Final ruling on Q 1

Similarly, publication of spousal, family, close relatives, or other persons known to the offical is WAY overbroad and goes beyond "what is strictly necessary."

This could easily be referred to as shared interests of the spouse (generically) or relatives (generically) -- not their specific name.

First question

This is a hilarious image - all based on the family tree. Mr Elliot is part of the family but Sir Walter is not his father, he is the head of the family estate

https://www.gutenberg.org/files/33295/33295-h/33295-h.htm#ar1

THE ENCYCLOPÆDIA BRITANNICA<br /> A DICTIONARY OF ARTS, SCIENCES, LITERATURE AND GENERAL INFORMATION<br /> ELEVENTH EDITION<br /> VOLUME V SLICE VI<br /> Celtes, Konrad to Ceramics

Rudolphus Agricola was one of Konrad Celtes' teachers.

A very euphemistic way to say that she bowed her head a lot. And nose and chin echo her appearance description.

Immediately there is a sense of stillness. Similar to the descriptions in "Araby", the usage of the senses are employed to suggest something about the time. Here it smells dusty and she is still, which refers to the themes in "Eveline" of dwelling on the past.

Even when thinking of a joyous holiday like Christmas, the "grey face," Father Flynn presumably, "follows" the young boy. The boy mentions the word paralysis at the beginning of the short story -- a direct reflection of his current state. The boy is paralyzed in bed with ruminating thoughts about Father Flynn's current state.

Es ist interessant, dass die ökologische Produktion nicht weniger effizient ist als die konventionelle. Es hat offenbar – wie bei der Energie, folgt man Malm – andere Gründe, wenn sie abgelehnt wird.

another example of how math does not describe nature beautifully...

good enough for my purposes (local git scripts/aliases)!

⟫ cat .git/refs/remotes/origin/HEAD ref: refs/remotes/origin/main

Minigames of life, to improve ourselfs with small improvements (energizers, boosters) that unlock points of life.

Reviewer #3 (Public Review):

The number of identified anti-phage defense systems is increasing. However, the general understanding of how phages can overcome such bacterial defense mechanisms is a black box. Srikant et al. apply an experimental evolution approach to identify mechanisms of how phages can overcome anti-phage defense systems. As a model system, the bacteriophage T4 and its host Escherichia coli are applied to understand genome dynamics resulting in the deactivation of phage-defensive toxin-antitoxin systems.

Strengths:<br /> The application of a coevolutionary experimental design resulted in the discovery of a gene-operon: dmd-tifA. Using immunoprecipitation experiments, the interaction of TifA with ToxN was demonstrated. This interaction results in the inactivation of ToxN, which enables the phage to overcome the anti-phage defense system ToxIN.<br /> The characterization of the genomes of T4 phages that overcome the phage-defensive ToxIN revealed that the T4 genome can undergo large genomic changes. As a driving force to manipulate the T4 phage genome, the authors identified recombination events between short homologous sequences that flank the dmd-tifA operon.<br /> The discovery of TifA is well supported by data. The authors prepared several mutant strains to start the functional characterization of TifA and can show that TifA is present in several T4-like phages.

In addition, they describe T4 head protein IPIII as another antagonist of a so far unknown defense system.

In summary, the application of a coevolutionary approach to discover anti-phage defense systems is a promising technique that might be helpful to study a variety of virus-host interactions and to predict phage evolution techniques.

Weaknesses:<br /> The authors apply Illumina sequencing to characterize genome dynamics. This NGS method has the advantage of identifying point mutations in the genome. However, the identification of repetitive elements, especially their absolute quantification in the T4 genome, cannot be achieved using this method. Thus, the authors should combine Illumina Sequencing with a long-read sequencing technology to characterize the genome of T4 in more detail.

To characterize the influence of TifA during infection, T4 phage mutants are generated using a CRISPR-Cas-based technique. The preparation of these phages is unclearly described in the methods section. The authors should describe in detail whether a b-gt deficient strain was applied to prepare the mutants. Information about the used primers and cloning schemes of the Cas9 plasmid would allow the community to repeat such experiments successfully.

The discovery of TifA would benefit from additional data, e.g. structure-based predictions, that describe the protein-protein interaction TifA/ToxN in more detail.

Several publications have described that antitoxins can arise rapidly during a phage attack. The authors should address that this concept has been described before as well by citing appropriate publications.

The authors propose that accessory genomes of viruses reflect the integrated evolutionary history of the hosts they infected. However, the experimental data do not support such a claim.

Author Response

Reviewer 1

Sadeh and Clopath analyze two mouse datasets from the Allen Brain Atlas and show that sensory representations can have apparent representational drift that is entirely due to behavioral modulation. The analysis serves as a caution against over-interpreting shifts in the neural code. The analysis of data is coupled with careful modeling work that shows that the behavioral state reliably shifts sensory representations independently of stimulus modulation (rather than acting as a gain factor), and further show that it is reproducibly shifted when the behavioral state is adequately controlled for. The methods presented point towards a more careful consideration and measurement of behavioral states during sensory recordings, and a re-analysis of previous findings. The findings held up for both standard drifting grating stimuli as well as natural movies.

The fact that neurons may have different tuning depending on the behavioral state of the animal raises obvious questions about readout. The authors show that neurons with strong behavioral shifts should simply be ignored and that this can be achieved if the downstream decoder weights inputs with more stimulus information. While questions remain about why behavior shifts representations and how that could be more effectively utilized by downstream circuits, the results presented clearly show that sensory representations might not always be simply drifting over time, and will spark some careful analysis of past and future experimental results.

Many thanks for a clear summary of the work and emphasizing the significance of the results.

Reviewer 2

Studies from recent years have shown that neuronal responses to the same stimuli or behavior can gradually change with time - a phenomenon known as representational drift. Other recent studies have shown that changes in behavior can also modulate neuronal responses to a given sensory stimulus. In this manuscript, Sadeh and Clopath analyzed publicly available data from the Allen Institute to examine the relationship between animal behavioral variability and changes in neuronal representations. The paper is timely and certainly has the potential to be of interest to neuroscientists working in different fields. However, there are currently several important issues with the analysis of the data and their interpretations that the authors should address. We believe that after these concerns are addressed, this study will be an important contribution to the field.

We really appreciate the time and the effort the reviewer(s) have taken to evaluate our results and analysis in detail. Their comments are very relevant and critical to the improvement of the manuscript. We explain below how we addressed their various comments and concerns

1. The manuscript raises a potential problem: while previous work suggested that the passage of time leads to gradual changes in neuronal responses, the causality structure is different: i.e., the passage of time leads to gradual changes in behavior, which in turn lead to gradual changes in neuronal responses. The authors conclude that "variable behavioral signal might be misinterpreted as representational drift". While this may be true, in its current form, the paper lacks critical analyses that would support such a claim. It is possible that both factors - time and behavior - have a unique contribution to changes in neuronal responses, or that only time elicits changes in neuronal responses (and behavior is just correlated with time). Thus, the authors should demonstrate that these changes cannot be explained solely by the passage of time and elucidate the unique contributions of behavior (and elapsed time) to changes in representations.

This is a very important point and we addressed it with new analyses, by dedicating a new figure (Figure 1–figure supplement 5) and a new part of the Results section to it. The results of our new analyses show that strong representational drift mainly exists in those animals/sessions with large behavioral changes between the two blocks, and that in animals/sessions with small behavioral changes, such drift is minimal, despite the passage of time (see our responses below to Major comments for further details).

1. There are also several issues with the analysis of the data and the presentation of the results. The most concerning of which is that the data shows a non-linear (and non-monotonic) relationship between behavioral changes and representational similarity. In many of the presented cases, the data points fall into two or more discrete clusters. This can lead to the false impression that there is a monotonic relationship between the two variables, even though there is no (or even opposite) relationship within each cluster. This is a crucial point since the clusters of data points most likely represent different blocks that were separated in time (or separation between within-block and acrossblock comparisons).

This is an important concern. To address this, we analyzed the source of the non-monotonic relationship / opposite trend in the data and demonstrated the results in a new figure (Figure 4–figure supplement 2). Our results show that the non-monotonic relationship does not compromise the result of our previous analysis. Furthermore, it suggests that the non-monotonic / opposite trend is emerging as a result of more complex interactions between different aspects of behavior. We have also shown, in separate analyses, that the passage of time is not the main contributing factor to representational drift, rather large behavioral changes are correlated with strong drifts between the two blocks of presentation (Figure 1—figure supplement 5, and Figure 3—figure supplement 2).

More generally, we did not intend to claim that the relationship with behavioral changes is linear or/and monotonic. We used linear analysis just to show the main trend of decrease in representational similarity with large behavioral changes. Any other analysis should assume some form of nonlinearity, but because the nonlinear relationships between behavior and activity were complex, it was not easy to assume such nonlinearity.

We in fact tried to use two other ways of analysis, nonlinear correlations and generalized linear models (GLM), but there were issues hindering a proper use of each analysis. Nonlinear correlations assume a specific type of nonlinearity, but the nature of nonlinearity underlying the data is not clear (in fact, it looks to be different in different example non-monotonic trends in the data). We could not, therefore, assume a nonlinearity that best fitted all the data; we believe the nature of this nonlinearity, or how behavior modulates neuronal activity in a nonlinear manner, is in itself an interesting and open question for future investigation, but beyond the scope of this study. GLM did not provide useful results either, as the relationship between behavioral changes and neural activity/representational similarity was state-dependent and transitioning between nonlinear states, therefore hindering the usage of linear methods.

We therefore opted for the simplest analysis which can show and quantify this dependence - emphasizing that further analyses are in fact needed to get to the bottom of the exact nonlinear relationship (for further details, see the responses below to Major comments).

1. The authors also suggest that using measures of coding stability such as 'population-vector correlations' may be problematic for quantifying representational drift because it could be influenced by changes in the neuronal activity rates, which may be unrelated to the stimulus. We agree that it is important to carefully dissociate between the effects of behavior on changes in neuronal activity that are stimulus-dependent or independent, but we feel that the criticism raised by the authors ignores the findings of multiple previous papers, which (1) did not purely attribute the observed changes to the sensory component, and (2) did dissociate between stimulus-dependent changes (in the cells' tuning) and off-context/stimulus-independent changes (in the cells' activity rates).

That’s a very valid point. As population vector correlations are used quite often in (experimental and theoretical) works on representational drift, we wanted to highlight the pitfalls of such a metric in dissociating between sensory-evoked and sensory-independent components. However, as the reviewers have mentioned, these two aspects have been separated and addressed independently in some of the past literature in the field. For instance, as we discussed in the Discussion, Deitch et al. (Current Biology, 2021) have calculated this for different metrics, including tuning curve correlations, which can potentially alleviate this problem:

“A recent analysis of similar datasets from the Allen Brain Observatory reported similar levels of representational drift within a day and over several days5. The study showed that tuning curve correlations between different repeats of the natural movies were much lower than population vector and ensemble rate correlations5; it would be interesting to see if, and to which extent, similarity of population vectors due to behavioural signal that we observed here may contribute to this difference.”

We tried to highlight these contributions better in the revised manuscript (see further on this below in our responses to Major comments).

1. Another important issue relates to the interchangeable use of the terms 'representational drift' and 'representational similarity'. Representational similarity is a measure to identify changes in representations, and drift is one such change. This may confuse the reader and lead to the misconception that all changes in neuronal responses are representational drift.

We thank the reviewer(s) for raising this point. We have clarified our use of the terms representational similarity and representational drift in the revised manuscript. Specifically, we have quantified representational drift index between the two blocks according to a previously used metric (RDI; Marks & Goard, 2021) in our new analysis (Figure 1–figure supplement 5).

For the main part of the paper, however, we have decided to base our analysis on representational similarity (RS), and to evaluate the drop of RS with changes in behavior. Our reasoning for this is twofold. First, any measure of representational drift should ultimately be a function of the representational similarity. The measure we used above, for instance, is calculated as RD = (RS_ws - RS_bs)/(RS_ws + RS_bs) (Marks and Goddard, 2021), with RS_ws and RS_bs referring to the average representational similarity within a session or between different sessions. However, RS contains more information, especially with regard to fine-tuned changes - the above metric, for instance, averages all the changes within each block of presentation. By focusing on the basic function of representational similarity, we could capture both the gross changes between the blocks as well as more nuanced changes that can arise within them, especially with regard to behavioral changes. Another aspect that would have been lost by only using the usual metric of representational drift is the direction of change. In our analysis, we in fact found that the average RS increased within the second block of presentation, which might be contrary to the usual direction of drift. We found this unconventional change of RS interesting and informative too. We could highlight that, presenting the raw RS provided a better analysis strategy. Based on these reasons, we think representational similarity would be a better metric to base our analyses upon, although we have now calculated a conventional representational drift index for comparison too.

Reviewer 3

Although it is increasingly realized that cortical neural representations are inherently unstable, the meaning of such "drift" can be difficult or impossible to interpret without knowing how the representations are being read out and used by the nervous system (i.e. how it contributes to what the experimental animal is actually doing now or in the future). Previous studies of representational drift have either ignored or explicitly rejected the contribution of what the animal is doing, mostly due to a lack of high-dimensional behavioural data. Here the authors use perhaps the most extensive opensource and rigorous neural data available to take a more detailed look at how behaviour affects cortical neural representations as they change over repeated presentations of the same visual stimuli.

The authors apply a variety of analyses to the same two datasets, all of which convincingly point to behavioural measures having a large impact on changing neural representations. They also pit models against each other to address how behavioural and stimulus signals combine to influence representations, whether independently or through behaviour influencing the gain of stimuli. One analysis uses subsets of neurons to decode the stimulus, and the independent model correctly predicts the subset to use for better decoding. However, one caveat may be that the nervous system does not need to decode the stimulus from the cortex independently of behaviour; if necessary, this could be done elsewhere in the nervous system with a parallel stream of visual information.

Overall the authors' claims are well-supported and this study should lead to a re-assessment of the concept of "representational drift". Nonetheless, a weakness of all analyses presented here is that they are all based on data in head-fixed mice that were passively viewing visual stimuli, such that it is unclear what relevance the behaviour has. Furthermore, the behavioural measurements available in the opensource dataset (pupil movements and running speed) are still a very low dimensional representation of what the mice were actually doing (e.g. detailed kinematics of all body movements and autonomic outputs). Thus, although the authors here as well as other large-scale neural recording studies in the past decade or so make it clear that relatively basic measures of behaviour can dramatically affect cortical representations of the outside world, the extent to which any cortical coding might be considered purely sensory remains an important question. Moreover, it is possible that lowerdimensional signals are overly represented in visual areas, and that in other areas of the cortex (e.g. somatosensory for proprioception), the line between behaviour parameters and sensory processing is blurred.

Many thanks for the clear and insightful summary of the results, significance and caveats of our analysis. We totally agree with this critical evaluation - and suggestions for future work.

Her hair is symbolized as a desert. Desert meaning sun golden hair large as the deserts in Africa. Majority of North Africa consist of deserts much like the hair on a woman's head.

This became apparent very early in my teaching career. I had a student who reminded me of another student in one of my other classes whose names were spelled identically, but pronounced differently. Throughout the first semester, I repeatedly pronounced the student's name incorrectly by using the other's student's pronunciation (Does this make sense so far :)? By mid November, she came to me upset and told me she didn't appreciate that I still confused her name with another student's name. After that, I made it a point to say her name correctly by phonetically writing it on my attendance sheets, seating chart, and in my head (over and over again while looking at her picture). Have I made the same mistakes over the years? Yes. but I employ the above strategies before it goes to hurt feelings.

• reality : pink plane
• insanity : blue plane
• takes one to know one
  • learning a new idea could require almost as much creativity as inventing it in the first place because you have to invent it inside your head

and seriously you don't mention visual code???

"A leader is one who is looked up to, whose personal judgement is trusted, and who can inspire and warm the heart of those he/she heads or leads, gaining their trust and confidence and explaining what is needed in a language which can be understood." Agenta (2005) cited in Uwaifo (2012)

As I'm reading this I'm thinking about my own interaction with everyone in my place of employment. I'm a scheduling manager at a fairly small manufacturing company. We make office products and I'm over 50+ people. Am I someone who is looked up to? I hope so, and try to be. Can my personal judgement be trusted? Can in inspire and warm the heart of those I head or lead? Am I able to gain their trust? Gain their confidence? Am I capable to explain what is needed in a language which can be understood. I try to speak Spanish every day, but realize I need to make it a point to learn more technical Spanish. "Can we change the flow of these pallets in a different direction to allow a "First In, First Out" route?" These are the hard types of things to translate, and fulfill understanding in all parties. (*noted "to do" next week)

Really good questions to ask yourself, whether you own your own business, are just starting in the work place, or anywhere in between.

--worktree: restablece el working (desde el Index) --staged: restablece el Index (desde el HEAD)

Nota: --worktree es el default.

Author Response

Reviewer 3

The number of identified anti-phage defense systems is increasing. However, the general understanding of how phages can overcome such bacterial defense mechanisms is a black box. Srikant et al. apply an experimental evolution approach to identify mechanisms of how phages can overcome anti-phage defense systems. As a model system, the bacteriophage T4 and its host Escherichia coli are applied to understand genome dynamics resulting in the deactivation of phage-defensive toxin-antitoxin systems.

Strengths: The application of a coevolutionary experimental design resulted in the discovery of a geneoperon: dmd-tifA. Using immunoprecipitation experiments, the interaction of TifA with ToxN was demonstrated. This interaction results in the inactivation of ToxN, which enables the phage to overcome the anti-phage defense system ToxIN. The characterization of the genomes of T4 phages that overcome the phage-defensive ToxIN revealed that the T4 genome can undergo large genomic changes. As a driving force to manipulate the T4 phage genome, the authors identified recombination events between short homologous sequences that flank the dmd-tifA operon. The discovery of TifA is well supported by data. The authors prepared several mutant strains to start the functional characterization of TifA and can show that TifA is present in several T4-like phages.

In addition, they describe T4 head protein IPIII as another antagonist of a so far unknown defense system.

In summary, the application of a coevolutionary approach to discover anti-phage defense systems is a promising technique that might be helpful to study a variety of virus-host interactions and to predict phage evolution techniques.

Weaknesses: The authors apply Illumina sequencing to characterize genome dynamics. This NGS method has the advantage of identifying point mutations in the genome. However, the identification of repetitive elements, especially their absolute quantification in the T4 genome, cannot be achieved using this method. Thus, the authors should combine Illumina Sequencing with a longread sequencing technology to characterize the genome of T4 in more detail.

We think the combination of Illumina-based sequencing and PCR analyses presented are more than sufficient to arrive at the conclusions drawn about the repeats that emerge in our evolved T4 clones.

To characterize the influence of TifA during infection, T4 phage mutants are generated using a CRISPR-Cas-based technique. The preparation of these phages is unclearly described in the methods section. The authors should describe in detail whether a b-gt deficient strain was applied to prepare the mutants. Information about the used primers and cloning schemes of the Cas9 plasmid would allow the community to repeat such experiments successfully.

We have added details to the Methods section to clarify and expand on our mutagenesis approach.

The discovery of TifA would benefit from additional data, e.g. structure-based predictions, that describe the protein-protein interaction TifA/ToxN in more detail.

We were unable to predict the ToxN-TifA interaction interface using AlphaFold, and we are currently conducting follow-up work to characterize how TifA neutralizes ToxN.

Several publications have described that antitoxins can arise rapidly during a phage attack. The authors should address that this concept has been described before as well by citing appropriate publications.

We believe that we have already addressed this point sufficiently in the Introduction of the manuscript, in which we discuss (1) the emergence of phage-encoded pseudo-toxI repeats to overcome P. atrosepticum toxIN and (2) the presence of the naturally-occurring antitoxins Dmd and AdfA in T4 and T-even phages, respectively. We also discuss the similarities between TifA, Dmd, and AdfA in the discussion of the manuscript. To our knowledge, these are the only known examples of antitoxins arising during phage attack outside of TifA, but we are happy to include additional citations of which the reviewers are aware.

The authors propose that accessory genomes of viruses reflect the integrated evolutionary history of the hosts they infected. However, the experimental data do not support such a claim.

We disagree with the reviewer’s comment, as our evolution experiment demonstrates the plasticity of the T4 genome during adaptation to different hosts, as well as showing that the T4 accessory genome includes genes necessary for infection of some, but not all hosts. The proposal also comes as the last sentence of the Abstract and is framed not as a conclusion, but as a proposal based on the work done here, with the clear intention of providing a sense of how future work may build off our work.

Author Response

Reviewer 1

They adopted a comprehensive experimental and analytic approach to understand molecular and cellular mechanisms underlying tissue-specific responses against 3-CePs. They used two cell lines - BxPC-3 and HCT-15 - as example models for responsive and non-responsive cell lines, respectively. Although mutation rates didn’t differ by the drug treatment, they observed changes in cell cycle and expression of genes involved in DNA damage, repair and so on. Furthermore, they combined RNA-seq and ATAC-seq data and applied two approaches, pairwise and crosswise, to identify a number of gene groups that are altered in each cell line upon the drug treatment. Finally, they calculated enrichment of up/down genes in different cell lines, tumor types and samples to estimate potential responsitivity against the drug. This study is unique in in-depth analysis of RNA-seq and ATAC-seq data in identifying genetic signature underlying drug treatment. This study has the potential to be applied to different drugs and cell lines.

We thank the reviewer for the precise and kind summary of our work.

However, several major concerns need to be resolved. First of all, the biological and clinical performance of 3-CePs is not clearly described. They referenced several papers but they seem to have focused on the chemical properties of the drug. Without proven activity of 3-CePs against cancers in vitro and in vivo, the rationale of the study would be compromised.

We apologize for not being clear enough when introducing previous findings on the differential sensitivity of HCT-15 and BxPC-3 cancer cell lines to 3-CePs. In the revised manuscript, we now cite references on the preferential activity of these agents against the pancreatic cancer cell line in 2D and 3D in vitro cancer models (see lines 71-74, 128-129). These compounds have been selected to exemplify the use of the pipeline in drug discovery and early-stage of drug development: indeed, only cellular data are available for these molecules, which have not yet been characterized in vivo. The pipeline itself offered a final perspective on directions to take for their further development, i.e. most sensitive tumor types to target (PAAD, KIRC).

Their RNA-seq analysis was focused on discovering differentially expressed genes between cell lines, time points, etc. Interestingly, they found that DNA damage and repair signal was specifically increased in HCT-15. But is this approach capable of finding signals that are constitutively expressed in different cell lines? In other words, what if the differential responsiveness to 3-CePs was already there even before the drug was introduced?

We thank the reviewer for pointing out such key concept. The premise for the developed approach is that factors determining the overall cellular sensitivity to a treatment must be determined by intrinsic characteristics of the cell line. For this reason, we built the sensitivity signature on basal transcriptome profiles, where we prioritized a subset of genes based on perturbational evidence (perturbation-informed basal signature).

Beyond signature genes, we show in figure R1 (see above) the results of a GSEA analysis on the whole overlap (300 genes) between DE genes from the baseline comparison (BxPC-3 ctrl vs HCT-15 ctrl) and those from the 6 h M treatment comparison, in the sensitive cell line (BxPC-3 M 6 h vs BxPC-3 ctrl). Pathways like ribosome biogenesis, ROS metabolism, UPR also arise, attesting that genes activated in response to the treatment also have a constitutively different expression in unperturbed cells.

Are there any overlapping signals between pairwise vs crosswise approaches?

We thank the reviewer for this question. To make it easier for the reader to compare the output from the two types of integration and to intuitively grasp their functional overlap, we changed the visualization of the results from the pairwise approach (Figure 4 D).<br /> Indeed, some functional pathways both new or already emerging from previous analysis, arise from both integrations. This overlap has now been directly discussed from the functional point of view in the main text (from line 348 and in the following crosswise integration paragraph).

Genes used as input in both types of integration are DE or DAR-associated, so this means that many of the hits that we find having the same double regulation (pairwise) also appear in CoCena modules. Among them, only few hits show both 1) the same double regulation in a specified comparison (as suggested by crosswise) and also 2) end up having the similar pattern of regulation across all conditions (contributing to the same CoCena module, one of the strengths of the crosswise integration). Indeed, while the pairwise integration checks one single comparison per time, CoCena checks the pattern throughout conditions providing a more holistic view of the gene regulation (e.g one gene can have a different pattern across conditions at the transcriptional and chromatin level). This is due to the biological fact that RNA and chromatin regulation is not 1:1 (also, for instance, from a timing perspective).

The major added value of the two approaches consists in their intrinsically different output information. Within a specific comparison, the pairwise integration detects genes consistently activated at the transcriptome and chromatin level. At this information level gene set enrichment can simplify the coherent functional role of this set of genes; we now report this extra information in figure 4 to provide a more granular description of the pairwise integration. Instead, CoCena analyzes the pattern throughout conditions, and clusters together genes and peaks that behave similarly. Functional annotation of genes behaving similarly can put together promoters and/or transcripts that together may orchestrate a specific process (as highlighted by GSEA on each module).

Probably a similar question with the above: is this methodology applicable to other drugs in addition to 3-CePs?

To address this extremely important point, that we agree with the reviewer would be key to prove the versatility of our approach, we further applied the pipeline to the prediction of cancer cell lines’ sensitivity to cisplatin, a thoroughly reported broad-acting chemotherapeutic also acting as a DNA damaging agent. Results strongly supported the broad applicability of our approach, which was able to predict sensitivity to this reference drug with extremely high accuracy.

Reviewer 2

Carraro et al. describe a framework to understand MoA and susceptibility of drug candidates by integrating RNA-seq and ATAC-seq information. More specifically, by collecting drug responses from high-sensitive and low-sensitive cell lines, the authors identified a key set of pathways with co-expression analysis, and further predicted sensitivity of different cancer cell lines.

The authors provided a new bioinformatics pipeline to integrate multi-omics data (RNA-seq and ATAC-seq) in a drug response study. This approach increased detection power and identified additional key pathways that are associated with drug 3-CePs. This framework has the potential to be applied to the general drug discovery process.

We thank the reviewer for the precise summary of our study.

However, the current manuscript failed to describe the integration methodology in a clear and concise way. Without a full understanding of the methodology, it’s tough to evaluate the downstream results in an unbiased manner.

We apologize for not having included sufficient details in describing the difference between CoCena and the other two horizontal and vertical approaches. As already discussed in the response to Reviewer 1, we now included a more detailed description not only in the Methods section (from line 894) but also in the main text (lines 393-400).

In addition, the authors didn’t mention how much additional value this multi-omics approach provided compared to the single-omic data set, as multi-omics approaches are more expensive and labor-intensive.

We thank the reviewer for this valuable point. To better support the claim for multi-omics approaches, we have extended the Introduction (lines 96-98), as successful integration of information derived from multiple omic layers usually strengthens the determination of the major observed cellular responses. Here, this information helps dissecting and predicting how perturbations (here by drugs) can affect the overall cellular dynamics and mechanisms underlying a certain niveau of sensitivity. We agree with the reviewer that current costs are still prohibitive for large scale use of multi-layer omics in many settings, mainly when it comes to clinical use or drug development. However, significantly less expensive technologies (90% cost-reductions, lines 53-55) have recently been announced, which assures us that approaches as outlined here, will be applicable to many more clinical questions in the near future. Further, we show evidence that some cellular responses to the drug-induced perturbation was only revealed by applying multilayer analysis, but not by a single omics layer, e.g. TGF beta and EMT signaling (see lines 456-459).

Reviewer 3

Carraro et al utilize systems biology approaches to decode the mechanism of action of 3chloropiperidines (a novel class of cancer therapeutics) in cancer cell lines and build a drugsensitivity model from the data that they evaluate using samples from The Cancer Genome Atlas and cancer cell lines. The approach provides a framework for integrating transcriptomic and open-chromatin data to better understand the mechanism of action of drugs on cancer cell types. The author’s approach is of sound design, is clearly explained, and is bolstered by validation via holdout sets and analysis in new cell lines which lends the findings and approach credibility.

The major strength of this approach is the depth of information provided by performing RNA-seq and ATAC-seq on cells treated with 3-CePs at various time points, and the author’s utilization of this data to perform pairwise and crosswise analyses. Their approach identified gene modules that were indicative of why one cell type was more sensitive to a particular drug compared to another. The data was then used to build a sensitivity model which could be applied to samples from The Cancer Genome Atlas, and the authors evaluated their sensitivity predictions on a set of cancer cell lines which validated the predictions.

We thank the reviewer for the accurate recapitulation of our work.

The major drawback to this type of approach is that it relies on next-generation sequencing (somewhat costly) and requires intricate bioinformatics analyses. While I agree with the author’s perspective that this approach can be applied to additional classes of drugs and cancer samples, I disagree with their view that it is efficient and versatile. However, for research teams with the means to perform both transcriptomic and open-chromatin studies, I think this integrated approach has promise for evaluating novel classes of drugs, particularly in cancer cell lines that are easy to manipulate in vitro.

We thank the reviewer for this insightful comment. As with almost every technology, the early years are more difficult and at times adventurous. However, we have seen enormous improvements in robustness of the technology and significant cost reduction with more to come. Only recently sequencing technologies have been introduced into the market with a further 90% cost reduction (as stated in line 53-55). We are convinced that due to their increasing affordability and robustness, RNA-seq and ATAC-seq will be implemented routinely into clinical contexts. As a group working at the cross-section between drug discovery and bioinformatics, we hope that our current work, accompanied by a fair and detailed sharing of our scripts, will become a head start to run this type of analysis also by others in the field who are not (yet) so close to bioinformatics and computational biology.

While there are examples of similar frameworks being applied to drug development, this work will add to the body of literature utilizing an integrated systems biology approach for pairing drugs with specific tumor or cancer types and understanding their mechanism of action on an epigenetic level.

We thank the reviewer for this very positive statement and the support for our approach and her/his interest in the described pipeline.

assessments

Bitcoin P2P e-cash paper November 09, 2008, 03:09:49 AM

The proof-of-work chain is the solution to the synchronisation problem, and to knowing what the globally shared view is without having to trust anyone.

A transaction will quickly propagate throughout the network, so if two versions of the same transaction were reported at close to the same time, the one with the head start would have a big advantage in reaching many more nodes first. Nodes will only accept the first one they see, refusing the second one to arrive, so the earlier transaction would have many more nodes working on incorporating it into the next proof-of-work. In effect, each node votes for its viewpoint of which transaction it saw first by including it in its proof-of-work effort.

If the transactions did come at exactly the same time and there was an even split, it's a toss up based on which gets into a proof-of-work first, and that decides which is valid.

When a node finds a proof-of-work, the new block is propagated throughout the network and everyone adds it to the chain and starts working on the next block after it. Any nodes that had the other transaction will stop trying to include it in a block, since it's now invalid according to the accepted chain.

The proof-of-work chain is itself self-evident proof that it came from the globally shared view. Only the majority of the network together has enough CPU power to generate such a difficult chain of proof-of-work. Any user, upon receiving the proof-of-work chain, can see what the majority of the network has approved. Once a transaction is hashed into a link that's a few links back in the chain, it is firmly etched into the global history.

Little parasites can skillfully navigate their course of life, using odd tricks that just happen to work. For example, consider the following description of the tick from "A foray into the worlds of animals and humans" (Uexküll)

ANY COUNTRY DWELLER who traverses woods and bush with his dog has certainly become acquainted with a little animal who lies in wait on the branches of the bushes for his prey, be it human or animal, in order to dive onto his victim and suck himself full of its blood. In so doing, the one- to two-millimeter-large animal swells to the size of a pea (Figure 1). Although not dangerous, the tick is certainly an unwelcome guest to humans and other mammals. Its life cycle has been studied in such detail in recent work that we can create a virtually complete picture of it. Out of the egg crawls a not yet fully developed little animal, still missing one pair of legs as well as genital organs. Even in this state, it can already ambush cold-blooded animals such as lizards, for which it lies in wait on the tip of a blade of grass. After many moltings, it has acquired the organs it lacked and can now go on its quest for warm-blooded creatures. Once the female has copulated, she climbs with her full count of eight legs to the tip of a protruding branch of any shrub in order either to fall onto small mammals who run by underneath or to let herself be brushed off the branch by large ones. The eyeless creature finds the way to its lookout with the help of a general sensitivity to light in the skin. The blind and deaf bandit becomes aware of the approach of its prey through the sense of smell. The odor of butyric acid, which is given off by the skin glands of all mammals, gives the tick the signal to leave its watch post and leap off. If it then falls onto something warm—which its fine sense of temperature will tell it—then it has reached its prey, the warm-blooded animal, and needs only use its sense of touch to find a spot as free of hair as possible in order to bore past its own head into the skin tissue of the prey. Now, the tick pumps a stream of warm blood slowly into itself.

Experiments with artificial membranes and liquids other than blood have demonstrated that the tick has no sense of taste, for, after boring through the membrane, it takes in any liquid, so long as it has the right temperature. If, after sensing the butyric acid smell, the tick falls onto something cold, then it has missed its prey and must climb back up to its lookout post. The tick's hearty blood meal is also its last meal, for it now has nothing more to do than fall to the ground, lay its eggs, and die.

