold GH, Axel R, Ngai J, Brunet L. 2000. Formation of precise connections in the olfactory bulb occurs in the absence of odorant-evoked neuronal activity. Neuron 26:69–80. doi:10.1016/s0896-6273(00)81139-3

Ma L, Wu Y, Qiu Q, Scheerer H, Moran A, Yu CR. 2014. A developmental switch of axon targeting in the continuously regenerating mouse olfactory system. Science 344:194–197. doi:10.1126/science.1248805

Nishizumi H, Miyashita A, Inoue N, Inokuchi K, Aoki M, Sakano H. 2019. Primary dendrites of mitral cells synapse unto neighboring glomeruli independent of their odorant receptor identity. Commun Biol 2:1–12. doi:10.1038/s42003-018-0252-y

Richard MB, Taylor SR, Greer CA. 2010. Age-induced disruption of selective olfactory bulb synaptic circuits. Proc Natl Acad Sci U S A 107:15613–15618. doi:10.1073/pnas.1007931107

Yamaizumi M, Mekada E, Uchida T, Okada Y. 1978. One molecule of diphtheria toxin fragment A introduced into a cell can kill the cell. Cell 15:245–250. doi:10.1016/0092-8674(78)90099-5

Author response:

The following is the authors’ response to the original reviews.

Reviewer #1 (Public Review):

Summary:

In their manuscript, "Nicotine enhances the stemness and tumorigenicity in intestinal stem cells via Hippo-YAP/TAZ and Notch signal pathway", authors Isotani et al claimed that this study identifies a NIC-triggered pathway regulating the stemness and tumorigenicity of ISCs and suggest the use of DBZ as a potential therapeutic strategy for treating intestinal tumors. However, the presented data do not support the primary claims.

Weaknesses:

My main reservation is that the quality of the results presented in the manuscript may not fully substantiate their conclusions. For instance, in Figure 2 A and B, it is challenging to discern a healthy organoid. This is significant, as the entirety of Figure 2 and several panels in Figures 3 - 5 are based on these organoid assays. Additionally, there seems to be a discrepancy in the quality of results from the western blot, as the lanes of actin do not align with other proteins (Figure 6B).

We directly count organoids under microscopy as described previously (Igarashi M et.al., Cell.2016 Igarashi M et.al., Aging Cell.2019). When we count the number of organoids, we exactly can discern which are alive or dead organoids under microscope. Hence, we will detail the method and show which are alive or dead organoids using arrows in our revised version (Figure2A and B).

Moreover, as reviewer1 pointed out, the number of organoids originated from intestinal or colonic crypts can be affected by dead organoids as in Figure2A and 2B. However, almost all colonies from isolated intestinal stem cells (ISCs) (Figure 2C and D) are alive, so the number of colonies are less affected by dead colonies in those experiments using isolated ISCs. Since all organoid data in Figure 3-5 are based on the same method as that of Figure2C and D, the data quality of Figures 3-5 cannot be affected by dead colonies.

Finally, to improve data quality of Figure6B, we repeated this experiments and replaced it by new figures.

Reviewer #2 (Public Review):

Summary:

The manuscript by Isotani et al characterizes the hyperproliferation of intestinal stem cells (ISCs) induced by nicotine treatment in vivo. Employing a range of small molecule inhibitors, the authors systematically investigated potential receptors and downstream pathways associated with nicotine-induced phenotypes through in vitro organoid experiments. Notably, the study specifically highlights a signaling cascade involving α7-nAChR/PKC/YAP/TAZ/Notch as a key driver of nicotine-induced stem cell hyperproliferation. Utilizing a Lgr5CreER Apcfl/fl mouse model, the authors extend their findings to propose a potential role of nicotine in stem cell tumorgenesis. The study posits that Notch signaling is essential during this process.