The family from "The Garden Party"

I haven't seen such kind of metaphor before. The comparison of the human body to food is quite interesting. On the one hand, it is very consistent with the situation at the time, and on the other hand, it also implies that the boy's head is very thin and his complexion is very dark. This metaphor is quite vivid.

A porter is a servant who is employed to carry luggage. The ridicule of the porter in their father's prized top-hat is emblematic of social class differences that seems to be a common thread so far in Mansfield's short stories.

There are several interesting similarities so far between this short story and “the Garden Party.” Both begin with a description and a dialogue before the characters’ names are introduced. Both include a dialogue between two people with opposite views—one sympathetic and the other indifferent. There is also a similar sentence to the highlighted one in Garden Party: “ ‘I say, you’re not crying, are you?’ asked her brother. Laura shook her head. She was.” This hints at the author’s style and a possible common theme between the stories.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

Referee #3

Evidence, reproducibility and clarity

Summary:

The study attempts to reconcile cryo-EM SPA structures of ODAs with in situ tomographic reconstructions.

Several key discrepancies between SPA structures and the native in situ structures (here) are highlighted in the study with a particular focus on the positions of various ODA motor head components (linker, tails etc.) during the powerstroke cycle.

The study also highlights largely concordant inter and intra-ODA connections between previous SPA structures and the tomographic reconstructions.

Major comments:

Overall, the key conclusions are convincing. No additional experiments are suggested. The manuscript is acceptable provided minor comments below are addressed.

Minor comments:

The text could be improved throughout for improved clarity. Overall, the figures are good, but some panels are over-annotated which is confusing. Simplification or cartoon illustrations could add clarity to the figures.

CROSS-CONSULTATION COMMENTS

The paper still represents a significant and sufficient advance. Correction of factual errors flagged up by other reviewers (use of correct references and citations, correct species for Lis1 models used etc.) is required and essential prior to acceptance. Addition of more details in the sample preparation methods section would also be useful. Depositing PDBs and maps is recommended.

Agree on the overall point of improving accessibility and readability of the text. Figures can be much improved to highlight the biological insights for the reader.

The point of contention between extra density corresponding to either Lis1 or LC5 is valid. Tempering the assertion and removing bias towards Lis1 in the text would resolve this issue. The authors are putting forth a speculative model which is valid; this model can be tested in future work.

Several minor comments highlighted by other reviewers are fair and should be addressed as best as possible.

Several major comments highlighted by other reviewers (specifically: use of structure prediction and modeling, filament distortion analysis etc.) are well beyond the scope of the present work and do not advance the specific and main conclusions of the current study.

Significance

The study presents structures of ODAs during their powerstroke cycles in situ in their native context and integrates previous structural models of ODAs to provide novel insights.

The identification of a Lis1 or LC5 like density adjacent to the alpha-HC and observation of a curved position of the beta-HC stalk in the native state adds further novelty to the study.

The study will be of interest to researchers like myself working on cilia motility and dynein motors.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

Referee #1

Evidence, reproducibility and clarity

Summary

Zimmermann et al. provide a comparison between recent atomic models of the ODA determined by single particle cryo-EM and their conformation within intact axonemes by cryo-ET subtomogram averaging. They observed slight changes in the position of the motors for the structures of Kubo et al. and Walton et al., but the structure of Rao et al. required more changes, indicating that within the axoneme, the conformation of the ODA is influenced by the MTD on which it is docked, and the neighboring MTD to which its motors bind. They then use the information from their newly fit models to interpret cryo-ET maps of axonemes in the presence of ATP, which activates the ODA and other axonemal dyneins. They observe two states of the ODA, and describe how the position of the motor, linker, MTBD and LC tower change during the powerstroke cycle. A revised model of the ODA and the ability to describe conformational changes at the subunit level provides an advance on previous work and will be of interest to the dynein and cilia fields. However, the comments below must be addressed prior to publication, and additional work is needed to make the paper accessible.

Major comments

1. Greater clarity is needed in the introduction to explain the differences between the recent atomic models of the ODA. This is essential to understanding the paper, including Fig. 3. Arguably, the top half of Fig. S2 provides a stronger case for the study than any of the current main figures.

2. In the manuscript, potential differences between Chlamydomonas and Tetrahymena ODAs are not considered but need to be explored. Comparison of Tetrahymena models within Chlamydomonas maps could result in misinterpretations.

3. Systematic quantification of the fit-to-map should be provided for the models before and after refitting (together with evidence - see the point below - that the model has not been inappropriately distorted to fit the map). This information could be inserted into an expanded Supplementary Table.

4. Because the revised pseudo-atomic model of the ODA is a chimera of PDBs from different organisms, it does not accurately represent the Chlamydomonas ODA. The modeling method also has the potential to introduce clashes between rigid-body fitted chains. Validation of the model is necessary, and alternative approaches to generate a more accurate model (e.g. AlphaFold and molecular dynamics flexible fitting) should be considered.

5. Additional evidence needs to be provided to demonstrate that the intermediate state observed in Figure 4 is robustly detected and does not simply represent the data that doesn't fall into the "good" classes. In Fig. S1, the map looks very noisy and requires denoising. Are there other changes observed in the IDAs that would support the existence of an intermediate state?

6. The speculation that the additional density bound to a-HC is Lis1 is not well-supported. Lis1 binds AAA4/5 (PDB: 5VH9), not AAA2/3. The fit of the Lis1 homolog into the cryo-ET density does not appear consistent with Lis1 binding the motor. The authors should consider other possibilities that could explain the additional density.

Minor comments

1. The results section "Post-PS structure and Fitting of the atomic models" is very dense. It should be split into subsections to help guide the reader through specific models or regions of the ODA.

2. ODA numbering should be made consistent with previous papers (i.e. ODA1-4 as in Bui et al., 2012)

3. The ODA-shulin model (PDB: 6ZYW) is inaccurately described as the state transported during IFT, but experimental confirmation of this hypothesis is lacking.

4. The term TTH for tail-to-head contacts is too similar to T/TH for the tether/tetherhead complex and should be changed. An abbreviation may not be necessary.

5. Please check to make sure that all figures and figure legends clearly specify which map/model/motor is being shown. This will make the figures easier to follow.

6. The structures in Fig. 3 are from Rao et al., not Walton et al.

7. Fig 5M-O is very difficult to interpret. Could the authors consider coloring by region, for one of the maps, or at least put the maps in a similar orientation to the ODA cores as in Fig 2?

8. The final processing step in panel Fig S1B is confusing. Additional information is needed to explain the supervised classification step and how the final particle set was derived.

9. Atomic resolution should not be used to describe structures determined to 4.3 Å resolution (e.g. EMD-11579).

10. Supervised classification is not a method of validation

11. Please check for grammatical and spelling errors throughout the manuscript.

Significance

While previous literature has interpreted ODA conformation in broad regions, this study goes farther by using recent atomic models to identify specific subunits that change conformations and interactions during the powerstroke. From my perspective as a structural biologist in the cilia field, I think this paper provides a conceptual advance to the study and interpretation of axonemes.

Provides some important context when thinking about the marketing that went into the marketing behind 50's household staples. As much as we may want to lean onto hating on the pop culture figure head that is the 19950's housewife, we also have to remember that the progression in home-economics technology and easy to prepare meals was considered a great progression and celebrated by many women. When the housewife archetype is not something you chose or want, it only makes sense to seek out things that would make it easier, more efficient. It was in the best interest of the women, who were predominantly made responsible for household duties, to popularize these products.

Indication that in July 1889, Beatrice Webb was developing her note taking methods and had a setback.

table 3

I feel as though this stereotype has recently been turned on its head. As more households have two working parents they split responsibilities much more evenly.

This is a struggle that I have dealt with in each district I have worked in. These are factors that are out of our hands as educators but we are expected to "correct" them. It is an added stressor in my classroom when I feel that it is my responsibility to meet the very basic needs of my students. I will do it and in a nurturing way because that seems to be what many of these students are missing but that doesn't mean it is fair. In many years of having student teachers and even brand new teachers in classrooms with populations they are not familiar with, the factors mentioned above are a huge learning curve. Wanting to refer an ELL student for SPED because they are not speaking is not the. first path we go down. I have been beating my head against a brick wall for some time trying to get two of my colleagues, one veteran and one novice, to understand the process that should be followed. Encouraging my mentees to get to know their families and have a better understanding of cultural differences without adding to their already overflowing plate is so important. I need to figure out how to encourage it and maintain the work/life balance for everyone.

!- question : does evolution have a purpose? * Language is a constraint - it forces us to form questions that may not necessarily make sense, such as "how many angels exist on the head of a pin?" * To say that it has one, or even more than one purpose may itself be a meaningless assertion, as much as insinuating that it has no purpose. If one asks "Is the sound of a bell red or yellow? It is neither the case that it is red, yellow or any color. So arguing about the right and wrong of a quality that is nonsensical is itself nonsensical. * The self-annihilating questioning of Nagarjuna's tetralemma are relevant to shed insight into these deep questions.

Author Response:

Reviewer #1 (Public Review):

The authors show that the unmitigated generation interval of the original variant of SARS-CoV-2 is longer than originally thought. They argue that in the absence of interventions that limit transmission late in the course of infection, the fraction of transmission events that occur before symptom onset would be considerably lower, and the fraction of transmission events occurring 10 days or more after infection of the index case would be substantially higher.

These findings improve our ability to accurately estimate the basic reproductive number (R0), to evaluate quarantine and isolation policies, and to model counterfactual intervention-free scenarios. Many applied analyses rely on accurate generation interval estimates. To head off confusion, it would be helpful if the authors could provide more comprehensive guidance about which applied analyses should be informed by the unmitigated generation interval, or the observed generation interval. (E.g. the unmitigated interval is useful for quarantine and isolation policies, but would we ever want to use the unmitigated interval to estimate R?).

The unmitigated generation-interval should be used for estimation of the R0 of the initial epidemic phase, but not for the R(t) of the current epidemics. Estimation of R(t) must account for changes in generation-interval distributions caused by the invasion of new variants and changes in behavior. When analyzing policies of quarantine, isolation or contact tracing, the unmitigated interval should also be adopted to account for late transmissions.

We added few sentences at the end of our introduction to clarify this point:

“The estimated unmitigated generation-interval distribution could be adopted for answering questions about quarantine and isolation policy, as well as for estimating the original R0 at the initial spread in China. However, estimation of instantaneous R(t) should account for changes in generation-interval distributions, reflecting mitigation effects and the current variant.”

The analysis estimates a longer generation interval after accounting for three main sources of bias or error that are common in other analyses: 1. Recently infected individuals are intrinsically overrepresented in data on a growing epidemic. Thus, shorter incubation periods and forward serial intervals are more likely to be observed, even in the absence of interventions. This analysis adjusts for these dynamical biases. 2. Interventions or behavioral changes can prevent transmission late in the course of infection. This can shorten the generation and serial intervals over the course of an epidemic. This analysis focuses specifically on transmission pairs observed very early, before the adoption of interventions. 3. The incubation period and generation interval should be correlated - infectors that progress relatively quickly to symptoms should also become infectious sooner (symptom onset occurs near the peak of viral titers). Most existing analyses assume these intervals are uncorrelated, but this analysis accounts for their correlation.

Overall, the conclusions seem reasonable and well-supported. The observation that the generation interval decreases over the course of an epidemic is also consistent with existing studies that show the serial interval has similarly decreased over time. But given the implications of the findings, I hope the authors can address a few questions about potential additional sources of bias:

1. It is somewhat reassuring that the generation interval decreases relatively smoothly as the cutoff date increases (Fig. S6), but it would be helpful if the authors address the potential impact of ascertainment biases. One of the main reasons that the authors estimate a shorter generation interval is that they define January 17th, early in the outbreak before interventions and behavioral changes had taken place, as the cutoff point for the infector's date of symptom onset. This cutoff eliminates biases from interventions, but it also severely limits the size of the transmission-pair dataset (Fig. S3), and focusing on this very early subset of cases may increase the influence of ascertainment bias. Prior to January 17th, should we expect observed transmission pairs to involve more severe cases on average? And is the unmitigated generation interval correlated with case severity?

We thank the reviewer for identifying a source of possible bias that we overlooked. Following the comment we performed a new sensitivity analysis for the inclusion of the severe cases, summarized in Appendix 1—figure 11.

Severity of the cases was reported only in Ali et al.’s data, for some of the individuals. In these cases, individuals are divided into one of three conditions: mild, severe (non-fatal) and death. As non-mild cases represent a small fraction of the dataset, we combine them into one category, which we denote as severe.

Severe cases (including deaths) were overrepresented in the period prior to January 17, with 8 out of 77 cases, compared to 18 out of 745 in the period of January 18-31. The effect of inclusion of severe cases was analyzed by comparing the means of the estimated generation-interval distribution, separately for the two periods in question, using the inference framework with 30 bootstrapping runs. For the earlier period, the estimated means were compared between the dataset with or without the severe cases. For the later period, we also consider the “enriched” dataset, in which severe cases are oversampled for each bootstrap such that the proportion of severe cases matches that during the earlier period (10%). In both cases we see that the effect on the estimated mean generation interval is small.

1. The analysis assumes the incubation period follows a fixed distribution, whose parameterization comes from a meta-analysis of previously estimated incubation periods. But p.5 discusses the idea that observed incubation periods are affected by the same dynamical biases as forward serial intervals, "For example, when the incidence of infection is increasing exponentially, individuals are more likely to have been infected recently. Therefore, a cohort of infectors that developed symptoms at the same time will have shorter incubation periods than their infectees on average, which will, in turn, affect the shape of the forward serial-interval distribution." Has the incubation period been adjusted for these dynamical biases, or should it be?

In our analysis we use the incubation period distribution from Xin et al. 2021 which already considers the backward bias caused by the expanding epidemic with the corrected growth rate of 0.1/d. Xin et al. showed in their meta-analysis that the mean incubation period reported by the various sources changed according to the dates used by the source. Incubation periods prior to the peak of the epidemic in China were lower than ones from after the peak, in a manner that coincided with the backward correction they performed (using a similar derivation to that suggested by Park et al. 2021). Accordingly, the distribution of incubation period they report is the intrinsic incubation period, after correction for the growth rate of the initial spread in China. We added two sentences in our methods section to clarify this point:

“In their meta-analysis, Xin et al. found an increase of the incubation period following the introduction of interventions in China, matching the theoretical framework shown above. Their inferred incubation period distribution includes a correction for the growth rate of the early spread, accordingly.”

Furthermore, we perform a sensitivity analysis for the shape of the incubation period distribution, and show that it has a minor effect on our conclusions (Appendix 1—figure 10).

1. It appears that correlation parameter estimates co-vary with estimates of the mean generation interval (Fig. S6; S13b). Are the authors confident that the correlation parameter is identifiable? How much would the median generation interval estimate in the main analysis change if the correlation parameter had been fixed to 0 (which isn't realistic) or to 0.5 (which might be plausible)?

As the reviewer pointed out, the correlation parameter estimates co-vary with estimates of the mean generation interval. We further analyzed this relation following the comment. The analysis is summarized in supplemental figures S19-20.

We first examine the relation between the mean generation interval and the correlation parameter based on the uncertainty estimates, consisting of 1000 bootstrap runs. Appendix 1—figure 12 shows a joint bivariate scatter plot of the parameters, together with contours of equal probability. As can be seen there is a connection between the parameters. The estimates centered around the maximum likelihood estimate with correlation parameter of 0.75 and mean generation interval of 9.7 days. The confidence interval for the correlation parameter of 0.45-0.95 corresponds to mean generation intervals in the range of 8-11 days, supporting the conclusion of this study.

Next, we reanalyzed the dataset while fixing the correlation parameter, as suggested by the reviewer. Appendix 1—figure 13 shows the estimated mean generation interval for fixed correlation parameters with values of 0, 0.25, 0.5, 0.75, 0.9. For each fixed correlation parameter 100 bootstrapping runs. As can be seen, the results reflect the same connection that can be seen in Appendix 1—figure 12, with probable values in the range of 8-11 days, for correlation parameters in the range of 0.5-0.9. Assuming no correlation would cause underestimation of the mean generation interval match previous literature (Hart, Maini, and Thompson 2021; Park et al. 2022).

Reviewer #2 (Public Review):

There have been several estimates of the generation time and serial interval published for SARS-CoV-2, but as the authors note, estimates can be subject to biases including shifted event timing depending on the phase of the epidemic, correlation in characteristics between infector and infectee, and impact of control measures on truncating potential infectiousness. This study, therefore, has several strengths. It first collates data on transmission events from the earliest phase of the COVID-19 pandemic, then makes adjustments for these potential biases to estimate the generation time in absence of control measures, and finally discusses implications for transmission.

Given many subsequent aspects of the COVID-19 pandemic have been defined relative to earlier phases (e.g. relative transmissibility of variants, relative duration of infectiousness), understanding the baseline characteristics of the infection is crucial. I thought this paper makes a useful contribution to this understanding, generating adjusted estimates for infectiousness (which is longer than previous estimates) and corresponding values for the reproduction number (which is remarkably similar to earlier estimates, presumably because of the different sources of bias in the growth rate and generation time distribution somehow end up canceling each other out).

However, there are some weaknesses at present. The study correctly flags several potential sources of bias in estimates, but in making adjustments uses estimates from the literature that themselves could suffer from these biases, e.g. the distribution of incubation period from a 2021 meta-analysis. Although the authors conduct some sensitivity analysis it would be worth including some more explicit consideration of whether they would expect any underlying bias to propagate through their calculations. The authors also conduct some sensitivity analysis around the underlying data (e.g. ordering of transmission pairs), but again it would be useful to know whether there could be systematic biases in these early data. Specifically, the paper references Tsang et al (2020), which highlighted variability in early case definitions - is it possible that early generation times are estimated to be longer because intermediate cases in the transmission chain were more likely to go undetected than later in the epidemic?

We recognize the potential biases in the transmission pairs data. We therefore developed an extensive framework of sensitivity analyses for identifying biases that could substantially affect the results. In the results section and figure 5, we show that the main study result, that the unmitigated generation-interval distribution is longer than previously estimated, is robust to reasonable amounts of ascertainment bias. We discuss this point at length and have added several supplemental figures to support this claim.

As reviewer #3 mentioned, we conducted a sensitivity analysis for the inclusion of the longest serial intervals, to investigate possible effects of missing links in the longest transmission pairs. We also discuss why we think it’s not necessary to explicitly model the short intervals that may be unobserved due to missing links.

“Second, we considered the possibility that long serial intervals may be caused by omission of intermediate infections in multiple chains of transmission, which in turn would lead to overestimation of the mean serial and generation intervals. Thus, we refit our model after removing long serial intervals from the data (by varying the maximum serial interval between 14 and 24 days). We also considered “splitting” these intervals into smaller intervals, but decided this was unnecessarily complex, since several choices would need to be made, and the effects would likely be small compared to the effect of the choice of maximum, since the distribution of the resulting split intervals would not differ sharply from that of the remaining observed intervals in most cases.”

We added to the discussion text regarding the effect of possible bias in the dataset, explicitly specifying the ascertainment bias.

“Our analysis relies on datasets of transmission pairs gathered from previously published studies and thus has several limitations that are difficult to correct for. Transmission pairs data can be prone to incorrect identification of transmission pairs, including the direction of transmission. In particular, presymptomatic transmission can cause infectors to report symptoms after their infectees, making it difficult to identify who infected whom. Data from the early outbreak might also be sensitive to ascertainment and reporting biases which could lead to missing links in transmission pairs, causing serial intervals to appear longer (For example, people who transmit asymptomatically might not be identified). Moreover, when multiple potential infectors are present, an individual who developed symptoms close to when the infectee became infected is more likely to be identified as the infector. These biases might increase the estimated correlation of the incubation period and the period of infectiousness. We have tried to account for these biases by using a bootstrapping approach, in which some data points are omitted in each bootstrap sample. The relatively narrow ranges of uncertainty suggest that the results are not very sensitive to specific transmission pairs data points being included in the analysis. We also performed a sensitivity analysis to address several potential biases such as the duration of the unmitigated transmission period, the inclusion of long serial intervals in the dataset, and the incorrect ordering of transmission pairs (see Methods). The sensitivity analysis shows that although these biases could decrease the inferred mean generation interval, our main conclusions about the long unmitigated generation intervals (high median length and substantial residual transmission after 14 days) remained robust (Figure 5).”

It would also be helpful to have some clarifications about methodology, particularly in how the main results about generation time are subsequently analyzed. For example, estimates such as the conversion of generation time to R0 and VOC scalings are described very briefly, so it is currently unclear exactly how these calculations are being performed.

Following the reviewer comments we made edits to the Methods section in order to make it more readable and clearer. We added subheadings for the various sections. Moreover, we added a section explaining the derivation of the basic reproduction number and clarified the section regarding the VOCs extrapolations.

We made some edits to the methods section in order to make it more accessible and clear, for example, we added subheadings for the various sections, added a section explaining the derivation of the basic reproduction number, and clarified the section regarding the VOCs extrapolations.

Reviewer #3 (Public Review):

Sender & Bar-On et al. perform robust analyses of early SARS-CoV-2 line list data from China to estimate the intrinsic generation interval in the absence of interventions. This is an important topic, as most SARS-CoV-2 data are from periods when transmission-reducing interventions are in place, which will lead to underestimation of the potential infectious period.

The authors highlight two shortcomings in previous approaches. First, the distribution of 'observed' serial intervals (the time between symptom onset in the infector and symptom onset in the infectee) depends not only on the timeline of each infector's infection, but also the epidemic growth rate, which weights the proportion of observed short vs. long serial intervals. The authors argue that by accounting for this weighting, more accurate estimates of the intrinsic generation interval - the metric on which isolation policies are based - can be obtained. Second, the authors find that the original SARS-CoV-2 generation interval distribution has both a higher mean and longer tail than previous estimates when using only data prior to the introduction of interventions. Finally, the authors use publicly available data on viral load trajectories to extrapolate their estimates to other SARS-CoV-2 variants, finding that alpha, delta, and omicron may have shorter generation intervals than original SARS-CoV-2. These findings are important, as case isolation policies are based on assumptions for how long individuals remain infectious. More broadly, these methods will be important for future work to correctly estimate generation intervals in other outbreaks.

The conclusions are well supported by the data, and a suite of sensitivity analyses give confidence that the findings are robust to deviations from many of the key assumptions. The code is well documented and publicly available, and thus the findings are easily reproducible. Key strengths of the paper include the clarity and rigor of the modeling methods, and the exhaustive consideration of potential biases and corresponding sensitivity analyses - it is very difficult to think of potential biases that the authors have not already considered! I think this is a well-written and well-executed study. The work is likely to be impactful for reconsidering SARS-CoV-2 isolation policies and revisiting generation interval estimates from other data sources. I also expect this to be a key reference and method for future studies estimating the generation interval.

I have some minor comments on potential weaknesses and interpretation:

1. Uncertainty in early generation interval estimates. One of the conclusions is that the estimated mean generation interval is longer than the observed mean serial interval. However, this conclusion does not seem justified given that the observed mean serial interval (9.1 days) is well within the 95% CI of 8.3-11.2 days for the mean generation interval. The confidence intervals for the serial interval in figure 2 are also wide for pre-Jan 17th (though presumably these would be reduced if all pre-Jan 17th serial intervals were combined). Further, only 77 of the ~1000 transmission pairs are actually from pre-January 17th. The actual sample size used for these estimates is much smaller than suggested by Figure S1 and thus this should be made clear. Therefore, although the intuition for why observed serial intervals may differ from the generation interval is correct, I do not think that the data alone demonstrate this. A related issue is on ascertainment bias - could the early serial interval data be biased longer because ascertainment is initially poor and thus more intermediate infectors are missed? The authors consider removing particularly long serial intervals to try and account for this, but that does not deal with e.g. chains of multiple short serial intervals being incorrectly recorded as a single long serial interval (but still within 16 days).

We agree with the reviewer that due the large uncertainty we cannot deduce that the mean generation interval is longer than the mean serial interval. We changed the phrasing to emphasize this statement is supported by mathematical theory.

“We note that our estimated mean generation-interval is longer than the observed mean serial-interval (9.1 days) of the period in question. This is supported by the theory (Park et al. 2021) of the dynamical effects of the epidemic -- in contrast to the common assumption that the mean generation and serial intervals are identical. During the exponential growth phase, the mean incubation period of the infectors is expected to be shorter than the mean incubation period of the infectee - this effect causes the mean forward serial interval to become longer than the mean forward generation interval of the cohorts that developed symptoms during the study period. However, these cohorts of infectors with short incubation periods will also have short forward generation (and therefore serial) intervals due to their correlations. When the latter effect is stronger, the mean forward serial interval becomes shorter than the mean intrinsic generation interval, as these findings suggest.“

Following the comment, we added to Figure S1 the filtering according to date, to reflect the true sample size we use for the main analysis (We renamed it: Appendix 1—figure 1).

We recognize the potential biases in the transmission pairs data. We therefore developed an extensive framework of sensitivity analyses for identifying biases that could substantially affect the results. In the results section and figure 5, we show that the main study result, that the unmitigated generation-interval distribution is longer than previously estimated, is robust to reasonable amounts of ascertainment bias. We discuss this point at length and have added several supplemental figures to support this claim.

As reviewer #3 mentioned, we conducted a sensitivity analysis for the inclusion of the longest serial intervals, to investigate possible effects of missing links in the longest transmission pairs. We also discuss why we think it’s not necessary to explicitly model the short intervals that may be unobserved due to missing links.

“Second, we considered the possibility that long serial intervals may be caused by omission of intermediate infections in multiple chains of transmission, which in turn would lead to overestimation of the mean serial and generation intervals. Thus, we refit our model after removing long serial intervals from the data (by varying the maximum serial interval between 14 and 24 days). We also considered “splitting” these intervals into smaller intervals, but decided this was unnecessarily complex, since several choices would need to be made, and the effects would likely be small compared to the effect of the choice of maximum, since the distribution of the resulting split intervals would not differ sharply from that of the remaining observed intervals in most cases.”

We added to the discussion text regarding the effect of possible bias in the dataset, explicitly specifying the ascertainment bias.

“Our analysis relies on datasets of transmission pairs gathered from previously published studies and thus has several limitations that are difficult to correct for. Transmission pairs data can be prone to incorrect identification of transmission pairs, including the direction of transmission. In particular, presymptomatic transmission can cause infectors to report symptoms after their infectees, making it difficult to identify who infected whom. Data from the early outbreak might also be sensitive to ascertainment and reporting biases which could lead to missing links in transmission pairs, causing serial intervals to appear longer (For example, people who transmit asymptomatically might not be identified). Moreover, when multiple potential infectors are present, an individual who developed symptoms close to when the infectee became infected is more likely to be identified as the infector. These biases might increase the estimated correlation of the incubation period and the period of infectiousness. We have tried to account for these biases by using a bootstrapping approach, in which some data points are omitted in each bootstrap sample. The relatively narrow ranges of uncertainty suggest that the results are not very sensitive to specific transmission pairs data points being included in the analysis. We also performed a sensitivity analysis to address several potential biases such as the duration of the unmitigated transmission period, the inclusion of long serial intervals in the dataset, and the incorrect ordering of transmission pairs (see Methods). The sensitivity analysis shows that although these biases could decrease the inferred mean generation interval, our main conclusions about the long unmitigated generation intervals (high median length and substantial residual transmission after 14 days) remained robust (Figure 5).”

1. Frailty of using viral loads to extrapolate generation intervals. The authors take the observation that variants of concern demonstrate faster viral clearance on average to estimate shorter generation intervals for alpha, delta, and omicron. The authors rightly point out in the discussion that using viral load as a proxy for infectiousness has many limitations. I would emphasize even further that it is very difficult to extrapolate from viral load data in this way, as infectiousness appears to vary far more between variants than can be explained by duration positive or peak viral load. Other factors are potentially at play, such as compartmentalization in the respiratory tract, aerosolization, receptor binding, immunity, etc. Further, there is considerable individual-level variation in viral trajectories and thus the use of a population-mean model overlooks a key component of SARS-CoV-2 infection dynamics. An important reference, which came out recently and thus makes sense to have been missed from the initial submission, is Puhach et al. Nature Medicine 2022 https://doi.org/10.1038/s41591-022-01816-0.

We agree with the reviewer about the frailty of using viral loads to extrapolate generation intervals. We therefore expanded our discussion of the limitation of using viral load data for inferring infectiousness including many of the points mentioned by the reviewer. We use viral load data in the most minimal way to try to enable some discussion of new VOC, and try to emphasize the needed caution.

Viral load trajectories data have potential for informing estimates of the infectiousness profile. However the relationship between viral load, culture positivity, symptom onset, and infectivity is complex and not well characterized. Due to this limitation we tried to use viral loads in a more limited way, extrapolating our results to variants of concerns (which lack unmitigated transmission data). Following the comment, we added a detailed discussion of the limitations of using viral loads as a proxy for infectiousness, including the variation of viral loads across individuals. We also added supplementary figures (Figure 6—figure supplements 1-2) to show the possible effect of an individual's viral loads in relation to the infectiousness and for comparison with new viral load and culture results (Chu et al. 2022; Killingley et al. 2022). As the viral load trajectories data for the different VOC is given only as a function of time from the onset of symptoms, it is not possible to directly link it to the fraction of transmission post 14 days from infection. We made changes to Figure 6 to clarify the possible connection of viral load with the TOST (time from symptoms onset to transmission) distribution and the resulting extrapolation to the unmitigated generation-interval distributions.

“SARS-CoV-2 viral load trajectories serve an important role in understanding the dynamics of the disease and modeling its infectiousness (Quilty et al. 2021; Cleary et al. 2021). Indeed, the general shapes of the mean viral load trajectories and culture positivity, based on longitudinal studies, are comparable with our estimated unmitigated infectiousness profile (Figure 6—figure supplements 1-2, comparison with (Chu et al. 2022; Killingley et al. 2022; Kissler et al. 2021)). However, the nature of the relationship between viral load, culture positivity, symptom onset, and real-world infectivity is complex and not well characterized. Therefore, the ability to infer infectiousness from viral load data is very limited, especially near the tail of infectiousness, several days following symptom onset and peak viral loads. Viral load models are usually made to fit the measurements during an initial exponential clearance phase and in many cases miss a later slow decay (Kissler et al. 2021). Furthermore, there is considerable individual-level variation in viral trajectories that isn’t accounted for in population-mean models (Kissler et al. 2021; Singanayagam et al. 2021). Other factors limiting the ability to compare generation-interval estimates with viral loads models are the variability of the incubation periods and its relation to the timing of the peak of the viral loads, and the great uncertainty and apparent non-linearity of the relation between viral loads and culture positivity (Jaafar et al. 2021; Jones et al. 2021). Due to these caveats and in order to avoid over interpretation of viral load data, we restrict our extrapolation of new VOCs’ infectiousness to a single parameter characterizing the viral duration of clearance.”

We also edited another paragraph in the discussion:

“Our extrapolations are necessarily crude given the complex relationship between viral load, symptomaticity, and infectiousness discussed above. Moreover, compartmentalization in the respiratory tract, aerosolization, receptor binding affinity, and immune history can also play important roles in determining the infectiousness profiles of SARS-CoV-2 variants (Puhach et al. ). ”

1. Lack of validation with other datasets This study hinges on data from a single setting in a short window of time. Although the data are from multiple publications, the fact that so many reported the same transmission pair data demonstrates that these are overlapping datasets. As the authors note, there are potential biases e.g., ascertainment rates and behavioral changes which will impact the generation interval estimates. Thus, generalizability to other settings is limited.

We agree with the reviewer that the dataset used in our study is limited, and consists of overlapping transmission pairs. We perform some analysis of the possible bias caused by inclusion of each dataset, as can be seen in Appendix 1—figure 4.

The best validation would have been a comparison with another independent dataset from the early spread of the epidemic, but no such dataset exists. We added a sentence to the discussion to emphasize this point.

“Due to the nature of early spread of a new unknown disease it is nearly impossible to find two completely unrelated datasets from the period prior to mitigation, limiting the ability of further validation of the current results.”

1. The impact of epidemic dynamics on infector vs. infectee serial intervals. It took me a long time to get my head around the assertion that the forward serial interval distribution will be longer during epidemic growth due to the overrepresentation of short incubation periods among infectors relative to infectees. A supplementary figure, similar to the way Figure 1 is laid out, to illustrate this concept may go a long way to aid the reader's understanding.

We added an explanation to the paragraph in order to make it clearer:

“A cohort of individuals that develop symptoms on a given day is a sample of all individuals who have been previously infected. When the incidence of infection is increasing, recently infected individuals represent a bigger fraction of this population and thus are over-represented in this cohort. Therefore, we are more likely to encounter infected individuals with a short incubation period in this cohort compared to an unbiased sample. The forward serial-interval is calculated for a cohort of infectors who developed symptoms at the same time and therefore is sensitive to this bias. These dynamical biases are demonstrated using epidemic simulations by Park et al."

1. Simulations to illustrate concepts and power Given the assertion that observed serial intervals will depend on epidemic growth rates, reporting, and timing of interventions, I think a simple simulation to illustrate some of these ideas would be very helpful. For example, a simple agent-based model with simulated infectivity profiles and incubation periods using the estimated bivariate distribution would be extremely helpful in illustrating how serial intervals and estimates of the generation interval can differ from the true intrinsic generation interval (I coded such a simulation to help me understand this paper in a couple of hours with <100 lines of R code, so I do not think this would be much work). This would also be very helpful for illustrating statistical power re. comment 1.

The current paper is based on a strong theoretical foundation provided by previous works, specifically Park et al. 2021, which used simulations similar to the reviewer’s suggestions to demonstrate the dynamical biases. We now mention these simulations somewhere in the introduction section:

“These dynamical biases are demonstrated using epidemic simulations by Park et al."

page 109 "I notice everything in this manner and always have and have no control over it; in fact it tends to annoy me, for it wearies my head; I thought this, and composing verses, happened to everyone, until experience showed me the contrary; and this is so much my character or custom that I see nothing without considering it further."

someone call 911 i need to be revived

the casual drama of harry vs the drama drama of draco

Reinforces my previous annotation in Ezra's narrative that his dialogue/character has a significantly Romantic influence

This passage highlights how strong Betteredge's sense of duty and service is, even against his own morals. This highlights how Betteredge is likely heavily biased against narrating that the people he serves would do much wrong, even if he thought differently.

If the Indians were indeed innocent, then what "purpose" do they have in this mystery besides a relatively minor red herring?

What's the role of self-deprecating humor in this novel, especially on the part of Betteredge the 'house-steward' narrator/character? So far, no other narrators/characters self consciously make fun of themselves, although Betteredge will describe the silliness or odd behavior of other characters. Which ones are not "clownish" and why? And how do these descriptions affect readers' judgements about various characters' reliability about the information and observations they offer?

This paragraph goes into depth about the dark methodical nature of the wealthy where they dissect and tear apart things around them because they're restless from having absolutely nothing to do. In the previous paragraph, Betteredge foreshadows that more information about what Mr. Franklin and Miss Rachel do with their time. Is this description of cruelty hint that Betteredge believes someone in the wealthy class (not Miss Rachel or Mr. Franklin, they just set up the idea) is capable of murder and would commit such an act? There's plenty of mention of heads and hands with the dissecting, so could this also hint that a murder was thought out by one (or many) then followed through by them with their own hands? This overall section is very ominous and there is many possible ways that we can hypothesize with the way Betteredge has worded it.

Ancient statuary wasn't white as we often see now in museums, but was brightly colored. Statuary that was outside would have been sun bleached over time as well as subject to other weathering to mute or entirely remove color.

https://www.npr.org/2022/07/12/1109995973/we-know-greek-statues-werent-white-now-you-can-see-them-in-color

Oh god, proponents of new philosophies being invited to test out their ideas on the economy of Chile by the head of state?

bless it

this is so open and honest ugh i love him

It was surprising to see that stories do not need to initially fit into a clear theme as long as you are able to tie it all together in a coherent way at the end. Your intention does not need to be in your face obvious—it can slowly become apparent as the story progresses. In fact, it's more effective this way than if you just bludgeon the reader over the head with your message. I learned that essays that incorporate many stories and essays that incorporate only one can be equally effective, although I think at this stage I'm partial to essays that don't include more than, say, four or so stories.

Reviewer #2 (Public Review):

In this study, the authors take advantage of unbiased scRNA-seq datasets of the developing mouse soft palate that they previously reported and performed a new bioinformatic analysis to identify differential signaling pathway activities in the heterogeneous palatal mesenchyme. They found a strong association of TGF-beta signaling pathway activity with the perimysial cells and validated through immunofluorescent detection of pSmad2, which led to their hypothesis that TGF-beta signaling in the perimysial cells might regulate palatal muscle formation. They generated and analyzed Osr2-Cre;Alk5fl/fl mice and showed those mice have cleft soft palate and disruption of the levator veli palataini (LVP) muscle. They then performed a comparative scRNA-seq analysis of the soft palate tissues from E14.5 Osr2-Cre;Alk5fl/fl and control embryos and showed that the Osr2-Cre;Alk5fl/fl embryos exhibited defects in the perimysial cells, in particular reduction in Tbx15+ perimysial fibroblasts that directly associate with the LVP muscle progenitors. The FGF18 is one of the most highly enriched signaling molecules in the perimysial cells and showed that the Osr2-Cre;Alk5fl/fl embryos exhibited reduced Fgf18 expression together with loss of MyoD+ myoblasts in the prospective LVP region. Further data showed that pSmad2 bound in the Fgf18 promoter region in the developing soft palate tissues. In addition, bioinformatic gene regulatory network analysis of the scRNA-seq data identified Creb5 as a potential tissue-specific transcription factor in the perimysial cells and RNAi knockdown assays in palatal mesenchyme culture suggested that Creb5 is required for Fgf18 expression. Further studies identified a subtle deficiency in LVP in Osr2-Cre;Fgf18fl/fl mice and showed that exogenous Fgf18 bead implantation in explants of E14 Osr2-Cre;Alk5fl/fl embryonic head increased the MyoD+ myoblast population in the prospective LVP region. The authors concluded that TGF-beta signaling and Creb5 cooperatively regulate Fgf18 to control pharyngeal muscle development. While the study used multiple complementary approaches and the data presented are solid, important questions need to be addressed to resolve reasonable alternative explanations of the data to the authors' main conclusion.

Major points:<br /> 1. TGF-beta signaling is known to be crucial for neural crest-derived palatal mesenchyme cell proliferation from E13.5 to E14.5. The Osr2-Cre;Alk5fl/fl mutant embryos exhibited obvious disruption of LVP myogenesis and reduced soft palatal shelf size at E14.5 (Fig3-Sup2A-D and Fig 4H-K). The cellular and molecular defects likely started prior to E14.5. Thus, it is important to examine at earlier stages (E13.5/E14.0) whether the palatal mesenchyme was already defective in cell proliferation/survival and/or perimysial cell marker expression, including Creb5 and Tbx15, to resolve whether the primary defect in the Osr2-Cre;Alk5fl/fl palatal mesenchyme could be a reduction in perimysial progenitor cell proliferation and/or differentiation of the myoblast-associated subset, for which Tbx15 and Fgf18hi act as marker genes rather than direct molecular targets. Furthermore, the apparent loss of Tbx15+ cells coincided with a specific reduction of Fgf18 expression in the myoblast-associated perimysial cells (Fig 4J/K versus Fig 5H-K), which raises the possibility that TGF-beta signaling regulates the differentiation of the Tbx5+ population from the mesenchymal progenitors while the reduction in Fgf18 expression might be a secondary consequence of the cellular defect. The data in Fig 6O showing a lack of significant induction of Fgf18 expression in the palatal mesenchyme culture in both control and Creb5-RNAi cells is also consistent with this alternative explanation.<br /> 2. Since the Osr2-Cre;Fgf18fl/fl mice exhibited much subtler palatal and LVP defects than the Osr2-Cre;Alk5fl/fl mice even though the latter still had a lot of Fgf18-expressing perimysial cells at E14.5, Fgf18 is likely a minor player in the TGF-beta mediated gene regulatory network regulating LVP formation. The major players acting downstream of TGF-beta signaling in the palatal mesenchyme, that control initial LVP progenitor migration to and/or proliferation in the soft palate region, remain to be identified and functionally validated. Whether and how Fgf18 directly regulates the perimysial-myoblast interaction is also not known.<br /> 3. While the title and the main conclusion of this manuscript imply a crucial role of Creb5 in the regulation of pharyngeal muscle development, there is no data supporting such a crucial role. Do Creb5-/- mice have specific defects in pharyngeal muscle development?<br /> 4. Data in Fig 6 are not sufficient to conclude that TGF-beta signaling and Creb5 cooperatively regulate Fgf18. The TGFb1 treatment did not significantly induce Fgf18 expression in either the control or Creb5-RNAi palate mesenchyme cells (Fig 6O). No data regarding how they act cooperatively to regulate Fgf18 expression.

Reviewer #2 (Public Review):

In this study, the authors examined the associations between brain morphology and ADHD symptoms and how the adjustments for confounders change these associations. While the socioeconomic and maternal behavioral confounders were overlooked in most of prior ADHD neuroimaging studies, the authors show that controlling for these confounders in addition to the demographic variables generally attenuated the associations. The authors proposed using and building Directed Acyclic Graphs (DAGs) to identify confounders, colliders, and mediators, which helps present the research questions clearly. Notably, the authors also examined the potential role of IQ in the brain morphology-ADHD associations and concluded that IQ may be a confounder, mediator, or collider, thus the adjustments for IQ are often unnecessary.

Importantly, the authors used two independent large datasets to replicate the findings which strengthen the confidence in the results (although the significant clusters are slightly different between the two datasets). Large cohort studies could have a large number of confounds, including scanner acquisition protocol processing parameters, head motion confounds, and so on. Although this study only focused on limited confounders related to ADHD, the suggestions from authors for minimizing confounding bias are useful for future studies.

Reviewer #2 (Public Review):

Abbassi-Daloii et al. investigated the molecular and cellular signatures associated with heterogeneity among human leg muscles using RNA-sequencing and immune-histology phenotypic characterization. They analysed 128 biopsies sampled from 6 leg muscles of 20 young healthy male individuals: two muscles from the hamstrings (semitendinosus (ST) and gracilis (GR)), three from the quadriceps (rectus femoris (RF), vastus lateralis (VL) and vastus medialis (VM)) and one lower leg muscle (gastrocnemius lateralis (GL)). They also analysed the middle and distal parts of the semitendinosus (STM and STD). Using the expression of known markers of specific cell types, they show that muscles cluster into three main groups based on cell type composition: group 1 (G1: GR, STM, and STD), group 2 (G2: RF, VL, and VM) and group 3 (G3: GL). Interestingly, these groups correspond to their anatomical location of origin. Group 1 was enriched in markers of fast-twitch muscle fibres, while Groups 2 and 3 were enriched in markers of endothelial cells and slow-twitch muscle fibres. Muscles clustered similarly after excluding genes related to cell type composition, indicating that a difference in cell-type composition between muscles is not the only factor accounting for their different molecular signatures. They further show that genes related to oxidative phosphorylation and mitochondria-associated metabolic processes are enriched in groups 2 and 3, consistent with their higher expression of markers of slow-twitch fibres and associated oxidative metabolism.

In addition, the authors developed a high-throughput quantitative immunohistochemistry procedure to measure the relative expression of specific myosin heavy chain isoforms on muscle cryosections from the same biopsies. They suggest that muscle fibres can be separated into three main clusters according to the expression of MyHC1 (slow-twitch fibres), MyHC2A (fast-twitch fibres type 2A) and MyHC2X (fast-twitch fibres type 2X). They show that myofibres from groups 2 and 3 are enriched in slow-twitch fibres compared to group 1, in agreement with the gene expression data, and that muscles from groups 2 and 3 can be discriminated according to their relative composition of type 2A fibres.

Gene expression and immunohistochemistry data showed that the GL muscle has a higher blood vessel density compared to other muscles and that HOX genes have a differential expression pattern among the analysed muscles.

These data provide an impressive dataset of human muscle gene expression from young healthy individuals and will serve as a reference for future investigation of muscle pathology or ageing. The conclusions of this paper are mostly well supported by the data, but some aspects of the analyses by immunohistochemistry need to be clarified.

1) The method to cluster fibres according to their relative expression of myosin isoforms needs to be further explained (Figure 3). Fibres were clustered into three main categories according to the fluorescence intensity of MyHC1, MyHC2A and MyHC2X after immunohistochemistry. Figure 3B shows a continuum of the three intensities rather than clearly separated clusters and the authors should explain how fluorescence intensity thresholds were used to define such clusters. In addition, the differences in average Mean Fluorescence Intensity (MFI) values between clusters appear quite low: Max(MyHC1)/Min(MyHC1)=0.725/0.653=1.11;<br /> Max(MyH2A)/Min(MyH2A)=0.806/0.535=1.50;<br /> Max(MyH2X)/Min(MyH2X)=0.718/0.651=1.10.<br /> These differences are low, notably compared to the gene expression data where these myosin isoforms have expression levels spread over 2 logs (Figures 3D-F). The authors should show specific examples of quantification, show specific quantification of fluorescence over MFI of background, and add statistics of expression differences based on MFI for each MyHC between clusters.

Figures 3D-F show correlations between the % of fibres in a given fluorescence cluster as a function of the expression by RNAseq of the corresponding MyHC defining this cluster. To strengthen these correlations, the authors should show the % of fibres in a given fluorescence cluster as a function of the expression of all three MyHC isoforms (e.g. % of fibres in cluster 1 = f(expression MYH7) or f(expression MYH1), same for clusters 2 and 3).

2) The authors assess blood vessel density in GL by immunohistochemistry (Figure 4). Figure 4A does not mention the muscle corresponding to the cryosection presented, and presenting STM and GL sections side-by-side would help to understand the conclusions of this figure. Displaying the green and red arrows as well on the CD31 and ENG panels and showing higher magnifications would help to understand which regions were defined or not as capillaries.

3) The authors validate HOX expression patterns by RNA-scope (Figure 8). Figure 8A does not mention the muscle corresponding to the cryosection presented. Maybe it would help show STM and GL sections stained for HOXA10 and HOXC10 side-by-side. Also, although the authors mention that signal specificity of RNA-scope probes was verified using negative controls, it would be helpful to show these controls and validate these probes on muscle sections known not to express HOX genes (head-derived?).

Finally, gene expression datasets from human muscle samples have already been generated. As discussed by the authors, these studies were limited in terms of the number of samples, large variation of donor ages, sample conservation before processing, etc. Nevertheless, it would be helpful to put the main findings of this paper (cell type composition, blood vessel density, fibre types ratios, HOX genes expression, mitochondrial processes, etc) into context and assess, if possible and if the data is available, whether similar findings can be concluded from previous datasets.

.c1

The proprietor seemed to not only have a prejudice against him but also a real hatred for the author which extended beyond merely fearing him.

The essay closes with a critique of the subject / object mind / body framework that now dominates modernity. Socrates, Kierkigaard and Becker's claims, when seen through the lens of Cartesian modernity, are relegated to the margins. materialism denies any legitimacy to such claims. Recent 4E cognition is an attempt to push back on this. Hughes notes that:

"In his last book, Escape from Evil, Becker wrote: “Hubris means forgetting where the real source of power lies and imagining that it is in oneself (37).” "

Reviewer #2 (Public Review):

Gradual increases in neural activity have been observed prior to execution of movements in several species, including in humans. Yet, the nature of this ramping activity and whether it signals the intent to act is debated. To bring novel insights into this debate, the authors examined the cortical activity that preceded self-initiated actions in mice and evaluated the predictive relationship between pre-movement activity in different regions of the dorsal cortex and the lever-pull behavior of mice.

Strengths: The manuscript addresses a timely and controversial topic in the field of neuroscience. The authors provide a novel and relatively rich dataset in mice with the goal of tackling a question that remains challenging to answer with current techniques used in human subjects. The data consists of longitudinal widefield imaging across the entire dorsal cortex of head-fixed mice performing a behavioral task paired with video monitoring of body movements. This dataset could be quite versatile and potentially useful to the community if shared publicly as the authors intend to. The manuscript is well written and easy to read, and the figures are quite self-explanatory.

Weaknesses: While I applaud the use of a "simplified" task in rodents to disambiguate controversial questions traditionally addressed in human studies, I found that the behavioral data were under-analyzed and thus not strongly supporting the central claim of the manuscript. Below are my main comments:

1. One of the goals of the authors was to study the neural mechanisms underlying "voluntary" movements. While they acknowledge (in the discussion) that they do not have evidence that actions are "intentional", they make the assumption that mice do "form the intent to act near the lever pull time". To back up this assumption, the authors should at least present some evidence that the action of interest (i.e., the rewarded lever-pull) is not just a random jerky movement that happens to be rewarded once in a while. In fact, mice seemed to pull the lever very frequently and impulsively (the majority of inter-pull intervals were way below 3 s in Supplementary Fig. 1.2) even for the last sessions of the training. Therefore, it is not readily apparent that mice apply any control to their lever-pull actions. Providing evidence that the action is goal-directed is important if the goal of the paper is to study neural signatures of the intention to act. A somewhat compelling analysis could be to compare rewarded lever-pulls with "spontaneous" movements, provided that these two types of movement can be convincingly characterized as goal-directed vs. incidental. In contrast, throughout the manuscript, the neural activity aligned to rewarded lever-pull events (which are assumed to be "voluntary" actions) is compared to the neural activity aligned to random times during the task (whether or not it involved movements), which may not be the most convincing control.

2. The learning trajectory of mice is also not well characterized (e.g. changes in inter-pull intervals are not quantified, nor the relative increase in rewarded actions across training sessions, etc.). Yet, several claims in the paper are directly based on the fact that mice have learned to pull the lever after 3 s interval to receive water rewards (which relates to point 1). In particular, one important assumption in the paper is that as mice learn, the lever-pull movements become more stereotyped, but this has not been shown explicitly. It would be helpful, for example, to see how analog traces of lever-pulling change throughout the learning stages and how the variance of the movement across trials decreases in late sessions.

3. The central claim of the paper is that rewarded lever-pulls can be predicted from pre-movement neural activity several seconds (even up to 10 s) prior to the action. However, obvious motor confounds and other alternative explanations have not been convincingly ruled out. In fact, the action of lever pulling may require a series of complex movements (like changing posture, extending the forelimb, reaching the lever, grabbing the lever, etc.). The authors themselves mentioned that they found strong correlations between lever pulls and body movements in all mice, but the data is not used nor shown in the paper. The motor commands preceding but related to lever-pull could unfold at least a few hundreds of milliseconds prior to the detection of lever-pull in the task, and thus be reflected in the neural activity that is predictive of the lever pull. Moreover, if this series of movements is highly stereotyped, and in turn leads to stereotyped neural activity (like the slow oscillations observed before the lever-pulls), it could explain why the detection of lever-pulling actions always occurs at a given phase of the neural oscillation. Such observations that stereotyped movements occur way before the lever-pull detection could partially rule out the fully "cognitive" explanation proposed in the paper, but would concur with recent findings that showed that ramping neural activity can be, for the most part, explained by movement-related activity (Musall et al., 2019).

4. Toward the end of the result section (Fig. 6), the authors briefly begin to address the issue about whether pre-movement activity can really be considered movement free. Here, "lockouts", i.e. periods where other movements (like licking, or previous lever-pulls) did not occur, were introduced in the analysis. The lockouts altered the earliest-decoding-time (EDT) of the lever-pull (in some mice EDT was even divided by half: from -4 s to -2 s). However, the effects of "micro-movements" like facial movements or changes in body posture may not be taken into account with the lockout approach. Such micro-movements have been shown to explain a large variance of the neural activity (see Stringer et al. 2019 and Musall et al. 2019). Therefore, to fully control for movement confounds, the effect of high dimensional/micro-movements extracted from video recordings should be removed from the neural activity. These analyses could yield a much shorter EDT (e.g., -0.15 s), more consistent with previous reports.

I'd be memorising music in case there is a time that I can only hear the soundtrack 'inside my head'. I'd be memorising things of beauty in case there is a time that I cannot see with my 'outer eye'

same with hair pulling- don't pull away- instead pull the client's hand toward your head and push up on the hand bending back the wrist to increase the likelihood of the patient having to let go from pain.

Uexkull also studies how animals synthetic apriori sense of space differ between species. Humans and other animals have semi-circular canals in the ear and this helps then determine forward/backwards, up/down and left/right of volumetric space. Limpets and paramecium do not have such a semi-circular canal and therefore do not sense "3 dimensional space" the way that we do.

In 1896, John Dewey wrote a famous paper called "The reflex arc concept in psychology". At that time, psychology was nascent and the reflex arc began to emerge as a framework to synthesis the many disparate findings.

the notion of the "cognitive sandwich" emerged which claimed that sensory inputs fed a mysterious cognitive function sandwiched between the input and motor system outputs.

Dewey said that the input output models relegates subjectivity to something trivial so cannot be correct.

This crude model still manifests in today's common belief that after sensing, there is a process of perception that feeds into higher cognition.

John explains how the 4 P's originated as a way to summarize and present in a palatable way of presenting the cognitive science “4E” approach to cognition - that cognition does not occur solely in the head, but is also embodied, embedded, enacted, or extended by way of extra-cranial processes and structures.

talking about the VR control of drone thingy

Reviewer #1 (Public Review):

Understanding how changes in microRNA (miRNA) levels affect human brain development remains a crucial task that has been largely understudied. This paper is an unbiased, large-scale, important contribution to this effort. The authors first performed small RNA sequencing from mid-gestation human cortical tissue to identify expressed miRNAs. Then, they mapped 85 local-miRNA-eQTLs that rarely colocalize with mRNA-eQTLs for the host mRNA. This is a very significant discovery, with a broad impact on the field. The lack of colocalization between miRNA- and mRNA-eQTLs reinforces the theory that miRNA transcriptional mechanisms are often independent of the host gene. Interestingly, miRNA-eQTLs map less frequently to risk loci. Considering the well-known importance of miRNAs during brain development, it is likely that the expression mechanisms of miRNAs are tightly regulated, as the authors suggest. The authors then focus on one specific miRNA-eQTL that affects miR-4707-3p and show colocalization with GWAS signals for head size (smaller) and educational attainment (decreased). They hypothesize that miR-4707-3p affects brain development by altering progenitors' proliferation and they partially support the hypothesis by over-expressing miR-4704-3p in human neural progenitor cells and showing an increase in neurogenesis.

Reviewer #2 (Public Review):

The majority of genetic effects discovered in genome-wide association studies (GWAS) of common human diseases point to non-coding variants with putative gene regulatory effects. In principle, studying genetic effects on gene expression phenotypes, as mediators between genotype and disease, can help understand the underlying function of GWAS variants.

Lafferty et al., set to study the regulation of microRNA (miRNA) levels in mid-gestation human neocortical tissues as a potential contributor to brain-related phenotypes. To this end they performed miRNA expression profiling via small-RNA sequencing, followed by assaying expression quantitative trait loci (eQTLs) that locally regulate miRNA genes.

In addition to reporting some properties of miRNA-eQTLs, e.g., their tissue-specificity, the authors searched for potential overlap or "colocalization" between these eQTL loci and GWAS loci for several putatively brain-related phenotypes. They reported colocalization at the locus containing the SNP rs4981455 which is an eQTL for miR-4707-3p and is also associated with global cortical surface area (GSA) and educational attainment phenotypes in GWAS. They further showed that exogenously increased expression of miR-4707-3p in primary human neural progenitor cells (as a model to study neurogenesis) derives an increased rate of proliferation.

The reported results are interesting and important, particularly for the understanding of miRNA biology. That said, as I detail below, the claim that miR-4707-3p expression modulates brain size and thus cognitive ability, although potentially consistent with the data, is not unequivocally supported by the analyses. As such, considering the potential social impact of the misinterpretations of these results, I believe the authors should explicitly discuss caveats, alternative explanations consistent with the data, and broader implications:

1. The colocalization analysis used effectively tests whether miRNA-eQTL and GWAS variants are in linkage disequilibrium (LD), and does not formally test whether the miRNA-eQTL and GWAS signals are explained by the same genetic variant which is necessary for establishing causality. In this study, a formal test of colocalization is challenging given that the LD patterns in the eQTL data (from mixed ancestries) are dissimilar to the GWAS data (from European-descent samples). Furthermore, even if GWAS and miRNA-eQTL signals are explained by the same variant, this could be due to confounding (a confounder affecting both), or pleiotropy (genotype independently affecting both), and not necessarily that the miRNA-eQTL signal mediates the GWAS signal. These are also true for colocalization analyses of miRNA-eQTLs with mRNA-eQTLs or splicing-QTLs. One practical suggestion is whether authors can perform the colocalization analysis better, e.g., with methods such as SMR (https://yanglab.westlake.edu.cn/software/smr/#Overview).

2. Although possible, there is no direct evidence that the GWAS signals at rs4981455 for educational attainment and GSA are driven by variation in miRNA levels in the studied tissue. As the authors noted, rs4981455 is also an eQTL for the gene HAUS4. Furthermore, rs4981455 is a significant e/sQTL across almost all adult tissues in GTEx, and so likely has regulatory activity across wide ranges of cell or tissue types. Therefore, pinpointing the causal contexts mediating the effect in GWAS is impossible with the current data.

3. Orthogonal to the issues above, the genotype-to-phenotype pathway as hypothesized, i.e., genotype → miRNA levels → brain structure → educational attainment, is oversimplistic and rests on an implicit prior belief that genetic associations with educational attainment can be trivially mapped to fundamental brain features that determine cognitive ability. To illustrate the problem with this prior I refer to an old example by Christopher Jencks: in a society that prevents red-hair kids to go to school, genetic effects on hair color would be associated with educational attainment, despite having no intrinsic biological relationship with cognition. I give two scenarios consistent with the specific case of rs4981455 that are fundamentally different from what is implied in the paper: (i) The case of indirect genetic effects (see Kong et al., Science 2018). In this case, rs4981455 affects the nurturing behavior of an individual's parents, which in turn influences the individual's educational achievements and consequently brain structure features. (ii) The case of confounding. In this case, the genetic effects on brain structure are shared with another feature, such as facial shape (see Naqvi et al., Nature Genetics 2021). Variation in facial shape in a discriminatory educational environment can covary with educational attainment.

4. The paper lacks a discussion on caveats to protect against potential misinterpretation of findings, especially considering the troubled history of linking facial shape and head morphology to human behavior and intelligence. I refer to the last paragraph of Naqvi et al., Nature Genetics 2021, as an example of such discussion. This is particularly crucial given that the frequency of rs4981455 varies across human populations. For example, it is important to state that the GSA and education attainment GWAS findings are in individuals of European descent, and may not necessarily point to an effect in other ancestries or even in European-descent individuals that differ from the GWAS samples in various ways, e.g., socioeconomic status (see Mostafavi et al., eLife 2020). In other words, these findings pertain to variation within the studied samples. On this note, it is important to state the amount of variation in multiple phenotypes explained by rs4981455 (which is likely tiny), and that it by no means determines the phenotype.

5. The main colocalization signal is tentatively shown for GSA. However, the authors casually refer to links with "brain size" or "head size" throughout the paper.

WSJ：中国警方的数据库对外网开放访问了一年多，导致泄密

今年早些时候发现安全漏洞的网络安全专家说，这可能是历史上最大的个人数据盗窃案之一，也是中国最大的已知网络安全漏洞，因为一个常见的漏洞使数据在互联网上被公开。

据网络安全专家称，上海警方的记录包括近10亿中国公民的姓名、政府身份证号码、电话号码和事件报告，是安全存储的。但他们说，一个用于管理和访问数据的仪表板被设置在一个公共网站上，而且没有密码，这使得任何具有相对基本技术知识的人都可以随意进入，复制或窃取信息库。

暗网情报公司Shadowbyte的创始人Vinny Troia说："他们将这么多数据暴露在外，这太疯狂了。"

网络安全研究公司SecurityDiscovery的老板鲍勃-迪亚琴科（Bob Diachenko）表示，该数据库从2021年4月到上个月中旬一直暴露在外面，当时其数据突然被清除，取而代之的是一张赎金字条，上海警方发现了这一点。

根据Diachenko先生提供的截图，"你的数据是安全的"，赎金字条上写道。"联系你的数据......恢复10btc"，意思是数据将以10个比特币，大约20万美元的价格被归还。

这个赎金数额与一位匿名用户上周四在一个在线网络犯罪论坛上开始要求的价格相吻合，以换取对一个数据库的访问权，该用户声称该数据库包含从上海国家警察数据库盗取的数十亿条中国公民的信息记录。

这个周末开始在社交媒体上流传的帖子引起了网络安全专家的警觉，不仅是因为泄漏的规模，还因为政府数据库中的信息的敏感性。

上海市政府和中国网信办（该国的互联网监管机构）没有对评论请求作出回应。

网络安全专家拼凑出了数据库真实性的新证据，以及这么多私人信息如何落入网络犯罪分子手中的细节。

他们说，仪表盘就像一扇通向数据库的大门，即使在所有数据丢失后也没有关闭，直到这个漏洞开始得到公众的广泛关注。特罗亚先生说，窃取数据的人很可能是正在兜售数据的同一个实体。

他说："很常见的是，如果赎金受害者不支付赎金，那么他们会试图在网上出售数据。"

无法确定该数据库被公开访问是偶然还是故意的，也许是为了在少数人之间更容易分享数据。特罗亚先生和迪亚琴科先生说，这种漏洞很常见，但他们都表示，发现如此规模的不安全数据库令他们感到震惊。

两人都说，他们也证实了匿名泄密者的说法，即它包括超过23兆字节的数据，涵盖多达10亿人。他们说，一个名为 "person_address_label_info_master "的文件包含了人们的姓名、生日、地址、政府身份证和身份证照片，长度接近9.7亿行，这表明它包括同样多的人的详细信息，假设没有重复的条目。

该文件标记了有犯罪史的个人，包括有交通违规行为的人、被认为是逃犯的人以及被指控犯有强奸或杀人罪的人。它还包括一个 "应该被密切关注的人 "的标签，这是中国政府监控系统中经常使用的一种称呼，以表示被认为对社会秩序构成威胁的人。

数据泄露突出了一些政策研究人员所描述的中国信息安全方法中的核心矛盾。

—— 华尔街日报 (https://www.wsj.com/articles/china-police-database-was-left-open-online-for-over-a-year-enabling-leak-11657119903?mod=djemalertNEWS)

China Police Database Was Left Open Online for Over a Year, Enabling Leak Cybersecurity experts say error allowed theft of records of nearly 1 billion people, leading to $200,000 ransom note By Karen HaoFollow in Hong Kong and Rachel LiangFollow in Singapore July 6, 2022 11:05 am ET PRINT TEXT What is likely one of history’s largest heists of personal data—and the largest known cybersecurity breach in China—occurred because of a common vulnerability that left the data open for the taking on the internet, say cybersecurity experts who discovered the security flaw earlier this year.

The Shanghai police records—containing the names, government ID numbers, phone numbers and incident reports of nearly 1 billion Chinese citizens—were stored securely, according to the cybersecurity experts. But a dashboard for managing and accessing the data was set up on a public web address and left open without a password, which allowed anyone with relatively basic technical knowledge to waltz in and copy or steal the trove of information, they said.

“That they would leave this much data exposed is insane,” said Vinny Troia, founder of dark web intelligence firm Shadowbyte, which scans the web for unsecured databases and found the Shanghai police database in January.

Leaked Shanghai police records contained names, ID information and incident reports for nearly 1 billion Chinese citizens. PHOTO: ALEX PLAVEVSKI/SHUTTERSTOCK The database stayed exposed for more than a year, from April 2021 through the middle of last month, when its data was suddenly wiped clean and replaced with a ransom note for the Shanghai police to discover, according to Bob Diachenko, owner of the cybersecurity research firm SecurityDiscovery, which similarly found the database—and later the note—through its periodic web scans earlier this year.

“your_data_is_safe,” the ransom note read, according to screenshots provided by Mr. Diachenko. “contact_for_your_data…recovery10btc,” meaning the data would be returned for 10 bitcoin, roughly $200,000.

The ransom amount matches the price that an anonymous user began asking for last Thursday on an online cybercrime forum in exchange for access to a database the user claimed contained billions of records of Chinese citizens’ information stolen from a Shanghai national police database.

The post, which began circulating on social media over the weekend, alarmed cybersecurity experts not just for the leak’s size but also because of the sensitivity of the information contained in the government database.

The Shanghai government and the Cyberspace Administration of China, the country’s internet regulator, didn’t respond to requests for comment.

Cybersecurity experts have pieced together new evidence of the database’s authenticity and details of how so much private information could have fallen in the hands of cybercriminals.

The dashboard acted like an open door to the data vault, they say, which wasn’t closed—even after all the data went missing—until the vulnerability began gaining widespread public attention. Whoever stole the data is likely the same entity that is peddling it, according to Mr. Troia.

“What’s pretty common is if the ransom victim doesn’t pay the ransom, then they’ll try to sell the data off online,” he said.

It couldn’t be determined whether the database was made publicly accessible by accident or on purpose, perhaps to share the data more easily among a few people. Such vulnerabilities are common, Mr. Troia and Mr. Diachenko said, though both said they were shocked to find an unsecured database of this size.

Both said they also corroborated the anonymous leaker’s claims that it includes over 23 terabytes of data covering as many as a billion individuals. One file named “person_address_label_info_master”—which contains people’s names, birthdays, addresses, government IDs and ID photos—runs close to 970 million rows long, they said, which suggests it includes details on just as many people, assuming no duplicate entries.