Strengths and Weaknesses:

One noteworthy research highlight in this study is the indication, as shown in Figure 2 and S2, that the trophic effect of nicotine on ISC expansion is independent of Paneth cells. In the Discussion section, the authors propose that this independence may be attributed to distinct expression patterns of nAChRs in different cell types. To further substantiate these findings, it is suggested that the authors perform tissue staining of various nAChRs in the small intestine and colon. This additional analysis would provide more conclusive evidence regarding how stem cells uniquely respond to nicotine. It is also recommended to present the staining of α7-nAChR from different intestinal regions. This will provide insights into the primary target sites of nicotine in the gut tract. Additionally, it is recommended that the authors consider rephrasing the conclusion in this section (lines 123-124). The current statement implies that nicotine does not affect Paneth cells, which may be inaccurate based on the suggestion in line 275 that nicotine might influence Paneth cells through α2β4-nAChR. Providing a more nuanced conclusion would better reflect the complexity of nicotine's potential impact on Paneth cells.

It was difficult to obtain nAchRs antibodies usable in immunostaining. Hence, we instead performed qPCR of nAchRs in ISCs and Paneth cells from isolated whole small intestine (new Figure3C), although we cannot know the difference of the nAchRs expression in different intestinal regions by this method. Although the comparatively high expression was observed in α7-nAChR and α8nAChR in both ISCs and Paneth cells, the significant difference between ISCs and Paneth cells were not observed (Figure3C). 

Interestingly, nicotine up-regulated only the expression of α7-nAChR in ISCs, suggesting the specifical response of α7-nAChR to nicotine (Figures 3C and D). We paraphrased the conclusion of the paragraph according to reviewer’s suggestion.

As shown in the same result section, the effect of nicotine on ISC organoid formation appears to be independent of CHIR99021, a Wnt activator. Despite this, the authors suggest a potential involvement of Wnt/β-catenin activation downstream of nicotine in Figure 4F. In the Lgr5CreER Apcfl/fl mouse model, it is known that APC loss results in a constitutive stabilization of β-catenin, thus the hyperproliferation of ISCs by nicotine treatment in this mouse model is likely beyond Wnt activation. Therefore, it is recommended that the authors reconsider the inclusion of Wnt/β-catenin as a crucial signaling pathway downstream of nicotine, given the experimental evidence provided in this study.

We appreciate for this important suggestion. Certainly, Wnt/β-catenin was activated in Nicotine treated ISCs. However, as reviewer points out, the hyperproliferation of ISCs by nicotine treatment is likely beyond Wnt activation.  According to the reviewer’s suggestion, we removed Wnt/β-catenin as a crucial signaling pathway downstream of nicotine (Figure 5G).

In Figure 4, the authors investigate ISC organoid formation with a panPKC inhibitor, revealing that PKC inhibition blocks nicotine-induced ISC expansion. It's noteworthy that PKC inhibitors have historically been used successfully to isolate and maintain stem cells by promoting self-renewal. Therefore, it is surprising to observe no effect or reversal effect on ISCs in this context. A previous study demonstrated that the loss of PKCζ leads to increased ISC activity both in vivo and in vitro (DOI: 10.1016/j.celrep.2015.01.007). Additionally, to strengthen this aspect of the study, it would be beneficial for the authors to present more evidence, possibly using different PKC inhibitors, to reproduce the observed results with Gö 6983. This could help address potential concerns or discrepancies and contribute to a more comprehensive understanding of the role of PKC in nicotine-induced ISC expansion.

Gö 6983 is a pan-PKC inhibitor against for PKCα, PKCβ, PKCγ, PKCδ and PKCζ with IC50 of 7 nM, 7 nM, 6 nM, 10 nM and 60 nM, respectively. Since we used Gö 6983 at the concentration of 10nM in our experiment, we consider PKCζ may not be possible target of nicotine. Additionally, we treated using 5nM Sotrastaurin, another pan-PKC inhibitor, which is supposed not to affect PKCζ. The observed result with Gö 6983 was reproduced by Sotrastaurin (Supplemental Figure 3E).

An additional avenue that could enhance the clinical relevance of the study is the exploration of human datasets. Specifically, leveraging scRNA-seq datasets of the human intestinal epithelium (DOI: 10.1038/s41586-021-03852-1) could provide valuable insights. Analyzing the expression patterns of nAChRs across diverse regions and cell types in the human intestine may offer a potential clinical implication.