That file marks individuals who have a criminal history, and includes people with traffic violations, those considered fugitives and those who have been accused of rape or homicide. It also includes a label for “people who should be closely monitored,” a designation often used in China’s government surveillance systems to denote people seen as posing a threat to social order.

One leaked Shanghai police file includes records ranging from traffic violations to murder accusations. PHOTO: ALEX PLAVEVSKI/SHUTTERSTOCK The data leak highlights what some policy researchers have described as the central contradiction in China’s approach to information security.

In recent years, Beijing has signaled that data security and privacy are a priority, passing a series of laws and regulations designed to restrict commercial collection of sensitive data, including personal information, and keep it within the country’s borders. At the same time, the government has itself continued to collect vast amounts of data through a nationwide digital surveillance apparatus to exert tighter control over Chinese society.

That the information was leaked from a government agency—and now has an unknown number of copies circulating outside of the country’s borders—could undermine Beijing’s argument that such a system protects national security, some China tech-policy experts say

“It’s unclear who holds who accountable,” Kendra Schaefer, the head of tech-policy research at Trivium China, a Beijing-based strategic advisory consulting firm, wrote on Twitter in response to the leak. She said it is typically the Ministry of Public Security, which oversees local police agencies such as the Shanghai police, that is responsible for cybercrime investigations.

The Chinese government hasn’t commented on the data leak, and references to it on Chinese social media are quickly being scrubbed.

Some Chinese-speaking users of Twitter, including the chief executive of cryptocurrency exchange Binance, speculated that the leak stemmed from a 2020 technical blog post published by a user on CSDN, a Chinese developer forum similar to Github, that appeared to inadvertently include the access credentials to a Shanghai police server.

Mr. Troia and Mr. Diachenko said the database, based on its configuration, in fact didn’t need access credentials at all, making that theory unlikely. The fault was with the person who set up the dashboard, they said.

Write to Karen Hao at karen.hao@wsj.com and Rachel Liang at rachel.liang@wsj.com

Copyright ©2022 Dow Jones & Company, Inc. All Rights Reserved. 87990cbe856818d5eddac44c7b1cdeb8 Appeared in the July 7, 2022, print edition as 'Police Database in China Was Open Online for Over a Year.'

中國警察數據庫在線開放一年多，導致洩密 網絡安全專家表示，錯誤導致近 10 億人的記錄被盜，導致 20 萬美元的贖金票據 經過 Karen Hao跟隨 在香港和 梁瑞秋跟隨 在新加坡 美國東部時間 2022 年 7 月 6 日上午 11:05 打印 文本 今年早些時候發現該安全漏洞的網絡安全專家表示，這可能是歷史上最大的個人數據盜竊案之一——也是中國已知最大的網絡安全漏洞——因為一個常見的漏洞導致數據可以在互聯網上被盜取.

據網絡安全專家稱，上海警方的記錄——包含近 10 億中國公民的姓名、政府身份證號碼、電話號碼和事件報告——被安全存儲。但他們表示，用於管理和訪問數據的儀表板是在公共網址上設置的，並且在沒有密碼的情況下保持打開狀態，這使得任何具有相對基本技術知識的人都可以進入並複製或竊取大量信息。

暗網情報公司 Shadowbyte 的創始人 Vinny Troia 說：“他們會讓這麼多數據暴露在外，這簡直是瘋了。”該公司在網絡上掃描不安全的數據庫，並在 1 月份找到了上海警方的數據庫。

洩露的上海警方記錄包含近 10 億中國公民的姓名、身份證信息和事件報告。 照片： 亞歷克斯·普拉韋夫斯基/SHUTTERSTOCK 網絡安全研究機構的負責人鮑勃·迪亞琴科（Bob Diachenko）表示，從 2021 年 4 月到上個月中旬，該數據庫暴露了一年多，當時它的數據突然被清除乾淨，取而代之的是一張贖金單，供上海警方發現。公司 SecurityDiscovery，該公司今年早些時候通過定期網絡掃描同樣發現了數據庫，後來發現了筆記。

根據迪亞琴科先​​生提供的屏幕截圖，勒索信上寫著“your_data_is_safe”。“contact_for_your_data...recovery10btc”，這意味著數據將返回 10 個比特幣，大約 200,000 美元。

贖金金額與一名匿名用戶上週四開始在在線網絡犯罪論壇上索要的價格相匹配，以換取訪問該用戶聲稱包含從上海國家警察數據庫竊取的數十億中國公民信息記錄的數據庫。

該帖子於週末開始在社交媒體上流傳，令網絡安全專家感到震驚，不僅因為洩密的規模，還因為政府數據庫中包含的信息的敏感性。

上海市政府和中國互聯網監管機構中國國家互聯網信息辦公室沒有回應置評請求。

網絡安全專家拼湊了數據庫真實性的新證據，以及如此多私人信息可能落入網絡犯罪分子手中的細節。

他們說，儀表板就像一扇通往數據保險庫的大門，即使在所有數據都丟失之後，它也沒有關閉，直到該漏洞開始引起公眾的廣泛關注。根據 Troia 先生的說法，竊取數據的人很可能是兜售數據的同一實體。

“很常見的是，如果贖金受害者不支付贖金，那麼他們會嘗試在網上出售數據，”他說。

無法確定該數據庫是偶然還是故意公開訪問的，也許是為了更容易在少數人之間共享數據。此類漏洞很常見，Troia 先生和 Diachenko 先生說，儘管他們都表示他們對發現如此規模的不安全數據庫感到震驚。

兩人都表示，他們還證實了匿名洩密者的說法，即其中包含超過 23 TB 的數據，涵蓋多達 10 億人。他們說，一個名為“person_address_label_info_master”的文件——其中包含人們的姓名、生日、地址、政府身份證和身份證照片——運行近 9.7 億行，這表明它包含了同樣多的人的詳細信息，假設沒有重複條目。

該文件標記有犯罪記錄的個人，包括違反交通規則的人、被視為逃犯的人和被指控犯有強姦或殺人罪的人。它還包括一個“應該被密切監視的人”的標籤，這個名稱經常在中國政府監控系統中用於表示被視為對社會秩序構成威脅的人。

一份洩露的上海警方檔案包括從交通違規到謀殺指控的記錄。 照片： 亞歷克斯·普拉韋夫斯基/SHUTTERSTOCK 數據洩露凸顯了一些政策研究人員所描述的中國信息安全方法的核心矛盾。

近年來，北京已發出信號，將數據安全和隱私放在首位，通過了一系列旨在限製商業收集包括個人信息在內的敏感數據並將其保存在中國境內的法律法規。與此同時，政府自身也在繼續通過全國性的數字監控設備收集大量數據，以加強對中國社會的控制。

一些中國技術政策專家說，這些信息是從一個政府機構洩露的——現在在中國境外流通的副本數量不詳——可能會削弱北京關於這種系統保護國家安全的論點

“目前尚不清楚誰對誰負責，”總部位於北京的戰略諮詢公司 Trivium China 的技術政策研究負責人 Kendra Schaefer 在 Twitter 上回應洩密事件時寫道。她說，負責網絡犯罪調查的通常是公安部，該部負責監督上海警方等地方警察機構。

中國政府尚未對數據洩露發表評論，中國社交媒體上對此事的提及也很快被刪除。

包括加密貨幣交易所 Binance 首席執行官在內的一些 Twitter 中文用戶推測，洩漏源於 2020 年用戶在類似於 Github 的中國開發者論壇 CSDN 上發布的一篇技術博客文章，該文章似乎無意中包含了訪問權限上海警方服務器的憑據。

Troia 先生和 Diachenko 先生說，根據其配置，該數據庫實際上根本不需要訪問憑證，因此這種理論不太可能。他們說，問題出在設置儀表板的人身上。

寫信給Karen Hao karen.hao@wsj.com和Rachel Liang rachel.liang@wsj.com

版權所有 ©2022 道瓊斯公司。保留所有權利。87990cbe856818d5eddac44c7b1cdeb8 出現在 2022 年 7 月 7 日的印刷版中，名稱為“中國警察數據庫在線開放一年多”。

break this up into two sentences- one for seclusion and one for restraint definitions.

The cumulative effect of one's perception of Sturgeon's law may be a driving force underlying imposter syndrome.

While one see's the entirety of their own creation process and realizes that only a small fraction of it is truly useful, it's much harder seeing only the finished product of others. The impression one is left with by availability heuristic is that there are thousands of geniuses in the world with excellent, refined products or ideas while one's own contribution is miniscule in comparison.

Contrast this with Matt Ridley's broad perspective in The Rational Optimist which shows the power of cumulative breeding and evolution of ideas. One person can make their own stone hand axe, but no one person can make their own toaster oven or computer mouse alone.

Link to: - lone genius myth (eg. Einstein's special relativity did not spring fully formed from the head of Zeus, there was a long train of work and thought which we don't see the context of)

Note: This rebuttal was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

Reply to the reviewers

General response to the reviewer

We thank all reviewers for their constructive comments on our manuscript. We were very pleased to see that the reviewers found our study ‘…represent new insight in the field’ (rev#1) and ‘…contains important and exciting novel findings’ (rev#2), and ‘…gives a more detailed perspective on how Src proteins (Src42A in Drosophila) control epithelial stability and the contraction of specific surfaces of epithelial cells’ (rev#3). The reviewers raised a number of specific points that we partially addressed already in a preliminary revision of the manuscript. Some more points will require some additional experiments that we will incorporate in a fully revised version of the manuscript.

Reviewer #1

(Evidence, reproducibility and clarity (Required)): Highest priority: 1) The Src42A knockdown and germline clone experiments both cause defects in cellularization (Fig. 2B and 9A), which could result in differences in the state of the blastoderm epithelium (cell size, cell number, structural integrity, organization, etc.) between the experimental and control conditions. In addition, Src42A knockdown appears to affect the size and shape of the egg (Fig. 9A and 9C). The manuscript would be strengthened if the authors included data to demonstrate that the initial structure of the epithelium is mostly normal (quantifications of cell size, number, etc.) in the Src42A RNAi condition, as this would bolster the argument that germband extension, rather than due to indirect effects resulting from the cellularization defects. The authors may have relevant data to do this on-hand, for example using data associated with figures 1, 3, 6, and 9.

Response:

The cellularization phenotype of src42A knockdown embryos has a penetrance of about 50% and exhibits a variable expressivity. We attempted to characterize this phenotype in detail, but failed to identify any dramatic differences in cellularization of the src42A knockdown embryos compared to wild type. The localization of E-cadherin, in turn is not affected, but occasionally, nuclei are dropping out of the blastoderm before cellularization is accomplished. This can result in patches of irregular cellularization, but the blastoderm epithelium in stage 6 embryos did not display major defects in overall structure. We will present additional data on the cellularization phenotypes in the fully revised manuscript. As the referee suggested, we will analyze our data to determine potential effects on the cell size, cell number and overall organization of the blastoderm before germband extension. We plan to present these data as an additional Suppl. Mat. Figure in the full revision.

Lower priority:

5) Figure 8 - in my opinion, using a FRAP or photoconversion approach would be a more convincing demonstration of differences in E-cadherin residency times / turnover rate than time-lapse imaging of E-cadherin:GFP alone. Authors should decide whether this improvement is worth the investment.

Response:

We thank the reviewer for this comment. While we believe that the data presented in Fig. 8 demonstrates a significant difference in the E-cadherin residence time based on E-cadherin-GFP fluorescence intensity, we agree with the referee that FRAP analyses would provide additional evidence to support our conclusion. For the full revision, we will therefore attempt to perform FRAP-experiments on src42A knockdown embryos expressing E-cadherin-GFP and compare the recovery time to the wild type.

Reviewer #1 (Significance (Required)):

The manuscript by Backer et al. examines the function of Src42A in germband extension during Drosophila gastrulation. Prior studies in the field have shown that Src family kinases play an important role in the early embryo, including cellularization (Thomas and Wieschaus 2004), anterior midgut differentiation (Desprat et al. 2008), and germband extension (Sun et al. 2017; Tamada et al. 2021). In this study, the authors showed that Src42A was enriched at adherens junctions and was moderately enriched along junctions with myosin-II. They then showed that maternal Src42A depletion exhibits phenotypes, starting with cellularization and including a defect in germband extension. The authors focus on defects in germband extension and found that Src42A was required for timely rearrangement of junctions and that the Src42A RNAi phenotype is enhanced by Abl RNAi. Finally the authors show that E-cadherin turnover is affect by Src42A depletion.

Overall, this study provided a higher resolution description of how Src42A regulates the behavior of junctions during germband extension. I thought the authors conclusions were well supported by the data and represent new insight in the field.

Reviewer #2 (Evidence, reproducibility and clarity (Required)):

Summary: Chandran et al. investigate the role of Src42A in axis elongation during Drosophila gastrulation. Using maternal RNAi and CRISPR/Cas9-induced germline mosaics, they revealed that Src42A is required to contract junctions at anterior/posterior cell interfaces during cell intercalations. Using time-lapse imaging and image analysis, they further revealed the role of Src42A in E-Cad dynamics at cell junctions during this process.

By analyzing double knockdown embryos for Src42A and Abl, they further showed that Src42A might act in parallel to Abl kinase in regulating cell intercalations. The authors proposed that Src42A is involved in two processes, one affecting tension generated by myosin II and the other acting as a signaling factor at tricellular junctions in controlling E-Cad residence time. Overall, the data are clear and nicely quantified. However, some data do not convincingly support the conclusion, and statistical analyses are missing for an experiment or two. Methods for several quantifications also need improvement in writing. Also, several figures (Figures 6-8) do not match the citation in the text and need to be corrected.

Page and line numbers were not indicated in the manuscript. For my comments, I numbered pages starting from the title page (Title, page 1; Abstract, page 2, Introduction, pages 3-6; Results, pages 7-14; Discussion, pages 15-18; M&M, 19-23; Figure legends, 28-30) and restarted line numbers for each page. For Figures 6-8 that do not match the citation in the text, I still managed to look at the potentially right panels. All the figure numbers I mention here are as cited in the text. My detailed comments are listed below.

Response:

We apologize for the lack of organization of the manuscript and the figure numbering. In the revised version we have added page numbers, line numbers and we corrected the figure numbers.

Major comments: 1. b-Cat/E-Cad signals at the D/V and A/P junctions in Src42Ai (Figs. 5-6). These data are critical for their major conclusion and should be demonstrated more convincingly.

In Fig. 5A, the authors said, "When the AP border was cut, the detached tAJs moved slower in Src42Ai embryos compared to control (Fig. 5A)". However, even control tAJs do not seem to move that much in the top panels, and I found the images not very convincing.

Response:

We thank the referee for commenting on the lack of clarity in the presentation of the data. The overall movement within the first 10 seconds after the laser cut (determined by movement of adjacent D/V tAJs from each other) was about 2 µm in the wildtype, while in the mutant it was 1 µm. Despite this 50% difference, it may be difficult to appreciate this difference from looking at Fig. 5A in our original submission. The yellow lines in Fig 5A only showed the region of the cut, but did not indicate the movement of the tAJ from each other, which may have led to a distraction from the actual movement. We will change the annotation and the marks within the figure to visualize the movement much more clearly in the full revision. In the fully revised manuscript, we will also add movies from the experiments including marks of the tricellular junctions to follow the displacement as part of the Supplemental Material.

Based on the genetic interaction between Src42A and Abl using RNAi (Fig. 7), the authors argue that Src42A and Abl may act in parallel. However, the efficiency of Abl RNAi has not been tested. It can be done by RT-PCR or Abl antibody staining. Also, the effect of Abl RNAi alone on germband extension should be tested and compared with Src42A & Abl double RNAi embryos. I expect the experiments can be done within a few weeks without difficulty.

Response:

We agree with the referee that it is important to determine the level of depletion in Abl RNAi embryos in order to interpret the genetic relationship between Abl and Src42A. In the full revision of the manuscript, we will follow the advice of the referee and analyze the knockdown, preferably by antibody labeling with an anti-Abl antibody. We will also generate single knockdowns of abl in embryos and determine their effect on germband extension compared to wildtype and src42/abl double knockdown.

Minor comments:

Fig. 2 - Fig. 2B: Higher magnification images of the defective cytoplasm can be shown as insets.

Response:

We will add some higher magnification images of the cellularization phenotype in the full revision of the manuscript. In addition, as mentioned in the response to reviewer #1, we will provide a more detailed analysis of the cellularization in src42Ai embryos in the fully revised manuscript.

• Fig. 2E: A simple quantification of the penetrance of cuticle defects in Src42A mutants and RNAi will be helpful, as shown in Fig. S3.

Response:

In the full revision, we will add the quantification of the occurrence of the different classes of cuticle phenotypes.

Fig. 9 - Fig. 9A: Magnified views of the cytoplasmic clearing can be added as insets.

Response: As described in our response to the comments made by referee #1, we will add a more detailed analysis of the cellularization phenotype in the full revision.

Page 14, lines 9-10: More explicit description of the phenotype rather than just "stronger compared to Src42Ai" will be helpful.

Response:

In the full revision, we will add a more detailed description of the phenotype and re-analyze and present data on the hatching rate, stage of lethality and cuticle phenotypes.

Reviewer #2 (Significance (Required)): This work revealed the role of Src42A in regulating germband extension. A previous study suggested the roles of Src42A and Src64 in this developmental process using a partial loss of both proteins (Tamada et al., 2021). Using different approaches, the authors demonstrated a role of Src42A in regulating E-Cad dynamic at cell junctions during Drosophila axis elongation. Most of the analyses were done with maternal knockdown using RNAi, but they successfully generated germline clones for the first time and confirmed the RNAi phenotypes. Overall, this work contains important and exciting novel findings. This work will be of general interest to cell and developmental biologists, particularly researchers studying epithelial morphogenesis and junctional dynamics. I have expertise in Drosophila genetics, epithelial morphogenesis, imaging, and quantitative image analysis.

Reviewer #3 (Evidence, reproducibility and clarity (Required)):

Chandran et al. report on the function of Src42A during cell intercalation in the early Drosophila gastrula. They create a Src42A-specific antibody (there are two Src genes in the fly genome) and examine the localization of Src42A and observe a planar-polarized distribution at cell interfaces. They then measure cell-contractile dynamics and show that T1 contraction is slower after Src42A disruption. The authors then argue that Src42A functions in a parallel pathway to the Abl protein, and that E-cadherin dynamics (turnover) is altered in Src42A disrupted embryos. Src function at these stages has been studied previously (though not to the degree that this study does), and in some respects the manuscript feels a little preliminary (please label figures with figure number!), but after editing this should be a polished study that merits publication in a developmentally-focused journal.

1) Does the argument that Src42A has two functions fully make sense? Myosin II function is known to affect E-cadherin stability (and vice versa), so it seems that Src42A could affect both MyoII and Ecad by either decreasing Myosin II function/engagement at junctions or by destabilizing Ecad.

Response:

We thank the referee for raising an important point that we may not have discussed appropriately in our initial submission. We agree that the reciprocal relationship between actomyosin and E-cadherin might not be reflected equivocally in our manuscript. As the referee points out, Src42A could affect both MyoII planar localization and E-cadherin dynamics through the same pathway. Previous studies showed that Src is involved in translating the planar polarized distribution of the Toll-2 receptor by recruiting Pi3-Kinase activity to the Toll-2 receptor complex resulting in planar polarized distribution of MyoII at the A/P interfaces. These data, however do not address the possibility that a well-known Src target, the E-cadherin/ß-Catenin complex, which is extensively remodeled in germband extension contributes to the delay in germband extension. The observed defects in both studies can be attributed to both a defect in abnormal planar polarization of MyoII and the abnormal dynamics of the E-cadherin/ß-catenin complex. In either of these cases, we suggest that Src42A phosphorylates distinct substrates, the Toll-2 intracellular domain in the MyoII planar polarity pathway and the E-cad/ß-Cat complex controlling E-cad dynamics. Given the relationship between MyoII and E-cadherin, however, it is not possible to decide whether these two effects are independent functions of Src42A or are consequences of each other. Since we cannot resolve a possible epistatic relationship between these potential two activities of Src42A, we decided to extend the discussion on this topic by taking both possible scenarios into account and discussing them appropriately. We will add this discussion in the full revision of the manuscript.

) One obvious question that arises is the nature of cleavage defects that are mentioned that happen previously to intercalation. For example, is E-cad normal prior to intercalation initiating? How specific are the observed defects to GBE?

Response:

please see response to referee #1

3) Pg. 10, "the shrinking junction along the AP axis strongly reduces its length with an average of 1.25 minute" - what is this measurement? How much is "strongly"?

Response:

We thank the referee for pointing out our inappropriate qualitative statement of the experimental data, which was indeed misleading. The measurement of the shrinking junction was based upon the time it takes for the AP interface junction between two adjacent vertices on the DV axis to shrink into a single 4-cell vertex. The time for this contraction was on average 1 minute 25 seconds. The data in Fig.4 A’,C show that after 2 minutes in the control embryo 100% of the observed AP junctions have collapsed and the extension of the new DV junction along AP axis has begun. At the same timepoint of 2 minutes in the src42A knockdown, we show in Fig. 4B’,D that the shrinking of the AP junction interface has still not been completed in 60% of the cases.

In the full revision, we will remove the qualitative statement and replace it with a correct description of the measurements taken and will refer to the data described in Fig. 4 A-D.

4) Also pg. 10, "the AP junction was not markedly reduced after 1 minute" - what is the criteria for this statement? X%? 1 minute is very specific, it feels like how much of a reduction/non-reduction should also be specific.

Response:

please see response to point 3.

Reviewer #3 (Significance (Required)):

This study gives a more detailed perspective on how Src proteins (Src42A in Drosophila) control epithelial stability and the contraction of specific surfaces of epithelial cells.

Description of the revisions that have already been incorporated in the transferred manuscript

Reviewer #2 and #3 noted that the manuscript was somewhat unorganized with regard to lacking the numbering of pages, lines and figures. We also noted that in the submission process the figures were not presented in the correct order. In the preliminary revision of the manuscript, we fixed these problems to facilitate the evaluation of our transferred manuscript by editorial boards.

In addition, we also addressed issues that the referees mentioned by editing the text according to their comments. We also addressed problems regarding the presentation of the figures and statistical analyses of the data. The following changes were made:

1. We added page numbers and line numbers.
2. We added figure numbers to the figure panels.
3. We corrected ordering of figures in the transferred manuscript.
4. We addressed the following comments by statistical analyses, editing the text and the figures:

   Regarding comments from Reviewer #1:

Highest Priority:

2) There is a discrepancy in the staging of embryos used between some of the analyses, which make it hard to interpret some of the data. For example, characterization of the knockdowns in Fig. 1A and B are based on stages 10 and 15, whereas the majority of the paper is focused on earlier stages 6 - 8 during germband extension (e.g., Fig. 1D). The analysis for Fig. 1B would be more meaningful if it was done on the same stages used for subsequent phenotypic analysis so they can be directly compared.

Response:

We thank the referee for pointing out an apparent misunderstanding caused by the description of Fig. 1A,B. The data presented in Fig.1A and 1B do not show RNAi knockdown experiments, but show a comparison between embryos that are heterozygous or homozygous for the loss-of-function allele src42A26-1. These data were intended to demonstrate that zygotic mutants still maintain levels of maternal Src42A protein up until late stages of development. Data for embryos at an earlier stage (stage 5) were shown in the Supplementary Fig. S1E, where no difference in protein levels of Src42A can be observed between heterozygous and homozygous zygotic src42A26-1 embryos.

At the beginning of the results sections 1 and 2 of the preliminary revised manuscript, we added a sentence to address the referee’s concern that earlier stages exhibit no difference in protein levels and will refer to Fig. S1E. We also more explicitly spelled that out that the experiment (referring to Fig.1A,B and S1) was intended to look at zygotic mutants and to demonstrate that our novel Src42A antibody was able to detect the reduction of maternal Src42A protein in mid- to late-stage homozygous zygotic embryos.

3) There is incongruence between figures in terms of which junctional pools (bAJs vs. tAJs) of beta-catenin and E-cadherin are quantified that makes it difficult to draw comparisons between analyses. For example, pTyr levels are examined for both bAJs and tAJs in Figure 3, however, only tAJs are considered in Fig. 8. Similarly, in some cases planar cell polarity is considered (e.g., comparison of levels at AP vs DV bAJs in Fig. 6 and 9), and in other cases (e.g. Fig. 8) it is not.

Response:

We thank the referee for commenting on the different readouts for different pools of cell junctions in our experiments. In our study we considered effects on src42A on both, bAJs and tAJs by RNAi knockdown of src42A. We decided to present the data for bAJ and tAJ in separate figures for clarity and structure. For example, the data for the effect of src42A knockdown on the planar polarized distribution on bAJs of E-cadherin were presented in Fig.6, while the effect on E-cadherin residence time in tAJs were presented in Fig.8. The analysis pTyr levels considered both pools in order to determine whether src42A knockdown leads to an overall reduction of pTyr levels or to a reduction in a specific junctional pool. From our data we conclude that pTyr levels show a similar reduction in both, the bAJ and the tAJ junctions.

In order to address the reviewer’s comment, we have linked the figures more stringently with the results text of the preliminary revision. We only referred to the reduction in PTyr levels in Fig. 3 to point out that both junctional pools are affected by reduced PTyr in src42i embryos. Furthermore, we referred to the individual figure panels when addressing junctional pools and explain the rationale to focus on particular pools (bAJs or tAJ) in the experiments in detail. For Fig. 6 we point out in the preliminary revised manuscript that we focus the analyses on the known planar polarized distribution of beta-catenin and E-Cadherin.

Lower priority: 1) Introduction, 2nd paragraph - The modes of cell behaviors described to drive cell intercalation leaves out another clear example in the literature - Sun et al., 2017 - which describes a basolateral cell protrusion-based mechanism. While the authors cite this paper later, leaving it out when summarizing the state of the field misrepresents the current knowledge of the range of mechanisms responsible.

Response:

We thank the referee for this remark. In the preliminary revision, we have added to the introduction that the cell behaviors associated with germband elongation include apical and basolateral rearrangements of the cells indicating that basolateral protrusions also contribute to the set of mechanisms that drive germ band elongation.

2) 'defective cytoplasm' - this term is confusing, and could perhaps be replaced with 'cellularization defect', or something similar.

Response:

We agree that the term we applied for the cellularization defect may be misleading. The observation, we intended to describe with the term was a defect in the cytoplasmic clearing which occurs in the last syncytial division and the beginning of the cell formation process. We changed the description of this observation according now refer to the defect in the preliminary revised manuscript as ‘cytoplasmic clearing defect’.

3) Tests of statistical significance are not uniformly applied across the figures. For instance, Figures 3G + H indicate statistical significance, but Fig. 3D + E do not. Performing statistical tests throughout the paper, or clearly articulating a rationale when they are not used, would strengthen the manuscript. Specifically, the authors should consider this for Fig. 3D + E, and Fig. 7D + E, to support their arguments that rates of germband extension are different between conditions.

Response:

We agree with the reviewer and have provided statistical analysis for the data displayed in Fig. 3D,E and Fig. 7D,E in the preliminary revision of the manuscript.

4) Page 12 - "We found that Src42A showed a distinct localization at the tAJs (Fig. 1B)": Figure 1B shows a quantification of levels at bAJs, not tAJs.

Response:

In the preliminary version of the revised manuscript, we added a quantification of the localization of Src42A at the tAJs as a part of Suppl Fig. S4. In Fig. S4A-C we show that Src42A is enriched in comparison to the bAJs.

Regarding comments from reviewer #2:

Major Comments:

In Fig. 6A, b-Cat signals look fuzzier and dispersed and have more background signals in the control, compared to the Src42Ai background. Also, b-Cat signals in the control image do not seem to show enrichment at the D/V border, as shown in Tamada et al., 2012.

Response:

We agree with the referee that the image in Fig. 6A for the control is fuzzier and looks dispersed. This is due to the fixation method that we used. In this experiment we did not apply heat fixation, but used formaldehyde fixation in which b-catenin protein, in addition to the junctional pool, is also maintained in the cytoplasm creating the fuzzy cytoplasmic staining. We chose to do this in order to be able to co-immunolabel the embryos with b-catenin and E-cadherin antibodies; the latter staining is not working with the heat fixation applied in the Tamada et al. 2012 paper. Despite the slightly lower quality of the staining, the quantification of the data clearly indicated an effect of src42A knockdown on the planar polarized distribution of E-cad/b-cat complex does show an enrichment. In the preliminary revision added a note to the figure legend to indicate the fact that the fixation procedure was not optimized for b-catenin junctional staining. In the preliminary revision we also added a quantification of live imaging data recording E-cadherin-GFP in wild-type and src42Ai embryos. We present these additional data in Fig. S5 in the preliminary revision of the manuscript. These data are consistent with the results in Fig. 6 from the immunolabeling and support our conclusion that E-cadherin AP/DV ratio is increased in Src42A knockdown embryos.

In Fig. 6B, C, it is not clear how the intensity was measured and how normalization was done. Was the same method used for these quantifications as "Protein levels at bicellular and tricellular AJs" on pages 21-22? Methods should be written more explicitly with enough details.

Response:

We thank the referee for pointing out the lack of detail in explaining how the quantification was done. In the preliminary revision of the manuscript, we extended a paragraph entitled ‘Protein levels at bicellular and tricellular junctions’ in the methods section that will serve this purpose and describe the methods that were applied for each quantification and the method as to how the data were normalized.

Does each sample (experimental repeat) for the D/V border in Fig. 6B match the one right below for the A/P border in Fig. 6C? It should be clearly mentioned in the figure legend. The ratio of the DV intensity to AP intensity will better show the compromised planar polarity of the b-Cat/E-Cad complex.

Response:

We thank the reviewer for pointing out a lack of clarity in our presentation. The experimental repeats for each measurement do indeed match, i.e. the measurement of the DV border matches the same adjacent 4-cell pair in the same embryo and in total 5 distinct embryos were analyzed for each experiment. In the preliminary revision of the manuscript, we explain this detail of the experimental design in the figure legend. In the preliminary revision, we also determined the ratios of DV/AP cell interfaces for b-Cat and E-Cad and added this quantification as panel 6C and 6E for a clearer presentation of the data.

Minor notes: Page 4, missing comma after "For example"

Response: The text was edited accordingly.

Page 4, "inevitable" does not make sense in this context

Response: We eliminated ‘inevitable’ and replaced it with ‘critical’ to better indicate the importance of Canoe protein for germband elongation.

Page 7, lines 6-7 - The localization of Src42A in control should be described in more detail and more clearly here.

Response: In the preliminary revised manuscript, we extended our description of the distribution of Src42A in more detail pointing out its dynamics and differential distribution at distinct plasma membrane domains.

Supplemental Fig S1 - Fig. S1D: Based on the head structure and the segmental grooves, the embryo shown here is close to late stage 13/early stage 14, not stage 15. - Fig S1E: It will be helpful if the predicted protein band and non-specific bands are indicated by arrows/arrowheads in the figure.

Response:

We thank the referee for their careful observation of the embryonic stage. We agree that the embryo was actually a younger stage. In the preliminary revision, we replaced the images with an example of an older stage. We will also add clear annotations as arrows to clearly mark the specific protein bands in Fig. S1E.

Page 7, lines 21-22 - "Src42A was slightly enriched at the AP interface" - To argue that, quantification should be provided.

Response:

We thank the referee for pointing out a qualitative statement that we made with regard to the distribution of Src42A at the AP cell interfaces. In the preliminary revision of the manuscript, we present an additional quantification of the imaging data of Src42A immunolabeling. In Figure S4A-C, we now present a quantification of the enrichment of Src42A at the tricellular junctions. In addition, the new Fig. S4D,E shows a quantification of the planar polarized distribution of Src42A at the AP cell interfaces.

Figure 1 - Fig. 1B: Src42A levels should be compared between control (Src42A/+) and Src42A/Src42A for each stage. It currently shows a comparison between Src42A/Src42A of stages 10 and 15.

Response:

We thank the referee for the comment. As indicated in our response to referee #1, the point of this analysis was to (1) provide evidence for the specificity of our new anti-Src42A antibody and (2) to demonstrate the presence of substantial material contribution of Src42A protein in zygotic mutant. We do not see the advantage to provide a detailed developmental Western-blot analysis, but we provide data in Suppl. Mat Fig S1E showing that the level of Src42A is unimpaired in stage 6 zygotic src42A[26-1] homozygous mutant embryos.

• Fig. 1B: The figure legend says, "dotted line represents mean value and error bars," but there are no dotted lines shown in the figure. Also, what p-value is for ****? It should be mentioned in the figure legend. It also says Src42A levels were normalized against E-Cad intensity here (stages 10 and 15). They have shown that E-Cad levels are affected in Src42A RNAi during gastrulation (Fig. 6). Is E-Cad not affected in Src42A26-1 zygotic mutants at stages 10 and 15?

Response:

We thank the referee for pointing out inaccuracies in the presentation and the description of Fig.1B. In the preliminary revision, we emphasized the marks on the graph and provide p-values throughout. Regarding the E-Cadherin levels: E-cadherin levels were altered in src42A RNAi knockdown embryos, but not in zygotic mutants, even at later developmental stages.

Page 8, line 14 - "Embryos expressing TRiP04138 showed reduced hatching rates with variable penetrance and expressivity depending on the maternal Gal4 driver used (Fig. 2B)" - Fig. 2B doesn't seem to be a right citation for this sentence.

Response:

We agree with the referee and in the preliminary revised manuscript we corrected the reference to the conclusion drawn from Figure 2A’, which does show the relationship of hatching rate to the various maternal Gal4 drivers.

• Fig. 2C: It will be helpful to indicate two other non-specific bands in the figure with arrows/arrowheads with a description in the figure legend.

Response:

In the preliminary revision, we added an arrow to mark the band specific for Src42A and asterisks to mark unspecific bands in Fig 2C.

Page 9, line 9 - This is the first time that the fast and the slow phases of germband extension are mentioned. As these two phases are used to compare the Src42A and Src42A Abl double RNAi phenotypes, they should be introduced and explained better earlier, perhaps in Introduction.

Response:

We thank the referee for pointing out that the two phases of germband extension were not introduced. We added a sentence to introduce and define the distinct phases of extension movements in the preliminary revision.

Fig. 3 - Fig. 3A: It will be helpful to mark the starting and the ending points of germband elongation with different markers (arrows vs. arrowheads or filled vs. empty arrowheads).

Response:

In the preliminary revision, we added distinct markers to indicate the start and endpoints of germband elongation to make this figure easier to read.

• Fig. 3C figure legend: R2 is wrongly mentioned in Fig. 3D, E. Also, R2 (coefficient of determination) needs to be defined either in the figure legend or Materials & Methods.

Response:

We thank the referee for pointing this misleading reference to us. In the preliminary revision we corrected the reference to R2 in Fig,3D,E and will describe the definition of R2 in the figure legend.

• Fig. 3D, E: statistical analysis is missing.

Response:

In the preliminary revision, we included a statistical analysis of the data (see ref #1). We changed the figure to indicate the data sets that were analyzed and added the p-values to the figure legend.

• Fig. 3G and H should be cited in the text.

Response:

In the preliminary revision, we added references to Fig 3G,H in the result section to the annotation of Fig.3F).

• Fig. 3F: It should be mentioned that the heat map is shown for pY20 signals in the figure legend, with an intensity scale bar in the figure.

Response:

In the preliminary revision, we added an intensity scale bar to the figure panel and mentioned the relationship to the PY20 signal.

Fig. 7A: Arrows can be added to mark the delayed germband extension.

Response:

In the preliminary revision, we added arrows to mark the anterior and posterior extent of the germband.

Fig. 8A: It should be mentioned that the heat map is shown for E-Cad signals in the figure legend, with an intensity scale bar in the figure.

Response:

In the preliminary revision, we added an intensity scale to the heat map and mention the relationship to the E-cadherin signal in the figure legend.

Fig. S3G: An arrowhead can be added to the gel image to indicate the band described in the legend.

Response:

In the preliminary revision, we added an arrow to help annotating the Src42A-specific bands on the Western blot.

• Fig. 9B: Arrow/arrowheads can be added to show the absence of the signals in the nurse cells.

Response:

In the preliminary revision, we added markers to help recognizing the reduced signal in the nurse cells and the oocyte.

• Fig. 9C: Indicate the ending point of the germband extension by arrows.

Response: In the preliminary revision, we added arrows to mark the anterior and posterior extent of the germband.

Regarding comments from reviewer #3:

Minor notes: Page 4, missing comma after "For example"

Response: The text was edited accordingly.

Page 4, "inevitable" does not make sense in this context Response:

In the preliminary revision, we eliminated ‘inevitable’ and replaced it with ‘critical’ to better indicate the importance of Canoe protein for germband elongation.

Description of analyses that authors prefer not to carry out

Referee #1 point2 and Referee#2 minor comment figure 1. Both referees suggest that figure 1 AB should include earlier developmental stages according to the stages looked at in the RNAi knockdown experiment.

Response:

The referees’ comments are likely based on a misunderstanding. The data that the reviewer are referring to present analyses of the zygotic phenotype of embryos homozygous for the src42A26-1 loss of function allele. They are not related to the maternal RNAi knockdown experiments, but were meant to demonstrate the existence and extent of a maternal pool of Src42A protein, that persists even to late stages in development. The maternal knockdown mutants are analyzed in detail at the appropriate stages in Fig. 2.

As described in our response above, we don’t feel that a detailed developmental stage Western analysis of wildtype and src42A26-1 embryos would provide significant additional insights. As mentioned in our response above, data for an earlier developmental stage (before germband elongation, as requested by the referees, were provided in Suppl. Fig. S1E.

Referee #1 Point 6) Figure 8E - showing images of multiple tAJs, rather than z-slices of a single vertex, would better support the claim here, as the assertion is that Src42a levels are different between control and sdk RNAi conditions, and not that it varies in the z-dimension.

Response:

The image series of Fig. 8E shows one representative example of multiple tAJs that have been imaged for this experiment (n=6 for wild type and n=10 for sdk RNAi). We think that the presentation of Z-slices for this experiment is important as the protein distribution needs to be considered for a larger area along the apical-lateral cell interface. In addition the quantification of the data for multiple tAJs was presented in Fig. 8F,G as a graph. We would therefore rather not change this figure in the revised manuscript.

Referee #3 suggests that anti MyoII staining should accompany the analysis of tension measurements in the germband.

As this analysis has already been performed by Tamada et al. 2021, we decided not to reproduce these data, but rather extend the analysis towards tension measurements, which support the findings by Tamada et al. 2021 on a functional level. We do not see the added value of adding MyoII labeling.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

Referee #2

Evidence, reproducibility and clarity

Summary:

Chandran et al. investigate the role of Src42A in axis elongation during Drosophila gastrulation. Using maternal RNAi and CRISPR/Cas9-induced germline mosaics, they revealed that Src42A is required to contract junctions at anterior/posterior cell interfaces during cell intercalations. Using time-lapse imaging and image analysis, they further revealed the role of Src42A in E-Cad dynamics at cell junctions during this process.

By analyzing double knockdown embryos for Src42A and Abl, they further showed that Src42A might act in parallel to Abl kinase in regulating cell intercalations. The authors proposed that Src42A is involved in two processes, one affecting tension generated by myosin II and the other acting as a signaling factor at tricellular junctions in controlling E-Cad residence time. Overall, the data are clear and nicely quantified. However, some data do not convincingly support the conclusion, and statistical analyses are missing for an experiment or two. Methods for several quantifications also need improvement in writing. Also, several figures (Figures 6-8) do not match the citation in the text and need to be corrected.

Page and line numbers were not indicated in the manuscript. For my comments, I numbered pages starting from the title page (Title, page 1; Abstract, page 2, Introduction, pages 3-6; Results, pages 7-14; Discussion, pages 15-18; M&M, 19-23; Figure legends, 28-30) and restarted line numbers for each page. For Figures 6-8 that do not match the citation in the text, I still managed to look at the potentially right panels. All the figure numbers I mention here are as cited in the text. My detailed comments are listed below.

Major comments:

1. b-Cat/E-Cad signals at the D/V and A/P junctions in Src42Ai (Figs. 5-6). These data are critical for their major conclusion and should be demonstrated more convincingly.

In Fig. 5A, the authors said, "When the AP border was cut, the detached tAJs moved slower in Src42Ai embryos compared to control (Fig. 5A)". However, even control tAJs do not seem to move that much in the top panels, and I found the images not very convincing.

In Fig. 6A, b-Cat signals look fuzzier and dispersed and have more background signals in the control, compared to the Src42Ai background. Also, b-Cat signals in the control image do not seem to show enrichment at the D/V border, as shown in Tamada et al., 2012.

In Fig. 6B, C, it is not clear how the intensity was measured and how normalization was done. Was the same method used for these quantifications as "Protein levels at bicellular and tricellular AJs" on pages 21-22? Methods should be written more explicitly with enough details.

Does each sample (experimental repeat) for the D/V border in Fig. 6B match the one right below for the A/P border in Fig. 6C? It should be clearly mentioned in the figure legend. The ratio of the DV intensity to AP intensity will better show the compromised planar polarity of the b-Cat/E-Cad complex. 2. Based on the genetic interaction between Src42A and Abl using RNAi (Fig. 7), the authors argue that Src42A and Abl may act in parallel. However, the efficiency of Abl RNAi has not been tested. It can be done by RT-PCR or Abl antibody staining. Also, the effect of Abl RNAi alone on germband extension should be tested and compared with Src42A & Abl double RNAi embryos. I expect the experiments can be done within a few weeks without difficulty.

Minor comments:

Page 2, line 14 - The abbreviation for tAJs was not introduced before.

Page 7, line 6 - A reference should be cited for the Src42A26-1 allele.

Figure 1 - Fig. 1B: Src42A levels should be compared between control (Src42A/+) and Src42A/Src42A for each stage. It currently shows a comparison between Src42A/Src42A of stages 10 and 15. - Fig. 1B: The figure legend says, "dotted line represents mean value and error bars," but there are no dotted lines shown in the figure. Also, what p-value is for ****? It should be mentioned in the figure legend. It also says Src42A levels were normalized against E-Cad intensity here (stages 10 and 15). They have shown that E-Cad levels are affected in Src42A RNAi during gastrulation (Fig. 6). Is E-Cad not affected in Src42A26-1 zygotic mutants at stages 10 and 15?

Page 7, lines 6-7 - The localization of Src42A in control should be described in more detail and more clearly here.

Supplemental Fig S1

• Fig. S1D: Based on the head structure and the segmental grooves, the embryo shown here is close to late stage 13/early stage 14, not stage 15.
• Fig S1E: It will be helpful if the predicted protein band and non-specific bands are indicated by arrows/arrowheads in the figure.

Page 7, lines 21-22

• "Src42A was slightly enriched at the AP interface" - To argue that, quantification should be provided.

Page 8, line 14

• "Embryos expressing TRiP04138 showed reduced hatching rates with variable penetrance and expressivity depending on the maternal Gal4 driver used (Fig. 2B)" - Fig. 2B doesn't seem to be a right citation for this sentence.

Fig. 2

• Fig. 2B: Higher magnification images of the defective cytoplasm can be shown as insets.
• Fig. 2C: It will be helpful to indicate two other non-specific bands in the figure with arrows/arrowheads with a description in the figure legend.
• Fig. 2E: A simple quantification of the penetrance of cuticle defects in Src42A mutants and RNAi will be helpful, as shown in Fig. S3.

Page 9, line 9

• This is the first time that the fast and the slow phases of germband extension are mentioned. As these two phases are used to compare the Src42A and Src42A Abl double RNAi phenotypes, they should be introduced and explained better earlier, perhaps in Introduction.

Fig. 3

• Fig. 3A: It will be helpful to mark the starting and the ending points of germband elongation with different markers (arrows vs. arrowheads or filled vs. empty arrowheads).
• Fig. 3G and H should be cited in the text.
• Fig. 3C figure legend: R2 is wrongly mentioned in Fig. 3D, E. Also, R2 (coefficient of determination) needs to be defined either in the figure legend or Materials & Methods.
• Fig. 3D, E: statistical analysis is missing.
• Fig. 3F: It should be mentioned that the heat map is shown for pY20 signals in the figure legend, with an intensity scale bar in the figure.

Fig. 7A: Arrows can be added to mark the delayed germband extension.

Fig. 8A: It should be mentioned that the heat map is shown for E-Cad signals in the figure legend, with an intensity scale bar in the figure.

Fig. S3G: An arrowhead can be added to the gel image to indicate the band described in the legend.

Fig. 9

• Fig. 9A: Magnified views of the cytoplasmic clearing can be added as insets.
• Fig. 9B: Arrow/arrowheads can be added to show the absence of the signals in the nurse cells.
• Fig. 9C: Indicate the ending point of the germband extension by arrows.

Page 14, lines 9-10: More explicit description of the phenotype rather than just "stronger compared to Src42Ai" will be helpful.

Significance

This work revealed the role of Src42A in regulating germband extension. A previous study suggested the roles of Src42A and Src64 in this developmental process using a partial loss of both proteins (Tamada et al., 2021). Using different approaches, the authors demonstrated a role of Src42A in regulating E-Cad dynamic at cell junctions during Drosophila axis elongation. Most of the analyses were done with maternal knockdown using RNAi, but they successfully generated germline clones for the first time and confirmed the RNAi phenotypes. Overall, this work contains important and exciting novel findings.

This work will be of general interest to cell and developmental biologists, particularly researchers studying epithelial morphogenesis and junctional dynamics.

I have expertise in Drosophila genetics, epithelial morphogenesis, imaging, and quantitative image analysis.

star fish leg grows a new start fish

Note: This rebuttal was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

Reply to the reviewers

Manuscript number: RC-2021-01016

Corresponding author(s): Dennis Klug

1. General Statements [optional]

Dear editor, dear reviewers,

thank you very much for the quick review of our manuscript as well as for the constructive criticism and the interesting discussion of our results. Reading the comments, we realized that we may have put too much emphasis on the in vivo microscopy of sporozoites and their interaction with the salivary gland. We believe that the generated mosquito lines can be used to address different scientific questions, the in vivo microscopy of host-pathogen interactions being only one of them. Because of this imbalance, and to address some of the reviewers' comments, we have partially rewritten the manuscript (particularly the introduction). At the same time, we have implemented additional data on the inducibility of the promoters used, as well as on the functionality of hGrx1-roGFP2 in the salivary glands. Furthermore, we created an additional figure to better present the expression patterns of trio and saglin promoters within the median lobe, and we expanded the section on in vivo microscopy of sporozoites. We hope that these results further highlight the significance of our study. Accordingly, we have also changed the title of the manuscript to „A toolbox of engineered mosquito lines to study salivary gland biology and malaria transmission” to indicate the broad applicability of the generated mosquito lines and we have included an additional co-author, Raquel Mela-Lopez, who conducted the redox analysis. We hope that these changes will adequately answer the questions of the reviewers and address any concerns they may have had. We look forward to hearing from you.

With our kind regards,

Dennis Klug

Katharina Arnold

Raquel Mela-Lopez

Eric Marois

Stéphanie Blandin

2. Point-by-point description of the revisions

Reviewer #1 (Evidence, reproducibility and clarity (Required)):

**Summary**

This manuscript reports the generation and characterization of transgenic lines in the African malaria mosquito Anopheles coluzzii that express fluorescent proteins in the salivary glands, and their potential use for in vivo imaging of Plasmodium sporozoites. The authors tested three salivary gland-specific promoters from the genes encoding anopheline antiplatelet protein (AAPP), the triple functional domain protein (TRIO) and saglin (SAG), to drive expression of DsRed and roGFP2 fluorescent reporters. The authors also generated a SAG knockout line where SAG open reading frame was replaced by GFP. The reporter expression pattern revealed lobe-specific activity of the promoters within the salivary glands, restricted either to the distal lobes (aapp) or the middle lobe (trio and sag). One of the lines, expressing hGrx1-roGFP2 under control of aapp promoter, displayed abnormal morphology of the salivary glands, while other lines looked normal. The data show that expression of fluorescent reporters does not impair Plasmodium berghei development in the mosquito, with oocyst densities and salivary gland sporozoite numbers not different from wild type mosquitoes. Salivary gland reporter lines were crossed with a pigmentation deficient yellow(-) mosquito line to provide proof of concept of in vivo imaging of GFP-expressing P. berghei sporozoites in live infected mosquitoes.

**Major comments**

Overall the manuscript is very well written with a clear narrative. The data are very well presented. The generation of the transgenic mosquito lines is elegant and state-of-the art, and the new reporter lines are thoroughly characterized.

This is a nice piece of work that is suitable for publication, although the in vivo imaging of sporozoites is somewhat preliminary and would benefit from additional experiments to increase the study impact.

We would like to thank the reviewer for his/her appreciation of our manuscript. In the revised version, we have included additional experiments on in vivo imaging of sporozoites, which allowed us to quantify moving and non-moving sporozoites imaged under the cuticle of live mosquitoes. Although this is still a proof of concept, we believe that these new data provide novel interesting data and will better illustrate potential applications.

The reporter mosquito lines express fluorescent salivary gland lobes, yet the authors only provide imaging of parasites outside the glands. It would be relevant to provide images of the parasite inside the fluorescent glands.

We have now included images showing sporozoites inside the salivary glands in vivo in Figure 8C and discuss possible ways to further improve resolution and efficiency of the imaging procedure in lines 563-586.

The advantage of the pigmentation-deficient line over simple reporter lines is not clear, essentially due to the background GFP fluorescent in figure 5C. Imaging of GFP-expressing parasites should be performed in mosquitoes after excision of the GFP cassette under control of the 3xP3 promoter. This would probably allow to document the value of the reporter lines more convincingly.

Indeed, by incorporating two Lox sites in the transgenesis cassette, we designed the yellow(-)KI line to permit removal of the fluorescent cassette and completely exclude expression of the transgenesis reporter EGFP. Still, EGFP expression in the yellow(-)KI adults is restricted to the eye and ovary, as we show now in Figure 7 supplement 1D. In contrast, no EGFP fluorescence was observed in the thorax area (Figure 7 supplement 1D). Therefore, we believe that the benefit of removing the fluorescence cassette for this study is limited. Moreover, the generation of such a line would take at least 3-4 months before experiments could be performed. Nevertheless, we agree with the reviewer that removal of the fluorescence cassette would be instrumental for follow-up studies. To draw the reader's attention to this issue, we now discuss background fluorescence in lines 378-387.

Along the same line, it is unclear if the DsRed spillover signal in the GFP channel is inherent to the high expression level or to a non-optimal microscope setting. This is a limitation for the use of the reporter lines to image GFP-expressing parasites.

We have discussed this problem with the head of the imaging platform at our institute, and we believe that it is not a problem that occurs due to incorrect settings. Rather, it seems to be due to the significant expression differences of the two fluorescence reporters used. We agree with the reviewer that this is a limitation and discuss the problem now in lines 416-412 and 565-567.

The authors should fully exploit the SAG(-) line, which is knockout for saglin and provides a unique opportunity to determine the role of this protein during invasion of the salivary glands. This would considerably augment the impact of the study. In this regard, line 131 and Fig S3E: why is there persistence of a PCR band for non-excised in the sag(-)EX DNA?

We definitely share the reviewer's enthusiasm about saglin and its role in parasite development in mosquitoes. We have thoroughly characterized the phenotype of sag(-) lines with respect to fitness and Plasmodium infection. These results are described in a spearate manuscript currently in peer review and available as a preprint on bioRxiv (https://doi.org/10.1101/2022.04.25.489337). Furthermore, in the revised manuscript, we have included additional data on the transcriptional activity of the saglin promoter with respect to the onset of expression and blood meal inducibility (Figure 2). In addition, we have included a completely new Figure 3 to highlight the spatial differences in transcriptional activity of the saglin promoter compared with the trio promoter. These new data are commented in lines 206-276.

There might be a misunderstanding in the interpretation of the genotyping PCR. The PCR shown in Figure 1 – figure supplement 3, displays PCR products for different genomic DNAs (sag(-)EX, sag(-)KI and wild type) using the same primer pair. „Excised“ refers to sag(-)EX while „non excised“ refers to sag(-)KI and „control“ to wild type. Primers were chosen in a way to yield a PCR product as long as the transgene has integrated, only the shift in size between „excised“ and „non excised“ indicates the loss of the 3xP3-lox fragment. We have now changed the labeling of the respective gel in Figure 1 – figure supplement 3 to make this clearer.

Did the authors search for alternative integration of the construct to explain the trioDsRed variability?

We validated trio-DsRed cassette insertion in the X1 locus by PCR. The only way to rule out an additional integration of the transgene would be whole genome sequencing, which we did not perform. Still, we believe that the observed expression patterns are due to locus-specific effects of the X1 locus. Indeed, several lines of evidence point in this direction: (1) transgenesis was realized using the phage Φ31 integrase that promotes site-specific integration (attP is 38bp long and very unlikely to occur as such in the mosquito genome) and for which we never detected insertion in other sites in the genome for other constructs inserted in X1 and other docking lines; (2) additional unlinked insertions would have been easily detected during the first backcrosses to WT mosquitoes we perform in order to isolate the transgenic line and homozygotise it; (3) we have often observed variegated expression patterns for other transgenes located in the X1 locus in the past, leading us to believe that this locus is subjected to variegation influencing the expression of the inserted promoters. Usually, the variation we observe is simpler (e.g. strong and weak expression of the fluorescent reporter placed under the control of the 3xP3 promoter in the same tissues where it is normally expressed), but some promoters are more sensitive to nearby genomic environment than others, which we believe is the case for trio. Finally, should there be additional insertions of the transgenesis cassette in the genome, they should all be linked to the X1 locus as we would otherwise have detected them in the first crosses as mentioned above, which is unlikely. Thus, although very unlikely, we cannot exclude a single additional and linked insertion possibly explaining the high/low DsRed patterns, but variegation would still be required to explain other patterns. We have mentioned this alternative explanation in the manuscript in lines 522-524.

Line 254-255. Does the abnormal morphology of SG from aapp-hGrx1-roGFP2 result in reduced sporozoite transmission?

This is an interesting question. For future experiments, it could indeed be important to test if the transmission of sporozoites by the generated salivary gland reporter lines is not impaired. However, the quantification of the number of sporozoites in aapp-hGrx1-roGFP2 expressing salivary glands did not reveal any significant differences from the wild type (Figure 5 – figure supplement 1B) and would definitely be sufficient to infect mice. As we have no evidence for reduced invasion of sporozoites in the salivary glands of aapp-hGrx1-roGFP2 and of the DsRed reporter lines, no good reason to believe that the expression of fluorescent proteins would interfere with parasite transmission, and as we produced these lines as tools to follow sporozoite interaction with salivary glands, we have not performed transmission experiments.

Of note, we have now included images of highly infected salivary glands of all reporter lines in Figure 5 – figure supplement 1D to confirm that expression of the respective fluorescence reporter does not interfere with sporozoite invasion. Also we have not observed that sporozoites do not invade salivary gland areas displaying high levels of hGrx1-roGFP2.

**Minor comments**

-Line 51: sporogony rather than schizogony

Schizogony was replaced with sporogony.

-Line 56: sporozoites are not really deformable as they keep their shape during motility

This sentence was removed.

-In the result section, it is not clearly explained where constructs were integrated.

We have now included the sentence „...with an attP site on chromosome 2L...“ (line 173) and the respective reference (PMID: 25869647) to give more information about the integration site.

Line 106 and 434-435: for the non-expert reader, it is not clear what X1 refers to, strain or locus for integration?

X1 refers to both, the locus and the docking line. We have rephrased the beginning of the result section (previously line 106) to give more information about the integration site as mentioned above.

-Line 112-115: the rational of integrating GFP instead of SAG is not clearly explained here, but become clearer in the discussion (line

We have slightly rephrased the sentence to better explain the reasoning for this procedure (lines 182-184).

-Line 140: FigS2A instead of S3A

This mistake was corrected in the revised manuscript.

-Perhaps mention that GFP reporters (SG) might be useful to image RFP-expressing parasites.

We have now included an image of the aapp-hGrx1-roGFP2 line infected with a mCherry expressing P. berghei strain in Fig. 7D.

-Line 236: the authors cannot exclude integration of an additional copy (as mentioned in the discussion line 367-368).

As discussed above, we removed „..as a single copy...“ and introduced the possibility of an additional integration linked to X1 (lines 522-524).

-Line 257-258. The title of this section should be modified as SG invasion was not captured.

The title was rephrased. It reads now „Salivary gland reporter lines as a tool to investigate sporozoite interactions with salivary glands” (line 356-357).

-Line 287: remove "considerable number" since there is no quantification.

This was removed. In addition, we included new data in this section of the manuscript and rephrased the results accordingly (lines 406-427).

-Line 400-402: Klug and Frischknecht have shown that motility precedes egress from oocysts (PMID 28115054), so the statement should be modified.

Thank you for this suggestion. The passage was modified accordingly.

-Line 404: remove "significant number" since there is no quantification.

This section was rephrased and the phrase "significant number" was removed (lines 406-427).

-Line 497: typo "transgenesis"

The typo was correct in the revised manuscript.

-FigS1: add sag-DsRed in the title

Thank you for spotting this inconsistency, we corrected this mistake (line 1134).

-Stats: Mann Whitney is adequate for analysis in fig 2C but not 2B, where ANOVA should be used (more than 2 groups).

We have performed now an one-way-ANOVA test and adapted figure and figure legend accordingly.

Reviewer #1 (Significance (Required)):

This work describes a technical advance that will mainly benefit researchers interested in vector-Plasmodium interactions. Invasion of salivary glands by Plasmodium sporozoites is an essential step for transmission of the malaria parasite, yet remains poorly understood as it is not easily accessible to experimentation. The development of transgenic mosquitoes expressing fluorescent salivary glands and with decreased pigmentation provides novel tools to allow for the first time in vivo imaging in live mosquitos of the interactions between sporozoites and salivary glands.

Reviewer's expertise: malaria, Plasmodium berghei, genetic manipulation, host-parasite interactions

Reviewer #2 (Evidence, reproducibility and clarity (Required)):

The first achievements of the Klug et al. study are the (i) genetical engineering of the Anopheles coluzzii mosquitoes reared in insectarium, that stably express distinct fluorescent reporters (DsRed and hGrx1-roGFP2 and EGFP) under the putative "promoters" of genes reported to encode proteins expressed differentially in the pluri-lobal salivary glands(Sg) of anthropophilic blood-feeding adult females, (ii) the analysis of the promoter activity - based on the selected fluorescent reporter - with a primary focus on the salivary gland/Sg (including at the Sg lobe level) of the adult female but also considering the preimaginal developmental time with larvae and pupa samples. Of note, some data confirm the already reported time-dependent and blood meal-dependent promoter activity for the related Anopheles species. The last part presents preliminary dataset on live imaging of Plasmodium berghei sporozoites with the aim of highlighting the usefulness of these A. coluzzii transgenic

lines to better understand how the rodent Plasmodium sporozoites first colonize and then settle as packed cells in Sg acinar host cells.

**Major comments**

The two first objectives presented by the authors have been convincingly achieved with (i) the challenging production of four different lines expressing different single or double reporters chosen by the authors (and appropriately presented in the result text and figure sections), (ii) the careful analysis of the spatiotemporal expression of the DsRed reporter under two "promoters" studied and with regards to the blood feeding event parameter. However, if the reason why the authors have put so much effort in the production of their transgenic mosquitoes is (and as mentioned) to provide a significant improved setting enabling the behavioral analysis of sporozoites upon colonization and survival in the Sg, it seems this part is kind of limited. Likely in relation with this perception is the fact I found the introductory section often confusing and not enough direct to the points: in particular distinguishing the rationale from the necessity to produce appropriate models, and clarifying what is/are the added value(s) offered by these new transgenic lines models when compared to what exist (in Anopheles stephensi) with specific evidence that argue for this knowledge gain. At this stage, it is unfortunately not clear to me, what is the bonus of imaging the Plasmodium fluorescent sporozoites in hosts with fluorescent salivary gland lobes if one can not monitor key events of the Sg-sporozoite interaction that were not reachable without the fluorescent mosquito lines. Furthermore, it should be better explained why the rodent Plasmodium species has been chosen rather Plasmodium falciparum (or other human species) for which A. coluzzii is a natural host; may be just mentioning that this study would serve as a proof of concept but bringing real biological insights would be fine.

We would like to thank the reviewer for his/her evaluation of our manuscript, which has helped us clarify our manuscript on several points. Our goal here was a proof of concept demonstrating potential applications for the fluorescent salivary gland reporter lines and for the low pigmented yellow(-) line we generated. In vivo imaging of sporozoites in salivary glands is one possible application that we intended to use as proof-of-concept, but we tailored the manuscript too restrictively with this aim in mind and neglected other applications as well as characterization of the biology of salivary glands in general. To improve this, we have included further data on the blood inducibility of the promoters tested (Figure 2), the functionality of roGFP2 in the salivary glands (Figure 5), and the use of the generated lines in the examination and definition of expression patterns of salivary gland proteins in vivo (Figure 6). Accordingly, we have adjusted the entire manuscript to adequately describe all the results presented. We have also rephrased major parts of the abstract and the introduction to better describe the impact of salivary gland biology on the transmission of pathogens, and to explain the anatomy of salivary glands in more detail.

We agree with the reviewer that it would be desirable to show direct salivary gland-sporozoite interactions in vivo. Still we believe that having mosquito lines expressing a fluorescent marker in the salivary gland as well as weakly pigmented mosquitoes are a first step to make this visualization possible, although we cannot provide a lot of quantitative data about this interaction yet.

1- The three genes and gene products selected by the authors should definitively be more systematically explained, which means for example the authors need to introduce the different mosquito species and the parasite-mosquito host pairs they are then referring to for the promoter/encoded proteins of their interest. In the same vein, I did not find any information as to the choice of the mosquito species (A. Coluzzii) for the current work. I was curious to know what is the advantage since better knowledge was available with Anopheles stephensi with respect to (i) Saglin and its promotor activity, (ii) aap driven dsRed expression (lines already existing) and (iii) sporozoite-gland interaction.

We have largely reworded the introduction to clarify the rationale for selecting these three promoters while providing a better understanding of salivary gland biology in general.

The choice of the mosquito species depends, in our opinion, strongly on the perspective and on the experiments to be performed. We agree with the reviewer that the malaria mosquito A. stephensi is a widely used model, based on its robustness in breeding and its high susceptibility to P. berghei and P. falciparum infections. However, in these cases, both vector-parasite pairs are to some extend artificial. Indeed, although it is also a vector of P. falciparum in some regions, A. stephensi mostly transmits P. vivax that cannot be cultured in vitro. Thus research efforts on this vector-parasite pair is limited. Also, due to the emerging number of observed differences between Anopheles species and their susceptibility to Plasmodium infection and transmission, more research has recently been conducted on African mosquito species. This effect is also reinforced by the fact that P. falciparum, unlike all other Plasmodium species infecting humans, causes the most deaths, making control strategies for species from the A. gambiae complex such as A. coluzzii particularly important. As a result, the number of available genetic tools in A. coluzzi/gambiae has overpaced A. stephensi. These include mosquito lines with germline-specific expression of Cas9 for site-directed transgenesis, lines expressing Cre for lox-mediated recombination, and several docking lines. Such tools are, as far as we know, not available in A. stephensi and were essential in reaching our objectives. Docking lines are of particular interest because they allow reliable integration into a characterized locus, which is an advantage over random transposon-mediated integration. Random insertion sites have generally not been characterized in the past, which can cause problems since integrations regularly occur in coding sequences. Docking lines also enable comparison of different transgenes as they are all integrated in the same genetic environment, which does not ensure some expression variation as illustrated in our manuscript. For all these reasons, we have thus chosen to work with A. coluzzii.

Concerning the use of the murine malaria parasite P. berghei instead of the human one P. falciparum, there are two reasons that motivated our choice. (1) For in vivo imaging of sporozoites, we needed a parasite line that is strongly fluorescent at this stage, and there is no such line existing for P. falciparum. Actually, there is no fluorescent P. falciparum line able to efficiently infect A. coluzzii reported thus far, as reporter genes have all been inserted in the Pfs47 locus that is required by P. falciparum for A. coluzzii colonization. (2) Imaging P. falciparum infected mosquitoes, especially with sporozoites in their salivary glands, requires to have access to a confocal microscope in a biosafety level 3 laboratory. Hence our objective here was indeed to provide a proof of principle of in vivo imaging of sporozoites in the vicinity or inside salivary glands using our engineered mosquitoes, and to provide a first analysis of this process using P. berghei as a model of infection. Nevertheless, we agree with the reviewer that the goal should be to work as close as possible to the human pathogen.

Despite the wide range of topics that this study touches on, we want to try and keep the manuscript as concise as possible. Therefore, we have not discussed the advantages and disadvantages of the different vector-parasite pairs and ask the reviewer to indulge us in this.

2- To help clarifying the added value of the present study, introducing the species names of the mosquito and the Plasmodium that serve as a model would be appreciated.

We have included now the name of the used Plasmodium species in line 361. At this position we also give now more details about the transgene this line is carrying. We mention the used mosquito species A. coluzzii now at different positions in the manuscript (e.g. lines 52, 162 and 177).

3- Since a focus is the salivary gland of the blood feeding female Anopheles sp., a rapid description of the glands with different lobes and subdomains the results and figure 1 nicely refer to, would help in the introduction.

We explain now the anatomy of female and male mosquito salivary glands in the introduction (lines 119-123). The different lobes are now also indicated in the salivary gland images shown in several figures including Figure 1.