We analyzed distribution pattern nAChRs of by scRNA-seq datasets of the human intestinal epithelium (DOI: 10.1038/s41586-021-03852-1). In consistent with mouse data (Figure3C), the expression of human α7-nAChR is higher than that of other nAChRs. The difference of the expression between ISCs and Paneth cells is not clear as in that of mouse (Supplemental Figure4A and B). From mouse and human data, we speculate the induction of specific nAChR by nicotine is essence of ISC response to nicotine, rather than the distribution of nAChRs.

Reviewer #2 (Recommendations For The Authors):

The manuscript could benefit from addressing a few minor points to enhance its quality before publication:

(1) Ensure all images are presented in higher resolution to improve visual clarity.

We replaced all images by those with higher resolution.

(2) Quantify Western blot results accurately for rigor and precision in data representation.

We quantified all blots.

(3) Include error bars in control groups where missing, particularly in Figures 3C and 4D, to enhance data interpretation.

We included error bars in control groups in new Figure 3C and 4D.

(4) The layout of Figure S3B, S4A and S4B should be corrected.

We corrected the layout of those Figures.

Author response:

The following is the authors’ response to the original reviews.

Reviewer #1 (Public Review): 

Petty and Bruno investigate how response characteristics in the higher-order thalamic nuclei POm (typically somatosensory) and LP (typically visual) change when a stimulus (whisker air puff or visual drifting grating) of one or the other modality is conditioned to a reward. Using a two-step training procedure, they developed an elegant paradigm, where the distractor stimulus is completely uninformative about the reward, which is reflected in the licking behavior of trained mice. While the animals seem to take on to the tactile stimulus more readily, they can also associate the reward with the visual stimulus, ignoring tactile stimuli. In trained mice, the authors recorded single-unit responses in both POm and LP while presenting the same stimuli. The authors first focused on POm recordings, finding that in animals with tactile conditioning POm units specifically responded to the air puff stimulus but not the visual grating. Unexpectedly, in visually conditioned animals, POm units also responded to the visual grating, suggesting that the responses are not modality-specific but more related to behavioral relevance. These effects seem not be homogeneously distributed across POm, whereas lateral units maintain tactile specificity and medial units respond more flexibly. The authors further ask if the unexpected cross-modal responses might result from behavioral activity signatures. By regressing behavior-coupled activity out of the responses, they show that late activity indeed can be related to whisking, licking, and pupil size measures. However, cross-modal short latency responses are not clearly related to animal behavior. Finally, LP neurons also seem to change their modality-specificity dependent on conditioning, whereas tactile responses are attenuated in LP if the animal is conditioned to visual stimuli.

The authors make a compelling case that POm neurons are less modality-specific than typically assumed. The training paradigm, employed methods, and analyses are mostly to the point, well supporting the conclusions. The findings importantly widen our understanding of higher-order thalamus processing features with the flexibility to encode multiple modalities and behavioral relevance. The results raise many important questions on the brain-wide representation of conditioned stimuli. E.g. how specific are the responses to the conditioned stimuli? Are thalamic cross-modal neurons recruited for the specific conditioned stimulus or do their responses reflect a more global shift of attention from one modality to another? 

To elaborate on higher-order thalamic activity in relationship to conditioned behavior, a trialby-trial analysis would be very useful. Is neuronal activity predictive of licking and at which relative timing? 

To elaborate on the relationship between neuronal activity and licking, we have created a new supplementary figure (Figure S1), where we present the lick latency of each mouse on the day of recording. We also perform more in-depth analysis of neural activity that occurs before lick onset, which is presented in a new main figure (new Figure 4). 

Furthermore, I wonder why the (in my mind) major and from the data obvious take-away, "POm neurons respond more strongly to visual stimuli if visually conditioned", is not directly tested in the summary statistics in Figure 3h.

We have added a summary statistic to Figure 3h and to the Results section (lines 156-157) comparing the drifting grating responses in visually and tactilely conditioned mice.  

The remaining early visual responses in POm in visually conditioned mice after removing behavior-linked activity are very convincing (Figure 5d). It would help, however, to see a representation of this on a single-neuron basis side-by-side. Are individual neurons just coupled to behavior while others are independent, or is behaviorally coupled activity a homogeneous effect on all neurons on top of sensory activity?

In lieu of a new figure, we have performed a new analysis of individual neurons to classify them as “stimulus tuned” and/or “movement tuned.” We find that nearly all POm cells encode movement and arousal regardless of whether they also respond to stimuli. This is presented in the Results under the heading “POm correlates with arousal and movement regardless of conditioning” (Lines 219-231).