4- That description could logically introduce the few proteins actually identified with lobe specific or cell domain specific expression (apical versus basal side, intracellular or surface expose, vacuole, duct...) profiles. The context with regards to sporozoite biology would then easily validate the "promoter choice". As a minor remark, I miss the reason why the authors wrote " the astonishing degree of order of the structures (referring to the packing of sporozoites within the Sg acinars) raise the question whether sporozoite can recognize each other". Please clarify since packing/accumulation can be passive due to cell mechanical constraints and explain what this point has to see with the question and experimental work proposed here?)

We thank you for this suggestion. We have reworded key parts of the introduction to make the reasons for using the three selected promoters clearer. We also mention now other proteins expressed in the salivary glands which have been characterized in more detail because of their effect on blood homeostasis (e.g. anticoagulants) (lines 136-139).

The mention of stack formation of salivary gland sporozoites served only to clarify that almost nothing is known about the behavior of sporozoites within the salivary glands in vivo to explain why new methods are needed to make these processes visible. We have now reworded this passage to make this clearer, and we also mention that stack formation could also occur due to mechanical constraints, as suggested by the reviewer (lines 101-102, 106-110).

5- The selection of hGrx1-roGFP2 is quite interesting and justified but there is then no use of this reporter property in the preliminary characterization of the Sg and Sg-sporozoite interaction. Could the authors provide such characterization?

We have now implemented data testing the functionality of hGrx1-roGFP2 in the salivary glands. We also show qualitatively that the redox state of glutathione does not change upon infection with P. berghei sporozoites (Figure 6). We now describe and discuss these new data in lines 337-354.

6- Figure 1: it would be nice to add in the legend at what time the dissection/imaging has been made (age, blood feeding timing?). I would also omit the double mutant trio-Dsred/aapDsred in the main figure (may be supplemental) since the two single mutants Dsred separately together with the double mutant (with different fluorescence) already provide the information. I would suggest to regroup the phenotypic presentation of the transgenic line made in the KI mosquitoes (current figure 5) in the main figure 1.

We have now added the missing information about the age of dissected mosquitoes and their feeding status in the legend of Figure 1. We also thank the reviewer for the suggestion to replace one image displaying aapp and trio promoter activity in trans-heterozygous mosquitoes with an image of the pigment deficient mutant yellow(-)KI. Still, due to the changes made to the manuscript based on the reviewers comments in general, we have now implemented new data highlighting the functionality of the generated salivary gland reporter lines investigating the redox state of glutathione as well as the expression pattern of the saglin and trio promoters at the single cell level (see Figure 3 and 6). Therefore it would no longer seem logical to introduce the yellow(-)KI mutant in Figure 1 while further data on this mutant are provided in the last two figures of the manuscript and discussed later in the manuscript (Figure 7 and 8). In addition we believe that co-expression of different transgenes (carrying fluorescent reporters) in the median and the distal lobes could potentially be interesting for certain applications. We believe that readers who might actually be interested in combining both transgenes in a cross would like to see the outcome to better evaluate the usefulness before experiments are planned and performed. This is especially true because localization as well as expression strength may differ between different fluorescence reporters while using the same promoter (e.g. the hGrx1-roGFP2 construct appears less bright and more localized to the apex of the distal-lateral lobes than dsRed, while expression of both reporters is driven by the aapp promoter in aapp-hGrx1-roGFP2 and aapp-DsRed, respectively).

7- Figure 2:

1. a) Is there anything known on the Sgs' size change overtime. It seems that between day 1 and 2 there is an increase of size and volume as much as I can evaluate the volume (Fig S4). Could that mean that there is increase in cell number in the lobes and therefore more cells expressing the transgene which would account for the signal intensity increase rather than more transcripts per cell? Thank you for this interesting question. The changes in the morphology of the salivary glands in Anopheles gambiae following eclosion have been studied in detail by Wells et al., 2017 (PMID: 28377572) which we cite now in the introduction (line 122-123). According to this reference, cell counts of the salivary gland are not changing upon emergence of the adult mosquito. However, we agree with the reviewer that the glands appear smaller and differ in morphology directly after eclosion. We noted that glands of freshly emerged females are more „fragile“ during dissections and lack secretory cavities, as reported by Wells et al., 2017. We believe that the increase in size occurs through the formation and filling of the secretory cavities which has been reported to take place within the first 4 days after emergence (Wells et al., 2017). This observation is in accordance with our observations that the promoters of the saliva proteins AAPP and Saglin display only weak activity after hatching, or, in the case of TRIO are not yet active directly after emergence. The timing of the formation of the secretory cavities is also in agreement with our time course experiment (Figure 2) which shows a strong increase in fluorescence intensity in dissected glands within the first 4 days after emergence.

2. b) why choosing 24h after the blood meal to assess promoter activity in the Sgs? Do we have any information on how the blood meal impact on the Sgs'development. At this time anyway the sporozoites are far from being made. Yosshida and Watanabe 2006 mentioned at significant decrease of Sg proteins post-blood feeding. Could the authors detail their rationale based on what the questions they wish to address Thank you for this question. Unfortunately, the data available in the literature on this topic are very sparse, so we could only refer to few previous publications. The decision to quantify the fluorescence signals as early as 24 hours after blood feeding was based on Yoshida et al, Insect Mol. Biol, 2006, PMID: 16907827. The authors of this study generated the first salivary gland reporter line in A. stephensi by using the aapp promoter sequence to drive DsRed expression, and showed by qRT-PCR that DsRed transcripts increase 1-2 days after blood feeding compared to controls. Consistent with this observation and because we were concerned that putative changes in protein levels would only be visible for a short period of time, we began quantification one day after feeding. Since we observed significant changes in fluorescence intensity for the aapp-DsRed and sag(-)KI lines 24 hours after blood feeding, we retained the experimental setup and did not change it further. Nevertheless, we agree with the reviewer that different time points could help determine how long the effect lasts, and whether trio expression might also be regulated by blood feeding, but at a later time point. Still, our main objective here was to validate that the ectopic expression of DsRed driven by the aapp promoter in the aapp-DsRed line was indeed induced upon blood feeding as previously reported (PMID: 16907827). This experiment allowed us to confirm the inducibility of aapp in a different way and to show for the first time that saglin, but not trio, is induced one day after blood feeding. Our transgenic lines could be used for follow-up studies investigating the inducibility of salivary gland-specific promoters by different stimuli, or after infection with Plasmodium sporozoites. For example, for trio, transcription has been shown to increase after infection of the salivary gland by Plasmodium (PMID: 29649443).

8- Figure 3: The figure is quite informative in terms of subcellular localization. Concerning the section "Natural variation of DsRed expression in trio-DsRed mosquitoes", I think it could be shortened because because it is a bit out of the focus the study.

We agree with the reviewer that this part of the manuscript sticks a bit out and is not perfectly in line with the remaining results because it doesn’t deal with the salivary gland. Still, we would like to emphasise that in this work, we particularly want to show possible applications of the generated mosquito lines to address unanswered questions in host-parasite interactions and salivary gland biology. As a result, this manuscript establishes potentially important tools. For this reason, we feel it is important to mention the natural variation in DsRed expression, as this natural variation can have a significant impact on crossing schemes (especially with lines inheriting other DsRed-marked transgenes) and experiments (e.g. visualizing DsRed expression by western blot in larval and pupal stages). Furthermore, it is important for the use of the line to show that the transgene is inserted only once, at the expected location, which we try to emphasize with figure 4 – figure supplement 1 and figure 4 – figure supplement 2.

We would also like to note that transgenesis in Anopheles is a relatively young field of research and altered expression patterns of ectopically used promoters have rarely been described so far, although this could have major implications e.g. in the case of gene drives. Therefore, we hope that the data shown will bring this previously neglected observation more into focus and highlight the importance of accurate characterization of generated transgenic mosquito lines.

9- In contrast the last section of live imaging of P. berghei sporozoites in the vicinity and within salivary gland should be expanded. The 2 sentences summarizing the data are quite frustrating "We also observed single sporozoites moving actively through tissues in a back and forth gliding manner (Fig. 6B, Movie 3) or making contact with the salivary gland although no invasion event could be monitored"

We have now implemented new data and extended Figure 8 showing the results of the in vivo imaging in a qualitative manner. We have rephrased the result and discussion section accordingly.

10- I am aware of the technical difficulties to perform live imaging of sporozoite on whole mosquitoes, even when the salivary gland lobe under observation is closely apposed to the cuticle but that seems to be the final aim of the authors. I looked very carefully to the three movies and I am sorry but at this stage I could not make meaningful analysis out of them, and could not agree with the conclusions: for instances, the authors specify that sporozoites were undergoing back and forth movements (movie 3) but I do not see that and do not see the Sg contours in the available movies? The authors should also add bar and time scales to their movies. Having an in-depth description with regards to the sub-domain marked by a relevant reporter would strengthen the study, even if images are not collected in the whole mosquito to get higher resolution.

We thank the reviewer for this comment. We have to admit that parasite imaging in fluorescent salivary glands in vivo is an ambitious goal given the complex biological system we are working with. We believe that the system presented in our manuscript is a first and important step to enable the analysis of the interaction of sporozoites with salivary glands, although in-depth analysis will require further optimization and considerable time, especially to generate quantitative data. Therefore, we now downstate the significance of our results in this respect and changed the title accordingly. Still, we also provide a more detailed analysis of the data we have already collected (Figure 8 and lines 406-427). Because we focus on the analysis of sporozoites in the thorax area in the revised manuscript, the outlines of the salivary gland are not necessarily visible in the images.

I am not sure I understand the relevance of this quite condensed sentence in the text. Could the authors rephrase and expand if they wish to keep the issues they refer to. "The sporozoites' distinctive cell polarization and crescent shape, in combination with high motility, allows them to „drill" through tissues". I would stress more on the main unknown in terms of sporozoite-Sg interactions and the need to get right models for applying informative approaches (i.e. here, imaging).

We thank you for this suggestion. The sentence mentioned has been removed in its entirety. We have also adjusted the text accordingly and reworded most of the introduction to make the narrative clearer (lines 91-119).

Of note, it could help to point that the "Sgs is a niche in which the sporozoites which egress from the oocyst could mature and be fully competent when co-deposited with the saliva into the dermis of their intermediary hosts"

We have now implemented a similar sentence in the introduction (lines 93-98).

Reviewer #2 (Significance (Required)):

1- Clear technical significance with the challenging molecular genetics achieved in the mosquito A. coluzzii.

2- More limited biological significance: fair analysis and gain of knowledge of spatio-temporal of reporter expression under the selected promoter but limited significance of the final goal analysis which concerns the Plasmodium sporozoite biology once egressed from oocysts

As stated above, we changed the title to place the focus on the engineered mosquito lines.

3- Previous reports cited by the authors have used the DsRed reporter and the aap promoter in another Anopheles (i.e. A. stephensi, Yoshida and Watanabe, Insect Mol Biol, 2006; Wells and Andrew, 2019) which is also a natural host and vector for human Plasmodium spp.) with significantly more resolutive 3D visualization of GFP-fluorescent P. berghei but in dissected salivary glands and not in whole mosquitoes. The Wells and Andrew publication entitled "Salivary gland cellular architecture in the Asian malaria vector mosquito Anopheles stephensi" in Parasite Vectors, 2015 would deserve to be reference and described.

Thank you very much for this suggestion. We considered citing Wells and Andrews (PMID: 26627194). However, this reference focuses very specifically on the subcellular localization of AAPP and shows only highly magnified sections of immunostained dissected and fixed salivary glands. Working only with the AAPP promoter, we felt it important to refer to the previously observed expression pattern along the entire salivary gland, as shown in Yoshida and Watanabe (PMID: 16907827). Nevertheless, we have cited two other publications by Wells and Andrews (PMID: 31387905 and 28377572) at various points in the manuscript.

4- Audience: I would say that this work should be of interest of mostly scientists investigating Plasmodium biology (basic and field research) or in entomology of Diptera.

5- To describe my fields of expertise, I can refer to my extensive initial training in entomology including at one point in the genetic basis of mosquito-virus interaction. I have also been working for more than 20 years in the field of Apicomplexa biology (Plasmodium and Toxoplasma) and I have long-standing interest in live and static high-resolution imaging.

Reviewer #3 (Evidence, reproducibility and clarity (Required)):

Klug et al. generated salivary gland reporter lines in the African malaria mosquito Anopheles coluzzii using salivary gland-specific promoters of three genes. Lobe-specific reporter activity from these promoters was observed within the salivary glands, restricted either to the distal lobes or the medial lobe. They characterized localization, expression strength and onset of expression in four mosquito lines. They also investigated the possibility of influences of the expressed fluorescent reporters on infection with Plasmodium berghei and salivary gland morphology. Using crosses with a pigmentation deficient mosquito line, they demonstrated that their salivary gland reporter lines represent a valuable tool to study the process of salivary gland colonization by Plasmodium parasites in live mosquitoes. SG positioning close to the cuticle in 20% of females in this strain is another key finding of this study.

The key findings from this study are largely quite convincing. The authors have created a suite of SG reporter strains using modern genetic techniques that aid in vivo imaging of Plasmodium sporozoites.

Vesicular staining within salivary acinar cells should be stated as "vesicle-like" staining unless a co-stain experiment in fixed SGs is conducted using antisera against the marker protein(s) and antisera against a known vesicular marker (e.g. Rab11). It may also be possible to achieve this in vivo using perfusion of a lipid dye (e.g. Nile Red), but this is not necessary. As is, in Fig. 3A, there are images in which it appears that the vesicle-like staining is located both within acinar cells' cytoplasm and in the secretory cavities (e.g. Fig. 3A: aapp-DsRed bottom and middle), and this is fine, but should be more inclusively stated. Fixed staining of the reporter strain SGs would allow for clarification of this point. In previous work, other groups have observed vesicle-like structures in both locations (e.g. PMID: 33305876).

Thank you very much for this suggestion. Indeed, when we observed the vesicle-like localization, we had similar ideas and considered investigating the identity of the observed particles in more detail. Ultimately, however, we concluded that the localization of DsRed does not play a critical role in the use of the lines as such and believe that a more detailed investigation of the trafficking of the fluorescent protein DsRed is beyond the scope of this study.

We have thus followed the suggestion of the reviewer and now use the phrase „vesicle-like“ throughout the manuscript. In addition, we extended the discussion on the different localizations observed and presented some explanations that might have led to this observation. We also included a new reference that investigated the localization of AAPP using immunofluorescence (PMID: 28377572).

Morphological variation is extensive among individual mosquito SGs, thought to impact infectivity, and well documented in the literature. The manuscript should be edited to make it much clearer (e.g. n = ?) exactly how many SGs, especially in microscopy experiments, were imaged before a "representative" image was selected from each data point and in any additional experiment types where this information is not already presented. Figure S8 is an example where this was done well. Figure 3A-B is an example where this was not well done. All substantial variation (e.g. "we detected a strangulation..." - line 189) across individual SGs within a data point should be noted in the Results. Because of the genetics and labor involved, acceptable sample sizes for minor conclusions may be small (5-10), but should be larger for major conclusions when possible.

Thank you for this comment. We have improved this point by specifying precisely the number of samples and of repetitions in the respective figure legends. For example, we have now quantified the proportion of moving sporozoites and report both the number of sporozoites evaluated and the number of microscopy sessions required (see Figure 8).

Thank you for this comment. We have improved this point by specifying precisely the number of samples and of repetitions in the respective figure legends. For example, we have now quantified the proportion of moving sporozoites and report both the number of sporozoites evaluated and the number of microscopy sessions required (see Figure 8). Regarding Figure 3, fluorescence expression and localization in salivary gland reporter lines was actually very uniform in each line. We added the following sentence in the legend of revised figures 3 and 5: “Between 54 and 71 images were acquired for each line in ≥3 independent preparation and imaging sessions. Representative images presented here were all acquired in the same session”.

Sporozoite number within SGs has been shown to be quite variable across the infection timeline, by mosquito species, by parasite strain, in the wild vs. in the lab, and according to additional study conditions. The authors mention that the levels they observed are consistent with their prior studies and experience, but they did not utilize the reporter strains and in vivo imaging to support these conclusions, instead relying on dissected glands and a cell counter. It is important for these researchers to attempt to leverage their in vivo imaging of SG sporozoites for direct quantification, likely using the "Analyze Particles" function in Fiji. The added time investment for this additional analysis would be around two weeks for one person experienced in the use of the imaging software.

Thank you for this interesting suggestion. Indeed, it would be beneficial to use an imaging based approach to quantify the sporozoite load inside the salivary glands. We already used „watershed segmentation“ in combination with the „Analyze Particles“ function in Fiji on images of infected midguts to determine oocyst numbers. Still, we believe this analysis cannot be applied to images of infected salivary glands mainly because of differences in shape and location of the oocyst and sporozoite stages. Sporozoites inside salivary glands form dense, often multi-layered stacks. Because of this close proximity, watershedding cannot resolve them as single particles which could subsequently be counted. This creates an unnecessary error by counting accumulations of sporozoites as one, likely leading to an underestimation of actual parasite numbers. Furthermore, given that the proximity issue could be resolved e.g. by performing infections yielding lower sporozoite densities, another problem would be that infected salivary glands prepared for imaging are often slightly damaged leading to a leak of sporozoites from the gland into the surrounding. These leaked sporozoites are likely not included on images which would then be used for analysis, potentially leading again to an underestimation of counts. Since these issues are circumvented by the use of a cell counter, we believe that this method is still the method of choice in acquiring sporozoite numbers.

Nevertheless, we can understand the reviewer's concern that counts performed with a hemocytometer do not reflect the variability in the sporozoite load of individual mosquitoes. To highlight that all generated reporter lines can have high sporozoite counts, we have now included images of highly infected salivary glands for each line in Figure 7D.

This manuscript is presented thoughtfully and such that the data and methods could likely be well-replicated, if desired, by other researchers with similar expertise.

The statistical analysis is appropriate for the experiments conducted. It is currently unclear if some experiments were adequately replicated. That information should be added to the paper throughout where it is missing.

We do appreciate your comments on our efforts to give all required information for other laboratories to replicate our experiments. We have added the missing information about the number of independent experiments in the respective figure legends wherever appropriate.

Studies from multiple groups should be more thoroughly referenced when the authors are describing the "vesicle-like" staining patterns observed in SGs from reporter strains (e.g. Fig. 3A). Is this similar to the SG vesicle-like structures observed previously (e.g. PMIDs: 28377572, 33305876, and others)?

Thank you for this comment. We did not discuss this observation in detail in the first version of our manuscript because the observed localization was rather unexpected, as DsRed was not fused to the AAPP leader/signal peptide. The observed localization is therefore difficult to explain, however, we have expanded the discussion on this (lines 465-482) and now cite one of the proposed references (PMID: 28377572, lines 468-469).

There are minor grammar issues in the manuscript text (e.g. "Up to date" should be "To date"). The figures are primarily presented very clearly and accurately. One minor suggestion: In cases such as Fig. S2A images 3 and 6, where some of the staining labels are very difficult to read, please move all labels for the figure to boxes located directly above the image.

We are sorry for the grammatical errors we have missed in the first version of our manuscript. We have now performed a grammar check over the whole manuscript. We have also increased the font size of the captions in the above figures and tried to make them better readable by moving the captions over the images.

The data and conclusions are presented well.

Reviewer #3 (Significance (Required)):

This report represents a significant technical advance (improved in vivo reporter strain and sporozoite imaging), and a minor conceptual advance (active sporozoite active motility), for the field.

This work builds off of previous SG live imaging studies involving Plasmodium-infected mosquitoes (e.g. Sinnis lab, Frischneckt lab, etc.), addressing one of the major challenges from these studies (reliable in vivo imaging inside mosquito SGs).

This work will appeal to a relatively small audience of vector biology researchers with an interest in SGs. Many in the field still see the SGs as intractable, instead choosing to focus on the midgut due to ease of manipulation. Perhaps work like this will spark new interest in tangential research areas.

I have sufficient expertise to evaluate the entirety of this manuscript. Some descriptors of my perspective include: bioinformatics, SG molecular biology, mosquito salivary glands, microscopy, RNA interference, SG infection, and SG cell biology.

Reviewer #4 (Evidence, reproducibility and clarity (Required)):

Klug et al generated transgenic mosquito lines expressing fluorescent reporters regulated by salivary gland specific promoters and characterized fluorescent reporter expression level over the time, subcellular localization of fluorescent reporters, and impact on P. berghei oocyst and salivary gland sporozoite generation. In addition, by crossing one of the lines (aapp-DsRed) with yellow(-) KI mosquitoes, they open up the possibility to perform in vivo visualization of salivary glands and sporozoites.

Overall the generation and characterization of these transgenic lines is well-done and will be helpful to the field. However, there are several concerns with the in vivo imaging data shown in Figure 6, which does not convincingly show fluorescent sporozoites in the lobe or secretory cavity of a fluorescent salivary gland lobe. This needs to be addressed. Points related to this concern are outlined below:

(1) Although the authors mention that the DsRed signal was strong enough to see with GFP channel, it would be more appropriate to show that the DsRed signal from salivary glands and GFP channel image co-localize.

We now show a merge of the GFP and DsRed signal in Figure 7 – figure supplement 2 The yellow appearance of the salivary gland in the merge likely indicates the spillover of the DsRed signal into the GFP channel. In addition we discuss the issue in lines 416-412 and 565-567.

(2) Mosquitoes were pre-sorted using the GFP fluorescence of the sporozoites on day 17-21. From figure 4B, median salivary gland sporozoite number was about 10,000 sporozoites/mosquito on day 17-18. However, in Figure 6A there are no sporozoites in the secretory cavities. They should be able to see sporozoites in the cavities at this time. Can the authors confirm that they can visualize sporozoites in secretory cavities in vivo and perhaps show a picture of this.

This is entirely correct. We also examined mosquitoes for the presence of sporozoites in the salivary glands and wing joints prior to imaging, as shown in Figure 7B and Figure 7 – figure supplement 2A, to increase the probability that sporozoites could be observed. Nevertheless, the area of the salivary gland that comes to the surface is often small and limited to a few cells that can be imaged with good resolution. Unfortunately, these same cells were often not infected although other regions of the salivary glands must have been very well infected based on the previously observed GFP screening (Figure 7B). In addition, with the confocal microscope available to us, we struggled to achieve the necessary depth to image sporozoites in the cavities of the salivary gland cells. For this reason, we were often able to detect a strong GFP signal in the background, but not always to resolve the sporozoites sufficiently well. Still, we have now included an image showing sporozoites in salivary glands (Figure 8C). However, we believe that the method can be further improved to be more efficient and provide better resolution. We discuss possible ways to further improve the imaging in lines 563-586.

(3) There is no mention of the number of experiments performed (reproducibility) and no quantification of the imaging data. In the results (line 287-288), the authors state that sporozoites are present in tissue close to the gland and sometimes perform active movement. How can this be? Do they believe these sporozoites are on route to entering? More relevant to this study would be a demonstration that they can see sporozoites in the secretory cavities of the salivary gland epithelial cells, this should shown. If they have already performed a number of experiments, I would suggest to do quantification of the number of sporozoites observed in defined regions . The mention that sporozoites are moving is confounded by the flow of hemolymph. How do they know that the sporozoites are motile versus being carried by the hemolymph. Perhaps it's premature to jump to sporozoite motility in the mosquito when they haven't even shown sporozoite presence in the salivary glands.

Thank you very much for this comment. We have followed the suggestions of the reviewer and have now quantified the behavior of sporozoites in the thorax area of the mosquito. For the analysis, we only considered sporozoites that could be observed for at least 5 minutes. This analysis revealed that 26% of persistent sporozoites performed active movements, which in most cases resembled patch gliding previously described in vitro. We adjusted the results section accordingly. In addition, we have changed the figure legend to accurately indicate the number of experiments performed. Likewise, we now also provide an image of sporozoites that we assume are located in the salivary gland (Figure 8C). Although we have not yet been able to image and quantify vector-sporozoite interactions extensively (further improvements would be required, as mentioned previously), we believe these results illustrate the potential of the transgenic lines.

(4) In vivo imaging has been performed with the mosquito' sideways. Was this the best orientation? Have you tried other orientations like from the front (Figure 5B orientation).

It is true that in the abdominal view as shown in Figure 7B the fluorescence in the salivary glands is very well visible. This is mainly due to the fact that in this area the cuticle is almost transparent and therefore serves as a kind of "window". Nevertheless, the salivary glands are not close to the cuticle in this position, which makes good confocal imaging impossible. Imaging always worked best where the salivary gland was very close to the cuticle, and this was always laterally. However, there were differences in the position of the salivary glands in individual mosquitoes, which also led to slight differences in the imaging angle.

Overall, the text is easy to follow and I have only few suggestions.

Thank you for this comment.

In the result section, the authors describe the DsRed expression during development of mosquito (line 194-236) after they describe subcellular localization of fluorescent reporters. I felt the flow was disrupted. Thus, this part (line 194-236) could summarize and move to line 135. In this way, the result section flow according to the main figures.

Thank you very much for this suggestion. We have considered your idea, but based on the changes we have made in response to reviewer comments and new data implemented in the form of two new figures, we believe the current order in the results section is more appropriate. The rationale was primarily to first characterize the expression of fluorescent reporters in the salivary glands of all lines before going into more detail on expression in other tissues of a single line. We then finish with potential applications like in vivo imaging of sporozoite interactions with salivary glands.

Also, and as mentioned previously (reviewer 2, point 8), we believe it is important to describe the variability of ectopic promoter expression at a given locus with sufficient details, as this has not been characterized thus far despite its importance.

In the result section, text line 186-190, the authors describe the morphological alternation of salivary gland in aapp-hGrx1-roGFP2. I would suggest to mention that this observation was only in one of lateral lobe. (I saw that it was mentioned in the figure legend but not in the main text.)

We believe there has been a misunderstanding. The morphological alteration in salivary glands expressing aapp-hGrx1-roGFP2 was observed in all distal-lateral lobes to varying degrees (quantification in Figure 6E). To include as many salivary glands as possible in the quantification and because in some images only one distal-lateral lobe was in focus, only the diameter of one lobe per salivary gland was measured and evaluated. We have now revised the legend to prevent further misunderstandings.

In the discussion section, author discuss localization of fluorescent reporters (line 322-331). When I looked at aapp-DsRed localization pattern (Figure 3A), the pattern looked similar to the previous publication by Wells et al 2017 (https://www.nature.com/articles/s41598-017-00672-0). This publication used AAPP antibody and stain together with other markers (Figure 4-7). This publication could be worth referring in the discussion section.

Thank you for this suggestion. According to the information available through Vectorbase, we did not fuse DsRed with any coding sequence of AAPP that could potentially encode a trafficking signal. Therefore, it is rather unlikely that the observed DsRed localization in our aapp-DsRed line and the localization observed by AAPP immunofluorescence staining in WT mosquitoes match. This is further exemplified by the cytoplasmic localization of hGrx1-roGFP2 in the aapp-hGrx1-roGFP2 line, where the reporter gene was cloned under the control of the same promoter. For this reason, we had not mentioned this reference in the first version of the manuscript. In the revised manuscript, we have included now the suggested reference (lines: 475-476) and extended the discussion on possible reasons which led to the observed localization pattern.

In the text, authors describe salivary gland lobes as distal lobes and middle lobe. It would be more accurate to refer to the lobes as the lateral and medial lobes. The lateral lobes can then be sub-divided into proximal and distal portions. I would suggest to use distal lateral lobes, proximal lateral lobes and median lobe as other references use (Wells M.B and Andrew D.J, 2019).

Thank you for this suggestion. We have corrected the nomenclature for the description of the salivary gland anatomy as suggested throughout the manuscript.

Overall, the figures are easy to understand and I have following suggestions and questions.

Figure 1C) It is hard to see WT salivary gland median lobe. If authors have better image, please replace it so that it would be easier to compare WT and transgenic lines.

We have replaced the wild-type images of salivary glands in this figure and labeled the median and distal-lateral lobes accordingly (see Figure 1).

Figure 2) While it was interesting to observe the significant expression differences between day 3 and day 4, have you checked if this expression maintained over time or declines or increases (especially on day 17-21 when author perform in vivo imaging)?

Thank you for this interesting question. We have not quantified fluorescence intensities in mosquitoes of higher age. Nevertheless, we regularly observed spillover of DsRed signaling to the GFP channel during sporozoite imaging, suggesting that expression levels, at least in aapp-DsRed expressing mosquitoes, remain high even in mosquitoes >20 days of age (see Figure 8A). We also confirmed this observation by dissecting salivary glands from old mosquitoes, whose distal lateral lobes always showed a strong pink coloration even in normal transmission light (data not shown).

Figure 3A) There is no description of "Nuc" in figure legend. If "nuc" refers to nucleus, have you stained with nucleus staining dye (example, DAPI)?

Thank you for spotting this missing information in the legend. Initial images shown in this figure were not stained with a nuclear dye. To test whether the observed GFP expression pattern really colocalizes with DNA, we performed further experiments in which salivary glands from both aapp-hGrx1-roGFP2 and sag(-)KI mosquitoes were stained with Hoechst. We have now included these new data in Figure 3 - figure supplement 1. It appears that GFP is concentrated around the nuclei of the acinar cells, which makes the nuclei clearly visible even without DNA staining.

Figure 4B) The number of biological replicates in the figure and the legend do not match (In the figure, there are 3-5 data points and, in the legend, text says 3 biological replicates.)

Thank you for spotting this inconsistency. The number of biological replicates refers to the number of mosquito generations used for experiments. The difference is due to the fact that sometimes two experiments were performed with the same generation of mosquitoes using two different infected mice. We have clarified the legend accordingly to avoid misunderstandings.

Figure 4C) The number of data points from (B) is 5. However, in (C) only 4 data points are presented.

We have corrected this mistake. In the previous version, the results of two technical replicates were inadvertently plotted separately in (B) instead of the mean.

Figure 5) I would suggest to have thorax image of P. berghei infected mosquito to show both salivary glands and parasites.

Thank you for this suggestion. Images in Figure 7B (previously Figure 5) were replaced with an infected specimen to show salivary glands (DsRed) and sporozoites (GFP) together.

Reviewer #4 (Significance (Required)):

The transgenic lines that authors created have potential for in vivo imaging of salivary gland and sporozoite interactions. Since the aapp and trio lines have distinct fluorescence expression, they could help elucidate why sporozoites are more likely to invade distal lateral lobes compare to median lobe.

My areas of expertise are confocal microscope imaging, mosquito salivary gland and Plasmodium infection and sporozoite motility.

Given that data is but people in disguise

all businesses ultimately aimed at serving individual human needs so somewhere in the chain people will appear

we need autonomous emergent people centred interpersonal spaces

This thought runs through my head a lot. Would the "best" gamification turn each workers tasks into something that would relate well in a fantasy type world? Or how about a futuristic world? Perhaps it could change every year to ensure everyone is satisfied.

There should have been a stronger push to make the connections between the two problems that were asking the same thing. The student had just conveyed that they know how to solve the problem, but in a different context. If more of an emphasis was made to understand that they are working on the same problem they could apply the same reasoning in their head to solve other problems.

I feel like this is really connected to my first annotation on this essay, your plastic bags are not what is killing our planet. As we "legislate good eating habits." as it is put, are we further disenfranchising the people from their food? "Better" food is head and shoulders more expensive, you go to the grocery store and they separate the organic produce out into it's own sections, sections I never shop in because of the price difference. I am all for making sustainable and healthy food a more mainstream option but the way we currently go about it just further widens the gap between those who can afford "good" food and those who can't.

Active inference is exemplified using a robot, but is really a model of how humans learn, process information and make decisions in the world.

Reviewer #1 (Public Review):

This article presents interesting proof of concept of how predictive coding based on visual inputs, coupled with a complex array of RNNs can produce head direction and egocentric/allocentric boundary responses akin (to some extent) to the neural responses found in mammalian hippocampal formation neurons. However, while an impressive technical feat, the model contradicts key experimental findings, and the developmental timeline of spatial cell responses does not support the sole reliance on visual inputs.

Developmental considerations:<br /> Developmental studies have shown that rudimentary HD signals emerge before the unfusing of rat pup eyelids (suggesting visual inputs are not necessary for these initial responses). Head direction is fully formed more than a week before allocentric boundary responses (boundary vector cells: BVCs) emerge, and initial head direction signals predate BVC coding by 5-6 weeks in rat pups (Min, Wills, Cacucci 2017). While this does not exclude separate emergence per se, it removes the need to insist on the absence of HD inputs for the formation of BVCs. Hence it is plausible that at least HD would be available to any learning/developmental process that enables the emergence of allocentric boundary responses in the rodent brain. Not using this information would make this learning task unnecessarily difficult. Similarly, it is more plausible that egocentric boundary responses are constructed by a network that forms in a developmental time window and can later instantiate boundary responses based on visual inputs, depth perception, etc, without additional learning. The staggered emergence of spatial responses reported by experiments also strongly suggests that developmental stages build upon each other. Granted, it remains a possibility that egocentric boundary responses and head direction coding could be generated by a predictive coding framework (though the principle inputs to HD are vestibular), but there is no need to assume the head direction signal and the egocentric boundary signal wouldn't be used in a subsequent learning/maturation step that forms allocentric boundary responses from head direction and possibly egocentric boundary inputs. While I share the authors' sentiment that those previous models have not sufficiently accounted for this learning step (generating the network that allows egocentric signals to be translated to allocentric boundary representations), the appeal to a staggered developmental process is much more plausible and in line with developmental data.

The parallel emergence of distinct spatial responses:<br /> In the paper, very little is actually said about the interaction of the learned representation in the model. Since the different RNNs learn in parallel, one yielding HD, one egocentric boundary cell (EBCs), and one BVCs, this means that EBCs and BVCs do not need to interact. Similarly, HD cells and BVCs do not need to interact. This incredibly salient prediction is not emphasised at all. It would suggest the EBCs could be lesioned without affecting BVCs in a novel environment. In alternative models, this is not the case. Such strong claims should be emphasised as this is actually one of the few novel, direct experimental predictions that can be made here. Whether or not EBCs and BVCs interact is an open, empirical question. However, taking this line of reasoning further, the present model also predicts that lesioning HD cells should leave EBC and BVC unperturbed. This is extremely unlikely for BVCs. Lesions to the mammillary bodies (where HD cells are found and where the HD attractor signal is likely generated) lead to severe memory deficits. The orientation of BVCs and place cells is likely set by head direction cells. The three populations have repeatedly been shown to rotate in concert. Object vector cells (not addressed in this article) similarly co-rotate with HD cells. The article does not present sufficient evidence (or gains in understanding) to abandon this well-established view.

Relating the model to biological function:<br /> The normative account of the paper is interesting, but it is unclear how much (if anything) the model tells us about the biological underpinnings of spatial cognition despite the overt claim that the model would be useful to neuroscientists. The modelling approach is far from biologically plausible. This creates the unfortunate impression that a bunch of RNNs has been thrown together (with considerable technical skill), which are known to be able to extract the information inherent in the inputs. What does this tell us about how the brain generates these responses, and how can experimenters test for properties specific to the model? To provide a normative model that outlines one way for the appearance of known mammalian spatial representations based solely on interaction with the sensory world, is fine (and interesting in itself) but the method employed being so far from real biological function makes it impossible to assess if it is the correct normative explanation (see also next point).

Experimental contradictions:<br /> BVC activity emerges immediately upon entry into a new environment, while the present model needs to be retrained on sets of environmental geometries to be able to respond correctly in all those environments. This discrepancy cannot be remedied by appealing to the theoretical notion that an animal might experience all possible geometries during some developmental phase. Given the developmental timeline of spatial responses and the fact that rat pups do not leave their nest straight away this can in all likelihood be excluded. Competing models claim that EBC responses are computed directly from perceptual inputs (utilising networks formed in development), with the consequence that EBC (and hence BVCs driven by EBCs) can straightforwardly represent any new geometry without additional learning. This would be consistent with BVC activity emerging immediately in a new environment, even when faced with a never-before-experienced environmental geometry.

Reviewer #2 (Public Review):

The authors wish to investigate how various allocentric representations, such as those observed in the brain's navigational system, can emerge from the interaction between action and sensory inputs. They use a predictive architecture, in which visual inputs are predicted from actions, to explain the emergence of multiple allocentric representations (HD cells, place cells, boundary vector cells). The major strength of the paper is the demonstration of the network's ability to develop spatial representations of multiple virtual environments and the demonstration that such representations can be used as a foundation to quickly represent new environments and to support further reinforcement learning tasks. However, the analysis is not yet sufficient to support a number of claims made in the paper about critical pieces of the findings. Further, two critical aspects of the model, namely the correction step, and the RNN-3 memory store, are not adequately described, rely on decisions that are not adequately justified, and their properties/significance are not adequately investigated. Thus, while the authors did demonstrate the emergence of spatial representation and the utility of their model, their presentation did not adequately support their conclusions. With significant revisions to the text and additional experiments/analysis, this work will have a significant impact on the field, and their model will be of further use to the community.

My major concern is that two critical aspects of the model, namely the correction step, and the RNN-3 Memory store, are not adequately described, rely on decisions that are not adequately justified, and their properties/significance are not adequately investigated, as discussed below.

Correction step

- In the results, the correction step is minimally described. However, the method is fairly involved. For example, lines 81-82 state that "visual information being communicated only by the activation of slots in the memory stores (Fig 1B)". Similar descriptions are given in lines 102-103 and 125-126. However, the nature of these predictions is not stated in the results or well-diagrammed in Figure 1B. It might help to specify, for example in the figure legend, that further details about this step are provided in Supplementary figure 1. As this is a crucial piece of the model, I recommend that at least a few more sentences be given to this step in the results, which outlines the high-level details of the correction step.

- In the methods, the description of the correction step is inadequate, it's given simply as G(x,x). While this may be appropriate for a machine learning conference proceeding, it's not appropriate for a general journal. The authors should include equations that specify G (as well as F), which could be included in the section "Sigmoid-LSTM and Sigmoid-Vanilla". Further, the authors might want to justify the need for an entirely new RNN cell, rather than another input to the existing RNN. In lines 318-319: "each x~ can be thought of as the result of a weighted reactivation of the RNN memory embeddings by the current visual input." It might be useful to explain the correction code as: "the expected RNN activation given the current visual input's activation of the memory cells".

- Lines 125-126 state that: "RNN-3 received no self-motion inputs, thus being dependent on temporal coherence, and corrections from mispredictions as its sole input". It's unclear why the corrections to this RNN are generated from "mispredictions", and not just visual "corrections", like in the other RNNs. Further, nothing in the implementation of the correction step enforces that it gives "corrections", only that it learns to incorporate information from the current visual input, via the memory store, to the action of the RNNs. They're just occasional information that the network learns to use to update the RNN state as best as possible. While this is presented as a correction, it's unclear what this RNN actually does. Does it learn to simply replace the existing x with what it should be from the memory store (i.e. a correction)? Or does it combine information from x^hat and x^tilde in some complicated way? To understand this, I recommend the authors could compare x^, x~, and x. During the correction step.

- Finally, the authors state that (Line numbers missing), "to correct for the accumulation of integration errors, the RNNs must incorporate positional and directional information from upstream visual inputs as well. This correction step should not be performed at every time step, or the integration of velocities would be unnecessary; in our experiments, it was performed at random timesteps with probability Pcorrection = 0.1." This entails a claim that for Pcorrection=0, errors will accumulate, while for Pcorrection=1, the integration of velocities will be "unnecessary". While this makes intuitive sense, no empirical justification for these claims is shown, and their implications for the model's function and representation are not demonstrated. I would suggest that the authors compare a range of Pcorrection values, for example, p=[1, 0.3, 0.1, 0.03, 0.01], and demonstrate how the network performance and spatial representation vary as a function of Pcorrection. Finally, though less important, it's unclear why this correction is probabilistic. This decision could be justified, e.g. with an experiment comparing the results of probabilistic versus deterministic/periodic corrections.

RNN-3 and Memory store<br /> This seems like a key feature of the model, yet its implementation gets very little attention in the results, and the description is conflicting and difficult to understand.

- Line 142 states that "the allocentric representations of RNN-3 were stored in the external memory slots as a second set of targets - being reactivated at each time step by comparison to the current state of the RNN-3". However, it's unclear what's meant by a "second" set of targets, or why this is unique to RNN-3. From the text, it seems that this could either refer to m(x)_3 (the memory map corresponding to RNN3), or s (the slots). However, from my interpretation of the methods as written, the m(x) parameters are learned, and s are activated by the joint activity of all three RNNs, not just RNN-3 (Equation 4). Why is this written as if it's a separate group of slots unique to RNN-3?

- Further, how is the activity of memory slots assessed? While I can imagine (though not found in the method), how the tuning curves of RNN-1-3 are calculated, because of the confusion with what this set of targets refers to I don't know how e.g. Figure 2E was calculated. I recommend this be included in the methods. Importantly, I recommend the authors expand the description of RNN-3 and its associated memory store in the results, and clarify its description in the methods section.

- Lines 320 and 322 states that the memory store contents corresponding to the RNNs m(x) are optimized parameters, while those corresponding to upstream inputs m(y) are not. However, Line 325 states that all contents are chosen and assigned (m(y), m(x)) := (y, x).

- Finally, no justification was given as to why RNN-3 was added. The authors justify the addition of RNN-2 by stating that "a single RNN receiving all the velocity inputs did not develop the whole range of representations" (Line 101). However, no justification is given for a third RNN that receives no input. As this is a key piece of the results, justifying and understanding its contribution is critical. Does this affect predictive performance, the ability to generalize to new environments, or utility for RL, or is it simply adding a representational similarity to hippocampal place fields and egoBVCs? I recommend that the authors show the results of a network with only RNN-1 and RNN-2, to justify the addition of RNN-3 and demonstrate its utility for prediction.

On the head direction attractor analysis<br /> - Lines 174-176 state, "To investigate how our model incorporates visual information in its representation of heading, we simulated the input of visual corrections (512 images from the training environment), However, this experiment does not tell you "how the model incorporates visual information", but only the response to selected images. The intuitive idea is that the network learns to map distal cues to specific angles, but not ambiguous images. To test this hypothesis, I would recommend that the authors compare the heading direction of the visual correction input to the direction on the attractor activated, i.e. to show that images that give an attractor point match the heading of that image. Further, because the corrections are given through an entirely different RNN cell (G), from that which (presumably) holds the attractor (F), I would recommend that the authors show how the correction input to G interacts with an existing action-driven point on the attractor via F. For example, what if an image is shown that disagrees with the current heading direction?

On the RL agent<br /> - Lines 200-202 state that "self-consistency is an adaptive characteristic allowing spatial behaviour learned in one environment to be quickly transferred to novel environments", and Line 223: "the spatial responses present in the SMP's RNN support rapid generalization to novel settings." While they've shown that the SMP can support RL and generalization, they haven't tested whether its spatial tuning is responsible for the performance. One way they could test this is to replace the SMP input to the RL agent with equivalent rate tuned units as inputs (whose rate is simply what would be expected from the tuning curve of each RNN unit). This experiment could be done for the pre-trained agent (to see if performance is maintained from the tuning curves alone, or if there's more information in the SMP that's being used), and possibly compared to a newly-trained agent.

namjoon's internalized homophobia fr.. he's like,, six years old, and its already ingrained into him that boys shouldn't like other boys :(

reminds me of elementary school :(

This math is far and away over my head, thankfully we employ some smart engineers with math degrees. But it is good to see these equations exist and can be used to help us figure out the best methods for our project. I will not be highlighting all of the equations, there are a lot of them.

Note: This rebuttal was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

Reply to the reviewers

Response to the reviewers

Manuscript number: RC-2022-01407

Corresponding author(s): Ivana, Nikić-Spiegel

1. General Statements

We would like to thank the reviewers for careful reading of our manuscript and for their insightful and useful comments. We are happy to see that the reviewers find these results to be of interest and significance. The way we understand reviewers’ reports, their main concerns can be roughly divided in following categories: 1) providing more quantitative data 2) interpretation of the Annexin V/PI assay 3) additional evidence for calpain involvement. We intend to address these experimentally or by modifying the text, as outlined below.

2. Description of the planned revisions

Reviewer #1

Fig1A/B o SYTO 16 staining suggests slight reshaping of nucleus upon spermine NONOate, showing less blurry punctae. From the SYTO 16 profile, this should be quantifiable.

By looking at the shown examples and the entire dataset, it appears to us as if neuronal nuclei are shrinking upon spermine NONOate treatment resulting in their less blurry appearance. We are not sure if this is what the reviewer is referring to, but this can also be quantified by measuring changes in neuronal nuclear size. We already have this data from the measurements shown in Fig4 and we intend to show it in the revised version of the manuscript. Line profile measurements are also possible, but the nuclear size quantification might be more suitable for this purpose.

o There is a subset of neuron nuclei that are SYTO 16 positive. Please quantify the ratio

We will use our existing dataset to quantify the ratio of NFL positive and SYTO16 positive nuclei.

FigS1A o Show NeuN with Anti-NFL merged figures

We will show merged NeuN and anti-NFL images, which might require rearrangement of the existing figures and figure panels. We will do this in the revised manuscript.

FigS1C o Show quantification and timeline. I want to know whether there is also a plateau reached here.

As the data shown in the FigS1C do not include NeuN staining, we will do additional experiments and perform proposed quantifications.

FigS2A-F o Though the statements might be true, selecting one nucleus for a line profile as a statement for the whole dataset seems problematic. Average a larger number of unbiased selected nuclei profiles across multiple cultures to make a stronger statement, or a percentage of positive nuclei as in FigS1b.

Corresponding images and line profiles are representative of the entire dataset. However, we agree with the reviewer that this is not obvious from the current manuscript version. Thus, to strengthen our findings, we intend to quantify the percentage of positive nuclei as in FigS1b. The only difference will be that instead of NeuN, we will use SYTO16 as a nuclear marker. The reason being that the existing datasets contain images of NFL and SYTO16 and not NeuN.

FigS3 • There are no fluorescence profiles, no quantification

As the reviewer suggests, we will quantify the ratio of NFL positive and SYTO16 positive nuclei, and include the quantifications in the revised manuscript.

General statement: There do seem to be punctated patterns of non-nucleus accumulating NFL fragments. Can they be localized to any specific structure?

We assume that the reviewer is referring to neuronal/axonal debris. They are present after injury but they do not colocalize with nuclear stains. We will address this in the revised manuscript.

Fig1C-F • I find it too simplistic to categorize c+f and d+e together. There is a huge difference in the examples of nuclear localization between d and e. To not comment on their distinction (if that is consistent) is problematic. Also, since we don't see a merge with either NeuN or SYTO 16, reader quantification is difficult.

We thank the reviewer for bringing this up. We will carefully check our entire dataset and we will update the figures and the text accordingly. We will also show the corresponding SYTO16 images, as the reviewer suggested.

Would the microfluidic device construction allow for time to transport any axonally damaged fragments to the soma?

Yes, the construction of the microfluidic devices allows the transport of axonal proteins back to the soma. Based on our experiments, it seems that damaged NFL from the axonal compartment could be contributing to the accumulation of NFL fragments in the nuclei. However, this contribution seems to be minimal as we cannot detect nuclear NFL upon the injury of axons alone. Alternatively, it could be that the processing of axonal NFL fragments proceeds differently if neuronal bodies are not injured and that this is the reason we don’t detect the NFL nuclear accumulation upon injury of axons alone. We will discuss this in the revised manuscript.

Fig2C+D • The statement ".... no annexin V was detected on the cell membrane" needs to be shown more clearly

We will modify figures to address this comment.

• Please provide merged AnnexinV/PI images

We will modify figures to address this comment.

• The conclusion about 2D, that nuclear accumulated NFL overlaps with PI is not supported by the example image shown. There are plenty of PI positive spots that are not NFL positive and even several NFL positive ones that do not have a clear PI staining. Please quantify and then show a very clear result in order to be able to suggest necrosis as the underlying process.

We are not sure if we understand the reviewer’s concern correctly. We will try to clarify it here and in the revised text. If necessary, we will tone down our conclusion, but the reason why not all of PI positive spots are NFL positive is most likely due to the fact that not all injured nuclei are NFL positive. We quantified in FigS1 that up to 60% of nuclei under injury conditions show NFL accumulations. That is why we are not surprised to see some PI positive/NFL negative nuclei. And the fact that there are some NFL positive nuclei which appear to be PI negative is most likely related to the fact that the PI binding is affected. In addition, upon closer inspection of NFL and PI panels in Fig2d it can be observed that NFL positive nuclei are also PI positive, albeit with a lower PI fluorescence intensity. We will modify the figure to show this clearly in the revised manuscript.

FigS5 C+D • If the case is made that nitric oxide damage induces necrosis, then why is it that the AnnexinV example of Staurosporine exposure (which induces apoptosis) looks similar to that of nitric oxide damage in Fig2d and necrosis induction with Saponin looks very different?

We thank the reviewer for bringing this up. We will try to clarify this in the revised manuscript. Regarding the specific questions, the most likely explanation why staurosporine treated neurons look similar to the ones treated with spermine NONOate is that in the late stages of apoptosis cell membrane ruptures and allows for the PI to label nuclei. This is probably the case here as illustrated by the nucleus in the middle of the image (FigS5c) that shows the fragmentation characteristic for the apoptosis. This is not happening in early apoptotic cells due to the presence of an intact plasma membrane. On the other hand, the reason why saponin treated cultures look different compared to spermine NONOate is that membranes are destroyed by saponin so that the PI can enter the cell. For that reason, there could have not been any AnnexinV binding to the membrane which would correspond to the AnnexinV signal of spermine NONOate treated neurons. As we will discuss below, we did not try to mimic spermine NONOate-induced injury with saponin treatment. Instead this was a control condition for PI labeling and imaging. We also used a rather high concentration of saponin which probably destroyed all the membranes which was not the case with spermine NONOate treatment. We intend to do additional control experiments to address this.

• Additionally, does necrosis induction with Saponin also cause NFL fragment accumulation in the nucleus? Please show a co-staining of them. Also, the authors want to make a claim about reduce PI binding in NFL accumulated necrotic cells. In these examples, the intensity of the nuclear stain of PI with Saponin looks dimmer than with Staurosporine. Are the color scalings similar? It might be that the necrotic process itself causes reducing binding of PI and is not related to the presence of NFL.

With regards to this question, it is important to note that Annexin V and PI imaging was done in living cells. To obtain the corresponding anti-NFL signal as shown in Fig 2c,d we had to fix the neurons, perform immunocytochemistry and identify the same field of view. We tried to do the same procedure after saponin treatment (Supplementary Figure 5d) but the correlative imaging was very difficult due to the detachment of neurons from the coverslip after the saponin treatment. For this reason, we could not identify the same field of view co-stained with NFL. However, other fields of view did not show NFL fragment accumulation. This could also be the consequence of the high saponin concentration that we used as we discuss above. We have also noticed the reduced intensity of PI binding in the nuclei of saponin-treated neurons. However, if the necrotic process itself reduces the binding of PI to the DNA, then all of the neurons treated with spermine NONOate would have an equally low PI signal. In our experiments, only the nuclei which contained NFL accumulations had a low PI signal, while the signal of NFL-negative nuclei was higher (as shown in Fig2d). We would also like to point out again that the saponin treatment was our control of the PI’s ability to penetrate cells and bind the DNA, as well as our imaging conditions, and not the control of the necrotic process itself. This is the reason why we didn’t go into details about neuronal morphology and NFL localization upon saponin treatment. We thank the reviewer for pointing this out since it prompted us to reevaluate what we wrote in the corresponding paragraph of the manuscript. We realized that the confusion might stem from our explanation of the AnnexinV/PI assay controls in the lines 196-198 (“Additional control experiments in which neurons were treated with 10 μM staurosporine (a positive control for induction of apoptosis) or with 0.1% saponin (a positive control for induction of necrosis) confirmed the efficiency of the annexin V/PI assay (Supplementary Fig. 5c,d).”). We will modify this portion of the text to clearly state that staurosporine and saponin treatments were controls of the AnnexinV and PI binding to their respective targets and not of the apoptosis/necrosis process. When it comes to the saponin treatment, our intention was only to permeabilize the membranes in order to allow PI penetration and DNA binding and not to induce necrosis or to mimic the effect of the spermine NONOate. We also intend to perform experiments with lower concentration of saponin to try to address this experimentally in addition to the text modifications.

Fig3d • Please show similarly scaled images from controls for proper comparison

We will show similarly scaled images of the control neurons so that they can be properly compared. They were initially not scaled the same for visualization purposes, but we will modify this in the revised manuscript.

• How do the authors scale the degree and kinetics of induced damage between application of hydrogen peroxide/CCCP and glutamate toxicity? Does glutamate toxicity take longer to affect the cell, not allowing enough time to accumulate NFL fragments in the nucleus?

It is challenging to scale the degree and kinetics of induced damage with different stressors. That is why we did not intend to do this. Instead we set different injury conditions based on the published literature. That is why can only speculate when it comes to this. In this regard, it can be that the glutamate toxicity takes “longer” to affect the cells even though it is very difficult to compare them on a timescale, especially when considering different mechanisms of action. We will discuss this limitation in the revised manuscript.

Fig4B • Some groups (like NO and NO + emricasan) have much larger numbers of close to 0 intensity, compared to the control group. Why?

We were wondering the same when we analyzed the data. The fact that our nuclear fluorescence intensity analysis picked up NFL signal in control neurons which had no nuclear NFL accumulation made us realize that the intensity measured in the nuclei of control group comes entirely from the out of focus fluorescence – from neurofilaments in cell bodies, dendrites and axons (an example can be seen in the FigS6). That is why we presented the corresponding data with a cut-off value based on the control signal (as mentioned in lines 238-240). Since the oxidative injury causes NFL degradation (not only in neuronal soma, but also neuronal processes), the overall fluorescence intensity of the NFL immunocytochemical staining is reduced in injured neurons. We can see that in all of our images. Consequently, there is no contribution of out of focus fluorescent signal to the measured fluorescence intensity in the majority of nuclei. Due to that, the nuclei without NFL accumulation (at least 40% of injured nuclei) will appear to have a close to 0 intensity of the fluorescent signal. We will discuss and clarify this additionally in the revised manuscript.

• Please add the ratio of above/below threshold (50/50 obviously in controls)

We will update the figure in the revised manuscript.

• The description of the CTCF value calculation seems a little... muddled? Several parameters are described whereas "integrated density" is not even used. Why not simply mean intensity of nuclear ROI-mean intensity of background ROI?

We included the integrated density in the description since it is measured together with the raw integrated density and can also be used for the CTCF value calculation. However, since we didn’t use it for the CTCF calculation, we will remove it from the corresponding section of the manuscript. We calculated the CTCF value instead of calculating mean intensity of the nuclear ROI - mean intensity of the background ROI, since the CTCF value also takes into account the area of the ROI and not just the mean intensity.

• Also, please tell me if the areas for nuclear ROIs change, as I noted for Fig1A/B

We will include this information in the revised manuscript.

• To make sure that one of the 3 experimental repeats didn't skew the results, please show the median fluorescence intensity for each individual experiment to clarify that the supposed effect is repeated across experiments.

We have already noticed that in the earliest of the three experiments overall fluorescence intensity was higher, but this was consistent across all the experimental groups and did not skew the results or affect the overall conclusion. However, we will double-check this and revise the figure.

• From the text "...and due to the NFL degradation during injury...": this seems to contradict the process? Either the NFL fragment accumulates in the nucleus or it is degraded during injury. And isn't the degradation through calpain what supposedly allows this fragment of NFL to go to the nucleus in the first place? I reckon that the authors are possibly trying to reconcile why there are many close-to-0 intensity nuclei in the NO and NO + emricasan groups, but I don't feel the explanation given here fits.

As we tried to explain in our response above, we think that the overall degradation of neurofilaments in neurons affects the fluorescence intensity originating from the out of focus neurofilaments. Therefore, the nuclei without NFL accumulation in injured conditions have a close to 0 fluorescence intensity. Additionally, we think that this is not an either/or situation, but that both degradation and nuclear accumulation of NFL happen simultaneously. We also think that degradation of axonal NFL and the transport of its tail domain to the soma will at least partially contribute to the accumulation in the nucleus. In any case, degradation and nuclear accumulation seem to be differentially regulated in individual neurons, as some of them show nuclear NFL accumulation and some not. Furthermore, calpain and other mechanisms could also cause NFL degradation up to the point at which these fragments can no longer be recognized by the anti-NFL antibody leading to the loss of signal. We will try to clarify this in the revised version of the manuscript.

Fig5 • Does the distribution of this GFP in B match any of the various antibody stainings of different NFL fragments? Perhaps this is still a valid fragment of NFL, just not picked up by any AB?

The GFP signal in B appears rather homogenous and it does not match any of the various antibody stainings of different NFL fragments. As the reviewer points out, this could also be a valid fragment of NFL fused to GFP that none of our antibodies is recognizing. We will clarify this in the revised manuscript.

• "... and was indistinguishable from the full277 length NFL-GFP." Based on what parameters?

We will clarify this in the revised text, but we meant in terms of overall neurofilament network and cell appearance, which is commonly used to test the effect of NFL mutations.

• The authors claim that b is different from d, but I am not convinced. I would like to see a time dependent curve from multiple cells showing a differential change in nuclear and cytosolic GFP signal.

As we also wrote in the manuscript, in the majority of neurons that were monitored during injury we were not able to detect an increase in the GFP fluorescence intensity in the nucleus. This is what prompted further experiments with NFL(ΔA461–D543)-FLAG. We will clarify this additionally in the revised manuscript and perform line profile intensity measurements to show the difference in nuclear and cytosolic GFP signal.

• Secondly, the somatic GFP intensity for NFL increases for full length NFL-GFP. How is this explained, if it is only a separation of NFL and GFP? If anything, GFP should float away. And if the answer is that NFL is recruited to the nucleus, you showed that inhibition of calpain activity partially prevents that. So, if calpain activity is necessary for the transport of NFL to the nucleus, then wouldn't it also cut the GFP from NFL before it reaches the nucleus?

We thank the reviewer for bringing this up and we apologize for the confusion. This can be explained by the fact that the images were scaled in a way that the GFP signal over time could still be seen easily (i.e. differently across different time points which we unfortunately forgot to mention in the figure legend). In the revised manuscript, we will either scale the images the same or we will alternatively show the displayed grey values in individual panels.

Fig6 • It is recommended to overlap the transfected cells with a stain for endogenous NFL to show that despite the absence of the FLAG-tag, there is still NFL.

We did not overlap the anti-NFL with anti-FLAG and SYTO16 staining, due to the space constraint and the intent to clearly show the overlap of FLAG and SYTO16 signals in the merged images above the graphs. However, the line profile intensity measurements were done in all three channels and show that despite the absence of FLAG, there is still NFL in the nucleus (Fig6b), or that both FLAG and NFL are present in the nucleus (Fig6d, NFL signal shown in gray). However, as this is not obvious and can easily be overlooked, we will show the endogenous NFL staining overlap in the revised version of the manuscript.

Fig7 • „ ...all disrupted neurofilament assembly...": this sounds like the staining for native NFL supposedly shows a distortion due to a dominant negative effect of the expression of these constructs? Please clarify.

Yes, we were referring to the disruption of neurofilament assembly due to a dominant negative effect of the expression of NFL domains. We will clarify this in the revised version of the manuscript.

Discussion: • The authors show that after overepression of the head domain only, it possibly passively diffuses into the nucleus even in the absence of oxidative injury. However, it seems to be suggested as well that the head domain would not be freely floating around if it wouldn't be for increased calpain activity as a result of oxidative injury in the first place. Therefore, a head domain fragment localized in the nucleus would still more prominently happen upon oxidative injury and interact with DNA through prior identified putative DNA interaction sites from Wang et al. Please comment.

That is correct. Upon injury and calpain cleavage, it is conceivable that a fragment containing the NFL head domain would also be present in the cell and could potentially diffuse to the nucleus and interact with the DNA. However, by staining injured neurons with an antibody that recognizes amino acids 6-25 of the NFL head domain, we were not able to detect an NFL signal in the nucleus (FigS2a,b). It could be that either the NFL head domain does not localize in the nuclei upon injury, or that the fragment localizing in the nucleus does not contain amino acids 6-25 of the NFL head domain. As the putative DNA-binding sites described by Wang et al involve 7 amino acids located in the first 25 residues of the NFL head domain, we would expect to detect it with the aforementioned antibody. However, as that was not the case we speculated that the interaction of NFL and DNA occurs differently in living cells, as opposed to the test tube conditions utilized by Wang et al. We will comment and clarify this in the revised version of the manuscript.

• Reviewer #2*

• Major Comments:

• The initial data presented in the paper is good, does response of oxidative damage with proper controls, testing the antibodies to NF-L and etc. (Fig. 1-Fig. 4). *

We thank the reviewer for their positive feedback.

1. The evidence for calpain involvement in NF-L cleavage during oxidative damage is missing. Provide the evidence for full length NF-L construct and deletion mutants transfected into cells by immunoblot for cleavage of NF-L, perform nuclear and cytoplasmic extract preparations and show that enrichment of the tagged cleaved NF-L fragment in nuclear fraction.

We thank the reviewer for their comments and suggestions. Since we saw in our microscopy experiments that calpain inhibition reduced the accumulation of NFL in the nucleus, and since it is known that NFL is a calpain substrate (Schlaepfer et al., 1985; Kunz et al., 2004 and others), we did not perform additional experiments to confirm the involvement of calpain in NFL degradation during injury. However, to strengthen our findings, we intend to perform the suggested experiments and include the results in the revised manuscript.

1. Show calpain activation during oxidative damage by performing alpha-Spectrin immunoblots identify calpain specific 150-kda Spectrin and caspase specific 120-kDa fragment generation in these cells. Also, calpain activation can be measured by MAP2 level alteration and p35 to p25 conversion. Without this evidence it's very hard to believe if the calpain activity is increased or decreased during oxidative damage and these markers are altered by using calpain inhibitors.

To confirm the calpain activation, we intend to perform anti-alpha spectrin and/or anti-MAP2 blots in lysates of control and injured neurons and include the results in the revised manuscript.

1. The premise that NF proteins are absent in cell bodies and present only in axons is not correct. It has been demonstrated by multiple investigators that NFs are present in the perikaryon and dendrites of many types of neurons (Dahl, 1983, Experimental Cell Research)., Dr. Ron Liem's group showed NF protein expression in cell bodies of dorsal root ganglion cells (Adebola et ., 2015, Human Mol Genetics) and also showed N-terminal antibodies for NF-L, NF-M and NF-H stain rat cerebellar neuronal cell bodies and dendrites (Kaplan et al., 1991, Journal of Neuroscience Research) when NFs are less phosphorylated. (Schlaepfer et al., 1981, Brain Research) show staining of cell bodies of cortex and dorsal root ganglion cell bodies with NF antibody Ab150, and Yuan et al., 2009 in mouse cortical neurons with GFP tagged NF-L.

We are not sure what the reviewer is referring to since we cannot find a corresponding section in which we claim that NF proteins are absent in cell bodies. We wrote the following “Anti-NFL antibody staining of neurons treated with the control compound showed the expected neurofilament morphology, that is, a strong fluorescence intensity in axons and lower intensity in cell bodies and dendrites (Fig. 1a)” in our results section (lines 119-121), but the claim we were trying to make there was that NF proteins are particularly abundant in axons. We will clarify this in the revised manuscript.

1. Quantifying NF-L signal or tagged NF-L fragment signals in the cell body by ICC has many problems and making conclusions. It's extremely difficult to have control over levels of proteins in transfected overexpression models and comparing two or three different constructs with each other by ICC. Not every cell expresses same levels of protein in transfected cells and quantifying it by ICC again has a major problem. This can be addressed if there are stable lines that express equal levels of protein in all cells that comparisons can be made. Under thesese circumstances validation of the hypothesis presented in the study has no strong direct evidence to demonstrate that calpain is activated and NF-L fragment translocate to the nucleus.

We agree that the results from overexpression-based experiments should be interpreted with caution as levels of expression vary between the cells. We intend to discuss this in the revised manuscript. However, we find it difficult to experimentally address this comment since we are not sure which specific experiments the reviewer is referring to. With regards to this, we would like to emphasize that most of the initial experiments in which we observed NFL accumulation in the nuclei of injured neurons were based on the ICC labeling of endogenous NFL and didn’t involve its overexpression. This includes labeling of endogenous NFL in various types of neurons, comparing the effects of different types of oxidative injury, as well as testing the effects of calpain inhibition on the observed nuclear accumulation (Figures 1-4; Supplementary Figures 1-6). We later resorted to the overexpression experiments in primary neurons (Figures 5-7; Supplementary Figure 7, 10) to gain more information about the identity of NFL fragment which was detected in the nucleus. Due to the low transfection efficiency of primary neurons, we performed an additional set of overexpression experiments in neuroblastoma ND7/23 cells (Figure 8; Supplementary Figures 8,9) and obtained similar results in a higher number of cells. We agree that having stable cell lines which e.g. express same levels of NFL domains would be a more elegant approach and we intend to make them for our follow-up studies, however the generation of said stable cell lines might be beyond the scope of this revision. Furthermore, looking at our data with overexpression of NFL domains in ND7/23 cells (Supplementary Figure 8,9), it appears to us as if different domains are rather homogenously expressed in different cells. While the expression levels might vary, it seems that they all show the same trend when it comes to their localization (which was the main point of those experiments).

1. The interpretation that NF-L preventing DNA labeling cells is misinterpretation. NFs have very long half-life compared to other proteins. Due to oxidative damage, DNA is degraded in the cells but NFs that have very long half-life you see as NFs rings in the dead cells. So, NFs do not prevent DNA labeling, but DNA or chromatin is degraded in dead cells.