The conclusions on flexible response characteristics in LP in general are less strongly supported than those in POm. First, the differentiation between POm and LP relies heavily on the histological alignment of labeled probe depth and recording channel, possibly allowing for wrong assignment. 

We appreciate the importance in differentiating between POm, LP, and surrounding regions to accurately assign a putative cell to a brain region. The method we employed (aligning an electrode track to a common reference atlas) is widely used in rodent neuroscience, especially in regions like POm and LP which are difficult to differentiate molecularly (for example, see Sibille, Nature Communications, 2022; and Schröder, Neuron, 2020). 

Furthermore, it seems surprising, but is not discussed, that putative LP neurons have such strong responses to the air puff stimuli, in both conditioning cases. In tactile conditioning, LP air puff responses seem to be even faster and stronger than POm. In visual conditioning, drifting grating responses paradoxically seem to be later than in tactile conditioning (Fig S2e). These differences in response changes between POm and LP should be discussed in more detail and statements of "similar phenomena" in POm and LP (abstract) should be qualified.  

We have further developed our analysis and discussion of LP activity. Our analysis of LP stimulus response latencies are now presented in greater detail in Figure S3, and we have expanded the results section accordingly (lines 266-275). We have also expanded the discussion section to both address these new analyses and speculate on what might drive these surprising “tactile responses” in LP.

Reviewer #2 (Public Review): 

Summary  

This manuscript by Petty and Bruno delves into the still poorly understood role of higherorder thalamic nuclei in the encoding of sensory information by examining the activity in the Pom and LP cells in mice performing an associative learning task. They developed an elegant paradigm in which they conditioned head-fixed mice to attend to a stimulus of one sensory modality (visual or tactile) and ignore a second stimulus of the other modality. They recorded simultaneously from POm and LP, using 64-channel electrode arrays, to reveal the contextdependency of the firing activity of cells in higher-order thalamic nuclei. They concluded that behavioral training reshapes activity in these secondary thalamic nuclei. I have no major concerns with the manuscript's conclusions, but some important methodological details are lacking and I feel the manuscript could be improved with the following revisions.

Strengths 

The authors developed an original and elegant paradigm in which they conditioned headfixed mice to attend to a stimulus of one sensory modality, either visual or tactile, and ignore a second stimulus of the other modality. As a tactile stimulus, they applied gentle air puffs on the distal part of the vibrissae, ensuring that the stimulus was innocuous and therefore none aversive which is crucial in their study. 

It is commonly viewed that the first-order thalamus performs filtering and re-encoding of the sensory flow; in contrast, the computations taking place in high-order nuclei are poorly understood. They may contribute to cognitive functions. By integrating top-down control, high-order nuclei may participate in generating updated models of the environment based on sensory activity; how this can take place is a key question that Petty and Bruno addressed in the present study.

Weaknesses  

(1) Overall, methods, results, and discussion, involving sensory responses, especially for the Pom, are confusing. I have the feeling that throughout the manuscript, the authors are dealing with the sensory and non-sensory aspects of the modulation of the firing activity in the Pom and LP, without a clear definition of what they examined. Making subsections in the results, or a better naming of what is analyzed could convey the authors' message in a clearer way, e.g., baseline, stim-on, reward.  

We thank Reviewer 2 for this suggestion. We have adjusted the language throughout the paper to more clearly state which portions of a given trial we analyzed. We now consistently refer to “baseline,” “stimulus onset,” and “stimulus offset” periods. 

In line #502 in Methods, the authors defined "Sensory Responses. We examined each cell's putative sensory response by comparing its firing rate during a "stimulus period" to its baseline firing rate. We first excluded overlapping stimuli, defined as any stimulus occurring within 6 seconds of a stimulus of a different type. We then counted the number of spikes that occurred within 1 second prior to the onset of each stimulus (baseline period) and within one second of the stimulus onset (stimulus period). The period within +/-50ms of the stimulus was considered ambiguous and excluded from analysis." 

Considering that the responses to whisker deflection, while weak and delayed, were shown to occur, when present, before 50 ms in the Pom (Diamond et al., 1992), it is not clear what the authors mean and consider as "Sensory Responses"? 

We have addressed this important concern in three ways. First, we have reanalyzed our data to include the 50ms pre- and post-stimulus time windows that were previously excluded. This did not qualitatively change our results, but updated statistical measurements are reflected in the Results and the legends of figures 3 and 7. Second, we have created a new figure (new Figure 4) which provides a more detailed analysis of early POm stimulus responses at a finer time scale. Third, we have amended the language throughout the paper to refer to “stimulus responses” rather than “sensory responses” to reflect how we cannot disambiguate between bottom-up sensory input and top-down input into POm and LP with our experimental setup. We refer only to “putative sensory responses” when discussing lowlatency (<100ms) stimulus responses.

Precise wording may help to clarify the message. For instance, line #134: "Of cells from tactilely conditioned mice, 175 (50.4%) significantly responded to the air puff, as defined by having a firing rate significantly different from baseline within one second from air puff onset (Figure 3d, bottom)", could be written "significantly responded to the air puff" should be written "significantly increased (or modified if some decreased) their firing rate within one second after the air puff onset (baseline: ...)". This will avoid any confusion with the sensory responses per se.

We have made this specific change suggested by the reviewer (lines 145-146) and made similar adjustments to the language throughout the manuscript to better communicate our analysis methods. 

(2) To extend the previous concern, the latency of the modulation of the firing rate of the Pom cells for each modality and each conditioning may be an issue. This latency, given in Figure S2, is rather long, i.e. particularly late latencies for the whisker system, which is completely in favor of non-sensory "responses" per se and the authors' hypothesis that sensory-, arousal-, and movement-evoked activity in Pom are shaped by associative learning. Latency is a key point in this study. 

Therefore, 

- latencies should be given in the main text, and Figure S2 could be considered for a main figure, at least panels c, d, and e, could be part of Figure 3. 

- the Figure S2b points out rather short latency responses to the air puff, at least in some cells, in addition to late ones. The manuscript would highly benefit from an analysis of both early and late latency components of the "responses" to air puffs and drafting grating in both conditions. This analysis may definitely help to clarify the authors' message. Since the authors performed unit recordings, these data are accessible.

- it would be highly instructive to examine the latency of the modulation of Pom cells firing rate in parallel with the onset of each behavior, i.e. modification of pupil radius, whisking amplitude, lick rate (Figures 1e, g and 3a, b). The Figure 1 does not provide the latency of the licks in conditioned mice.

- the authors mention in the discussion low-latency responses, e.g., line #299: "In both tactilely and visually conditioned mice, movement could not explain the increased firing rate at air puff onset. These low-latency responses across conditioning groups is likely due in part to "true" sensory responses driven by S1 and SpVi."; line #306: "Like POm, LP displayed varied stimulus-evoked activity that was heavily dependent on conditioning. LP responded to the air puff robustly and with low latency, despite lacking direct somatosensory inputs."  But which low-latency responses do the authors refer to? Again, this points out that a robust analysis of these latencies is missing in the manuscript but would be helpful to conclude.

We have moved our analysis of stimulus response latency in POm to new Figure 4 in the main text and have expanded both the Results and Discussion sections accordingly. We have also analyzed the lick latency on the day of recording, included in a new supplemental Figure S1. 

(3) Anatomical locations of recordings in the dorsal part of the thalamus. Line #122 "Our recordings covered most of the volume of POm but were clustered primarily in the anterior and medial portions of LP (Figure 2d-f). Cells that were within 50 µm of a region border were excluded from analysis." 

How did the authors distinguish the anterior boundary of the LP with the LD nucleus just more anterior to the LP, another higher-order nucleus, where whisker-responsive cells have been isolated (Bezdudnaya and Keller, 2008)? 

Cells within 50µm of any region boundary were excluded, including those at the border of LP and LD. We also reviewed our histology images by eye and believe that our recordings were all made posterior of LD. 

(4) The mention in the Methods about the approval by an ethics committee is missing.  All the surgery (line #381), i.e., for the implant, the craniotomy, as well as the perfusion, are performed under isoflurane. But isoflurane induces narcosis only and not proper anesthesia. The mention of the use of analgesia is missing.