We thank the reviewer for their useful insight. DNA degradation could certainly be the reason why we observe a lower fluorescence intensity of the propidium iodide fluorescence in the nuclei of injured neurons. We intend to discuss this in the revised manuscript. However, if the DNA degradation is the only reason for the lower PI fluorescence intensity, then the PI fluorescence intensity would be the same in all injured nuclei. In our experiments, we saw the reduced PI fluorescence intensity in nuclei that contained NFL accumulations and not in other nuclei. Additionally, we observed a reduction of SYTO16 fluorescent labeling of nuclei which contained accumulations of the NFL tail domain, even in the absence of oxidative injury. Due to these reasons we speculated that NFL accumulation in the nucleus might hinder nuclear dyes from interacting with the DNA. But this is only a speculation and we will try to clarify this further in the revised manuscript including alternative explanations.

Minor comments: 1. In the introduction on page 4 reference is missing for NF transport, aggregation and perikaryal accumulation (on line 93).

We will add a reference to the revised manuscript.

1. The statement in discussion on page 14 line 454 for Zhu et al., 1997 study is not accurate. It should be modified to sciatic nerve crush not spinal cord injury.

We will correct this mistake in the revised manuscript.

1. What is the size of the calpain cleaved NF-L tail domain? If you perform immunoblots on cell extracts treated with oxidative agents one would know it.

We will perform immunoblots on cell lysates and incorporate the corresponding results in the revised manuscript.

1. Authors could make their conclusions clear. This is particularly true for the experiments in Figure 4 panels c and d. It is very difficult to understand the conclusions of the experiments. First state the expectation and then described whether the expectation is true or different.

We will do as the reviewer suggested in the revised manuscript.

1. The ICC images are at extremely low magnification. They should be shown at 100x or 120x so that details of the cell body and the nucleus can be seen.

Our intention was to show larger fields of view and wherever appropriate insets, but we will try to improve this in the revised manuscript by either zooming in, cropping or adding additional insets with individual cell bodies and nuclei. In general, images were taken with an optimal resolution/pixel size in mind for any of the used objectives (60x/1.4 NA or 100x/1.49 NA) and we can easily modify our figure panels to show more details.

1. Oxidative damage leads to beaded accumulation of NF-L in neurites and axons. Authors should address this issue.

We will discuss this in the revised manuscript.

1. The combination treatment of the inhibitors (last 3 sets of the Fig. 4 b) has no statistical significance should be removed.

Actually, these differences were statistically significant (Supplementary Table 1). For clarity and as described in the figure legend (line 516: “The most relevant significant differences are indicated with an asterisk”) we showed only a subset of them on the graph, but we will change this in the revised manuscript.

1. Why only two antibodies recognize cleaved NF-L? If the antibodies at directed at tail region, they should recognize it unless the phosphorylated tail at Ser473 may inibit the antibody binding. In that case NF-L Ser473 specific antibody (EMD Millipore: MABN2431) may be used to test this idea.

This is a very good point that we also wonder about. Even if all antibodies are directed at tail region, exact epitopes are not described for all of them. That makes it also difficult for us to understand and speculate on this. However, we have already ordered the new antibody as suggested by the reviewer and we will experimentally test it.

**Referees cross-commenting**

I agree with the reviewer#1 about presenting the quantification data for the indicated figures to make conclusions strong and see how much of variation is there among sampled cells.

As discussed in our response to reviewer #1, we will provide additional quantifications.

3. Description of the revisions that have already been incorporated in the transferred manuscript

4. Description of analyses that authors prefer not to carry out

Reviewer #2, major comment 7. Authors could do chromatin immunoprecipitation (chip) analysis to identify NF-L binding sites on chromatin and perform gel shift assays to show NF-L tail domain binding to specific consensus DNA sequences.

We thank the reviewer for their suggestion. We are very interested in performing additional experiments and identifying the NFL binding sites on the DNA (either by chromatin immunoprecipitation or DamID-seq) and we intend to perform these experiments as soon as possible. Unfortunately, at the moment we do not have the expertise to perform such experiments in our lab. Instead, this type of follow-up project requires establishing a collaboration which is beyond the scope of this revision.

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

Referee #1

Evidence, reproducibility and clarity

Summary:

The manuscript presented by Arsić and Nikić-Spiegel investigates a physiological consequence when neurons in vitro are exposed to oxidative stress injury, specifically a supposed interaction of the tail subdomain of the neurofilament light chain (NFL), after cleavage of the full NFL protein by calpain.

General comments:

The conclusions the authors draw from individual non-quantified example images are sometimes seen to be too simplistic when the shown examples ask for a more thorough investigation, especially when specific merged images are not available. It is highly recommended that the authors use the available data to come to more comprehensive answers across the entire acquired dataset. This for instance happens only in figures 4 and 8 and should be extended to other figures as well. There is not necessarily doubt about the author's general claims, but convincing the reader requires showing the variability and effect size of the entire group beyond a single selected example.

If these more thorough quantifications continue to support the author's claims, then I find no objections for publication of this data.

Specific comments:

Fig1A/B - SYTO 16 staining suggests slight reshaping of nucleus upon spermine NONOate, showing less blurry punctae. From the SYTO 16 profile, this should be quantifiable. - There is a subset of neuron nuclei that are SYTO 16 positive. Please quantify the ratio

FigS1A - Show NeuN with Anti-NFL merged figures

FigS1C - Show quantification and timeline. I want to know whether there is also a plateau reached here.

FigS2A-F - Though the statements might be true, selecting one nucleus for a line profile as a statement for the whole dataset seems problematic. Average a larger number of unbiased selected nuclei profiles across multiple cultures to make a stronger statement, or a percentage of positive nuclei as in FigS1b.

FigS3 - There are no fluorescence profiles, no quantification

General statement:

There do seem to be punctated patterns of non-nucleus accumulating NFL fragments. Can they be localized to any specific structure?

Fig1C-F - I find it too simplistic to categorize c+f and d+e together. There is a huge difference in the examples of nuclear localization between d and e. To not comment on their distinction (if that is consistent) is problematic. Also, since we don't see a merge with either NeuN or SYTO 16, reader quantification is difficult. - Would the microfluidic device construction allow for time to transport any axonally damaged fragments to the soma?

Fig2C+D - The statement ".... no annexin V was detected on the cell membrane" needs to be shown more clearly - Please provide merged AnnexinV/PI images - The conclusion about 2D, that nuclear accumulated NFL overlaps with PI is not supported by the example image shown. There are plenty of PI positive spots that are not NFL positive and even several NFL positive ones that do not have a clear PI staining. Please quantify and then show a very clear result in order to be able to suggest necrosis as the underlying process.

FigS5 C+D - If the case is made that nitric oxide damage induces necrosis, then why is it that the AnnexinV example of Staurosporine exposure (which induces apoptosis) looks similar to that of nitric oxide damage in Fig2d and necrosis induction with Saponin looks very different? - Additionally, does necrosis induction with Saponin also cause NFL fragment accumulation in the nucleus? Please show a co-staining of them. Also, the authors want to make a claim about reduce PI binding in NFL accumulated necrotic cells. In these examples, the intensity of the nuclear stain of PI with Saponin looks dimmer than with Staurosporine. Are the color scalings similar? It might be that the necrotic process itself causes reducing binding of PI and is not related to the presence of NFL.

Fig3d - Please show similarly scaled images from controls for proper comparison - How do the authors scale the degree and kinetics of induced damage between application of hydrogen peroxide/CCCP and glutamate toxicity? Does glutamate toxicity take longer to affect the cell, not allowing enough time to accumulate NFL fragments in the nucleus?

Fig4B - Some groups (like NO and NO + emricasan) have much larger numbers of close to 0 intensity, compared to the control group. Why? - Please add the ratio of above/below threshold (50/50 obviously in controls) - The description of the CTCF value calculation seems a little... muddled? Several parameters are described whereas "integrated density" is not even used. Why not simply mean intensity of nuclear ROI-mean intensity of background ROI? - Also, please tell me if the areas for nuclear ROIs change, as I noted for Fig1A/B - To make sure that one of the 3 experimental repeats didn't skew the results, please show the median fluorescence intensity for each individual experiment to clarify that the supposed effect is repeated across experiments. - From the text "...and due to the NFL degradation during injury...": this seems to contradict the process? Either the NFL fragment accumulates in the nucleus or it is degraded during injury. And isn't the degradation through calpain what supposedly allows this fragment of NFL to go to the nucleus in the first place? I reckon that the authors are possibly trying to reconcile why there are many close-to-0 intensity nuclei in the NO and NO + emricasan groups, but I don't feel the explanation given here fits.

Fig5 - Does the distribution of this GFP in B match any of the various antibody stainings of different NFL fragments? Perhaps this is still a valid fragment of NFL, just not picked up by any AB? - "... and was indistinguishable from the full277 length NFL-GFP." Based on what parameters? - The authors claim that b is different from d, but I am not convinced. I would like to see a time dependent curve from multiple cells showing a differential change in nuclear and cytosolic GFP signal. - Secondly, the somatic GFP intensity for NFL increases for full length NFL-GFP. How is this explained, if it is only a separation of NFL and GFP? If anything, GFP should float away. And if the answer is that NFL is recruited to the nucleus, you showed that inhibition of calpain activity partially prevents that. So, if calpain activity is necessary for the transport of NFL to the nucleus, then wouldn't it also cut the GFP from NFL before it reaches the nucleus?

Fig6 - It is recommended to overlap the transfected cells with a stain for endogenous NFL to show that despite the absence of the FLAG-tag, there is still NFL.

Fig7 - „ ...all disrupted neurofilament assembly...": this sounds like the staining for native NFL supposedly shows a distortion due to a dominant negative effect of the expression of these constructs? Please clarify.

Discussion:

• The authors show that after overepression of the head domain only, it possibly passively diffuses into the nucleus even in the absence of oxidative injury. However, it seems to be suggested as well that the head domain would not be freely floating around if it wouldn't be for increased calpain activity as a result of oxidative injury in the first place. Therefore, a head domain fragment localized in the nucleus would still more prominently happen upon oxidative injury and interact with DNA through prior identified putative DNA interaction sites from Wang et al. Please comment.

Significance

This in vitro study, despite its acknowledged caveats, can provide novel support for the claim that calpain induced cleavage of the NFL may play a role in downstream gene expression in order to regulate a neural response upon oxidative injury. Further investigation into this topic may provide further understand of physiological gene expression through interaction with cleavage products as well as yield possible therapeutic targets for pathological conditions. This study therefore may be of interest to a broad audience.

I made a highlight here and this is public

Author Response

Reviewer #1 (Public Review):

The authors set out to consider more the role of the predator in predator-prey interactions, particularly from a collective locomotion aspect. This is an aspect which at times has been overlooked, with many theories, experiments and models focusing largely on the prey response, independent of how the predator behaves. The major strengths are the (1) excellent writing, (2) quality of the figures, (3) quantity of data, and (4) question tackled. The major weaknesses are (1) the volume of information (as a reader, it is quite hard to distil key points from the sheer volume of what has been presented), (2) the confined captive environment making it difficult to draw comparisons with a wild-type scenario, and (3) lack of clarity about the wider implications of the work outside of the immediate field.

We thank the reviewer for their thoughtful review and positive comments. To address the weaknesses highlighted by the reviewer, we have revised our manuscript throughout.

Reviewer #2 (Public Review):

The manuscript describes a laboratory-based predator-prey experiment in which pike hunt shiner fish as a way to gain insight into the selective pressures driving the evolution of collective behavior. Unlike the predictions of classical theoretical work in which prey on the edge of social groups are considered to be at highest risk of predation, the fish in the center of the school were primarily targeted by the pike. This is because the pike uses a hunting behavior in which it slowly moves to the center of the school, seemingly undetected, until it rapidly attacks prey directly in front of its snout. This study also differs from previous studies in that both the predator and prey motion are examined, and the success of predation attempts was precisely determined. While the study demonstrates why shiners would be under selective pressure to avoid the center of a school, I am not convinced that the results explain why shiners evolved to have schooling behavior.

The reviewer indeed highlights one of the main findings of our study, that fish closer to the group center are more at risk of being attacked by pike. They also give a proper account of its possible explanation, and highlight some of the main ways in which our study differs from previous work. The reviewer states that our results do not explain why shiners evolved to school. We agree and note that we also don’t claim this anywhere in the manuscript. Rather, we state our study provides important new insights about differential predation risk in groups of prey and highlight the important role of predator attack strategy and decision-making and prey response, with potential repercussions for the costs and benefits of grouping.

We have considerably revised our introduction to better explain the importance of understanding differential predation risk in animal groups (lines 36-50): A key challenge in the life of most animals is to avoid being eaten. Via effects such as enhanced predator detection (Lima, 1995; Magurran et al., 1985), predator confusion (Landeau and Terborgh, 1986), and risk dilution effects (Foster and Treherne, 1981; Turner and Pitcher, 1986), individuals living and moving in groups can reduce their risk of predation (Ioannou et al., 2012; Krause and Ruxton, 2002; Pitcher and Parrish, 1993; Ward and Webster, 2016). This helps explain why strong predation pressure is known to drive the formation of larger and more cohesive groups (Beauchamp, 2004; Krause and Ruxton, 2002; B. Seghers, 1974). However, the costs and benefits of grouping are not shared equally among individuals within groups, and besides differential food intake and costs of locomotion, group members themselves may experience widely varying risks of predation (Handegard et al., 2012; Krause, 1994; Krause and Ruxton, 2002). Where and who predators attack within groups not only has major implications for the selection of individual phenotypes, and thereby the emergence of collective behaviour and the functioning of animal groups (Farine et al., 2015; Jolles et al., 2020; Ward and Webster, 2016), but also shapes the social behaviour of prey and the properties and structure of prey groups. Hence, a better understanding of the factors that influence predation risk within animal groups is of fundamental importance.

And in the discussion now better explain the potential evolutionary consequences of the findings of our work (lines 456-466): Predation is seen as one of the main factors to shape the collective properties of animal groups (Herbert-Read et al., 2017) and has so far generally been seen as to drive the formation of larger, more cohesive groups that exhibit collective, coordinated motion (see e.g. Beauchamp, 2004; Ioannou et al., 2012; B. H. Seghers, 1974). Our finding that central individuals are more at risk of being predated could actually have the opposite effect, with schooling having a selective disadvantage and over time result in weaker collective behaviour and less cohesive schools. However, we do not deem this likely as selection is likely to be group-size dependent, as discussed above. Furthermore, our multi-model inference approach revealed that, despite more central individuals experiencing higher predation risk, being close to others inside the school was still associated with a lower risk of being targeted. As most prey experience many types of predators, including sit-and-wait predators and active predators that hunt for prey, the extent and direction of such selection effects will depend on the broader predation landscape in which prey find themselves.

Major strengths of the paper include the precise recording of the location and orientation of all fish at all times during the experiments. This indeed provides a rich dataset that can be used to search for the factors that predict the likelihood of attack and escape with higher statistical power.

The major concern I have about the manuscript is that the results somewhat contradict the aim of the paper as expressed in the introduction and discussion: that predator-prey interactions explain the emergent evolution of collective behavior. Figure 2C shows that fish in smaller clusters or those that were totally isolated experienced lower rates of predation and were not included in any subsequent analyses. This would suggest that shiners experiencing predation from pike would be under strong selection to avoid schooling behavior altogether. Can you compare the likelihood of predation for individuals in non-central school locations compared to individuals outside of schools altogether? It might be helpful to investigate whether other predators of shiners use predation strategies that target prey on the edge of the school to help explain why schooling could be useful. Did the likelihood of schooling decrease throughout the trials?

The reviewer makes a good point regarding the observation that pike tended to mainly attack individuals in the main school, questioning if this would result in a selective disadvantage for schooling. We would like to point out that this result is regarding the likelihood to attack an individual, not the likelihood for a successful attack. If we look at the later we find 5 out of 8 attacks away from the main school were successful, a ratio that is actually similar to that of the main school. More importantly, when wanting to understand how predation risk is linked to group size one needs to look at the per capita risk. If we do that for the group size we used in our study, despite a moderately elevated risk of being predated in a large group, the shiners in the main school still had considerably lower individual risk to be killed than those that occurred in small sub-groups or were alone. We would like to note that in our study the shiners did not really show proper fission-fusion behaviour and by far the majority of the time the shiners were in one large cohesive school. Therefore, we feel our dataset is not suitable for a proper investigation about the role of group size in predation risk.

We now clarify these points in the discussion (lines 467-471): While the finding that pike were more likely to attack the main school may also appear to indicate a selective disadvantage to school, calculating the per-capita-risk for each individual would actually reveal it is still safest to be part of the main school. Nevertheless, as the shiners in our study rarely exhibited fission-fusion dynamics we feel our dataset is not appropriate to make proper inferences about how predation risk is linked to group size.

We have also slightly extended the relevant sentences in the results to further clarify the clustering results (lines 144-150): We found that, by and large, the shiners were organised in one large, cohesive school at the time of attack and rarely showed fission-fusion behaviour (merging and splitting of schools) during the trials. Only occasionally there were one or two singletons besides the main school (25 attacks) or multiple clusters of more than two fish (12 attacks Figure 2C), which tended to exist relatively briefly (mean school size: 36.5 ± 0.8). In more than 80% of these cases, pike still targeted an individual in the main cluster (Figure 2C).

We now also provide more discussion about other predator types being likely to attack central prey (lines 343-354): That predators may actually enter groups and strike at central individuals is not often considered (Hirsch and Morrell, 2011), possibly because it contrasts with the long-standing idea that predation risk is higher on the edge of animal groups (Duffield and Ioannou, 2017; Krause, 1994; Krause and Ruxton, 2002; Stankowich, 2003). However, our finding is in line with the predictions of theoretical work that suggest that the extent of marginal predation may depend on attack strategy and declines with the distance from which the predator attacks (Hirsch and Morrell, 2011). Furthermore, increased risk of individuals near the centre of groups may be more widespread than currently thought. Predators not only exhibit stealthy behavioural tactics that enable them to approach and attack central individuals, as we show here, but may also do so by attacking groups from above (Brunton, 1997) or below (Clua and Grosvalet, 2001; Hobson, 1963; but see Romey et al., 2008), and by rushing into the main body of the group (Handegard et al., 2012; Hobson, 1963; Parrish et al., 1989).

We furthermore discuss the potential role of group size on the observed effects (lines 441-455): In particular, while group size is not expected to effect much whether ambush predators are likely to attack internal individuals, the specific risk of central individuals could both be hypothesized to decrease with group size, such as if the predator is more likely to attack when surrounded by prey, or to not be affected by it, such as if the predator actively targets central individuals. Whatever the process, the observed findings are likely for prey that move in groups of somewhat intermediate size; for very large groups, such as the huge schools encountered in the pelagic, ambush predators may simply not be able to attack the group centre due to spatial constraints. More generally, the tendency for predators to attack the centre of moving groups may depend on the medium in which the predator-prey interactions occur. As in the air there is potential for (fatal) collisions, and on land it is physically difficult for predators to enter groups and predators’ size advantage tends to be more limited, predators may be less likely to go for the group centre as compared to in aquatic or mixed (e.g. aerial predator hunting aquatic prey) systems. Hence, the important interplay we highlight between predator attack strategy and prey response may have different implications across different predator prey systems and warrants concerted further research effort.

Finally, in response to the reviewer’s question if the likelihood to school decreased through the trials, we did not see a change in packing faction (median nearest-neighbour distance) with repeated exposure to the pike, but shiners increasingly avoided the area directly in front of the pike’s head (lines 182-186): While the shiners did not show a change in their packing fraction (median nearest-neighbour distance) with repeated exposure to the pike (F1,52 = 1.81, p = 0.185), they increasingly avoided the area directly in front of the pike’s head (Appendix 2 – Figure 1A) resulting in the pike attacking from increasingly further away (target distance: F1,52 = 45.52, p < 0.001, see Appendix 2 – Figure 1B,C). See also further Appendix 2.

I am also curious whether tank size affects the behavior of the fish, both of the shiners and the pike. The pike seem to be approximately 1/3 the shortest length of the tank, and 6 inches of depth have constrained the movement to be mostly in the 2D plane. A lack of open space might limit the pike's ability to hunt in any way other than this stealthy strategy. Has this stealthy hunting strategy been described in other experiments in larger or more naturalistic conditions? Does open space affect the shiners' propensity to school? Although the manuscript describes that shiners tend to school near the surface of water, does the shallow depth affect the pike's behavior? The manuscript states that some pike never attacked -- were these the largest in the study?

While the tank is small relative to the real world, we actually decided on this size of ~2m2 based on previous experimental work on predator-prey dynamics. As we stated in the methods of the original manuscript (lines 543-545) we expect that if a much larger space would have been used, pike would actually still show the same approach and attack behaviour linked to their stealthy attack strategy. The stealthy hunting behaviour of pike and similar predators and their ability to thereby get very close to their prey has been described elsewhere (see e.g. references on lines 332-344 of the original manuscript).

We now better explain the potential limitation of the arena size in the discussion (lines 472-480): Laboratory studies on predator-prey dynamics like ours do, of course, have their limitations. Although the size of the arena we used (~2m2) is in line with behavioural studies with large schools of fish (e.g. Sosna et al., 2019; Strandburg-Peshkin et al., 2013) and experiments with live predators attacking schooling prey (Bumann et al., 1997; Magurran and Pitcher, 1987; Neill and Cullen, 1974; Romenskyy et al., 2020; Theodorakis, 1989), compared to conditions in the wild the prey and predator had limited space to move. However, as pike are ambush predators they tend to move relatively little to search for prey and rather rely on prey movement for encounters (Nilsson and Eklöv, 2008). Increasing tank size would have made effective tracking extremely difficult, or impossible, and while a much larger tank is expected to considerably increase latency to attack, we expect it to have relatively little effect on the observed findings.

We agree that the shallow depth of the tank is a limitation of our study and may have somewhat restricted the pikes’ natural behaviour, although pilot experiments showed that the pike exhibited normal movements and attack behaviours. Fish were tested in very shallow water to be able to acquire detailed individual-based tracking of the schools as well as compute features related to the visual field of the fish. We would also like to note that both shiners and pike can often be found in the littoral zone and come in very shallow water of only a few 10s of cm (see e.g. Krause et al., 2000b; Pierce et al., 2013; Skov et al., 2018), with some experimental work furthermore showing that pike may actually prefer shallow water (Hawkins et al., 2005). We don’t think that increasing the depth of the tank would have considerably changed the predatory behaviour of the pike, as the pike would be expected to still use their stealthy approach to get close to their prey even if the prey school would be more three-dimensional.

We now provide a much more extensive discussion of the limited depth used in the discussion (lines 480-494): In terms of water depth, fish were tested in relatively very shallow water. This was primarily done to be able to keep track of individual identities and compute features related to the visual field of the fish. Shiners naturally school in very shallow water conditions as well as near the surface in deeper water in the wild (Hall et al., 1979; Krause et al., 2000b; Stone et al., 2016) and also pike primarily occur in the shallow littoral zone, sometimes only a few of tens of cm deep (Pierce et al., 2013; Skov et al., 2018). Furthermore, pilot experiment showed the pike did exhibit normal swimming and attack behaviour with attack speeds and acceleration comparable to previous work (Domenici and Blake, 1997; Walker et al., 2005). Recent other work on predator-prey dynamics did not find a considerable impact of adding the third dimension to their analyses (Romenskyy et al., 2020). Still, the water depth used is a limiting factor of our study and in the future this type of work should be extended to deeper water while still keeping track of individual identities over time. We expect that adding the third dimension would not change the stealthy attack behaviour of the pike and therefore still put more central individuals most at risk, but possibly attack success would be reduced because of increased predator visibility and prey escape potential in the vertical plane, which remains to be tested.

We did not observe a relationship between pike size and tendency to attack.

Reviewer #3 (Public Review):

While it has long been clear that animals in groups (e.g., fish schools) benefit in terms of safety in numbers, there has also been a keen interest in which animals in the group are at higher versus lower risk (e.g., those in front, or along the edges) and how that might depend on the predator's attack strategy. This study addresses these important predator-prey details using a common predatory fish (northern Pike) attacking schools of prey fish (golden shiners). A strength of the study is that it uses cutting-edge video tracking and computational/statistical methods that allow it to quantify and follow each fish's (1 predator and 40 prey in a group) spatial position, relative spacing, orientation and even each individual's visual field and movement throughout each of 125 attacks. Most (70%) of these attacks were successful, but many were not. The variation in attack success allowed the investigators to do statistical analyses to identify key predator and prey behaviors that are associated with successful vs. unsuccessful attacks.

The study yielded numerous interesting insights. While conventional wisdom pictures predators initiating an attack from outside of the group thus putting individuals at the group's edge at greatest risk, this study found that pike typically approached the school of prey headon both in terms of the group's orientation and direction of movement, and often stealthily moved within the group before initiating an attack. To understand which prey individual was targeted by the predator, the highly quantitative video analyses examined 11 measures of each individual prey's position and orientation at the time that the pike initiated its attack. Of course, pike showed a strong tendency to target one of the 3 closest prey, particularly prey that were more or less directly in front of the pike. However, contrary to conventional wisdom, the analysis showed that targeted prey were closer to the center than the edge, and that an individual's position and orientation relative to other nearby prey also played an important role in whether it might be targeted by the predator. Not surprisingly, analyses showed that targeted prey were more likely to escape if they were further from the predator's head and if they exhibited higher maximum acceleration. Interestingly, during the actual strike, on average, the predator accelerated to a speed about 50% faster than the velocity of the targeted prey.

A limitation of the study (that the authors describe and discuss) is that it was conducted in a tank with no spatial refuges whereas in nature, pike are often found in areas with vegetation, and schools of prey can often potentially respond to the presence of a predator by moving towards refuge (e.g., vegetation). Also, the study was done in very shallow water (6 cm) -- likely shallower than many, if not most, natural predator-prey interactions for these species. In deeper water, the predator-prey interaction might be better analyzed in three dimensions (i.e., also accounting for variation in vertical height in the water), though the authors argue that this conventional idea is not necessarily true.

Overall, this study provides an impressive example of the use of modern technology and statistical analyses allows us to better describe and understand the fine-scale behaviors that affect an interaction of high importance for ecology and evolution.

We thank the reviewer for the care and attention put in their review and their detailed objective assessment of our study.

Regarding refuge use, it is true that in the wild pike are often found in areas with vegetation, but it is actually predominantly younger pike seeking refuge among vegetation from predators themselves, including from cannibalism by larger pike (see Skov & Lucas, 2018 Chapter 5). Vegetation is also used by pike as background camouflage rather than a refuge per se, but due to their elongated body and narrow frontal body pike are able to approach and ambush prey when no vegetation is available, as we show in our study. During pilot experiments we did provide pike with refuges, but as they never used them, and it would provide a hiding place for hiding, which would have considerably impacted our ability to investigate predation risk within the schools, no refuges were provided during the experiment.

We now added an explanation about not using refuges in the discussion (lines 495502): For our experiments we used a testing arena without any internal structures such as refuges. This was a strategic decision as providing a more complex environment would have impacted the ability of the shiners to school in large groups and would have led fish to hide under cover. Although studying predator-prey dynamics in more complex environments would be interesting in its own regard, it would not have allowed us to study the questions we are interested in about the predation risk of free-schooling prey. Furthermore, pilot experiments indicated that the pike never used refuges (consistent with previous work, see Turesson and Brönmark, 2004), so they were not further provided during the actual experiment.

Regarding the shallow depth of the tank, we now better acknowledge this limitation and explain our reasoning (lines 480-482): In terms of water depth, fish were tested in relatively very shallow water. This was primarily done to be able to keep track of individual identities and compute features related to the visual field of the fish. We would also like to note that both shiners and pike spent a lot of their life in the littoral zone and occur in very shallow water of only a few 10s of cm (see e.g. Krause et al., 2000b; Pierce et al., 2013; Skov et al., 2018). Although the limited vertical space may have restricted the pikes’ natural behaviour to some extent, they did exhibit normal swimming and attack behaviour with attack speeds and acceleration comparable to previous work (Domenici and Blake, 1997; Walker et al., 2005). We now better discuss the limitation of the shallow depth used in the discussion on lines 477-494 (see also our responses above).

Reviewer #3 (Public Review):

While it has long been clear that animals in groups (e.g., fish schools) benefit in terms of safety in numbers, there has also been a keen interest in which animals in the group are at higher versus lower risk (e.g., those in front, or along the edges) and how that might depend on the predator's attack strategy. This study addresses these important predator-prey details using a common predatory fish (northern Pike) attacking schools of prey fish (golden shiners). A strength of the study is that it uses cutting-edge video tracking and computational/statistical methods that allow it to quantify and follow each fish's (1 predator and 40 prey in a group) spatial position, relative spacing, orientation and even each individual's visual field and movement throughout each of 125 attacks. Most (70%) of these attacks were successful, but many were not. The variation in attack success allowed the investigators to do statistical analyses to identify key predator and prey behaviors that are associated with successful vs. unsuccessful attacks.

The study yielded numerous interesting insights. While conventional wisdom pictures predators initiating an attack from outside of the group thus putting individuals at the group's edge at greatest risk, this study found that pike typically approached the school of prey head-on both in terms of the group's orientation and direction of movement, and often stealthily moved within the group before initiating an attack. To understand which prey individual was targeted by the predator, the highly quantitative video analyses examined 11 measures of each individual prey's position and orientation at the time that the pike initiated its attack. Of course, pike showed a strong tendency to target one of the 3 closest prey, particularly prey that were more or less directly in front of the pike. However, contrary to conventional wisdom, the analysis showed that targeted prey were closer to the center than the edge, and that an individual's position and orientation relative to other nearby prey also played an important role in whether it might be targeted by the predator. Not surprisingly, analyses showed that targeted prey were more likely to escape if they were further from the predator's head and if they exhibited higher maximum acceleration. Interestingly, during the actual strike, on average, the predator accelerated to a speed about 50% faster than the velocity of the targeted prey.

A limitation of the study (that the authors describe and discuss) is that it was conducted in a tank with no spatial refuges whereas in nature, pike are often found in areas with vegetation, and schools of prey can often potentially respond to the presence of a predator by moving towards refuge (e.g., vegetation). Also, the study was done in very shallow water (6 cm) -- likely shallower than many, if not most, natural predator-prey interactions for these species. In deeper water, the predator-prey interaction might be better analyzed in three dimensions (i.e., also accounting for variation in vertical height in the water), though the authors argue that this conventional idea is not necessarily true.

Overall, this study provides an impressive example of the use of modern technology and statistical analyses allows us to better describe and understand the fine-scale behaviors that affect an interaction of high importance for ecology and evolution.

for the present token, with the considered head on the logit..

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

Referee #3

Evidence, reproducibility and clarity

Summary:

The manuscript describes how a rhythmically active transcription factor is important for molting cycles. The first part of the manuscript focuses on oscillating genes, and the authors nicely show a rhythmic transcription of these genes. Indeed, using RNAPolII Chip-Seq experiments, they show a rhythmic recruitment of the RNAPolII to the promoters of the oscillatory genes they have previously described. They then demonstrate that GFPs driven by the promoter of these oscillating genes, and inserted as a single copy, can very accurately recapitulate the rhythmic transcription of oscillating genes. It is interesting to see the weak impact of introns and 3'UTR on the rhythmic expression. In the second part of the manuscript, the authors perform an RNAi screen, looking for oscillating transcription factors (TF) important for molting. The goal of this approach is to identify core oscillators that could control molting cycles. Focusing their screen on oscillating TF allows them to exclude TF that would be required for embryonic or larval viability unrelated to molting. Among the 6 candidates they have identified, 5 have already been linked to molting. They focus their study on grh-1, which has never before been described during larval development. They characterize the molting phenotype of grh-1 defective worms using the AID degron system. They monitor the molting cycles using a luciferase assay in a liquid culture where transgenic worms for luciferase are grown in the presence of bacteria supplemented with luciferin. Using this approach (which is quantitative and allows high-throughput analysis), they show that grh-1 is required for each molt in a dose-dependent manner. The GRH-1 protein oscillates and peaks shortly before each molt entry, reinforcing the idea that GRH-1 is an important core TF for molting. The authors finally show that the oscillating activity of GRH-1 is crucial for the molting.

Major comments:

Overall, the data are clear and convincing, and the results are quantified with care. The first part of the manuscript represents a significant amount of work with the RNAPolII Chip-Seq, the different single-copy integrants and RT-PCRs. Then, the authors provide a quantitative assessment of the molting process by combining their grh-1-AID construct with the luciferase system. This strengthens the quality of the manuscript.

My substantial suggestion is that the authors could consider extending the scope of the study in two alternative ways:

• One question that immediately springs to mind is: what are the targets of grh-1? By applying their luciferase assay to further possible downstream targets of grh-1, they could attempt to phenocopy the grh-1 molting defect, and then look if the oscillating expression of these targets is eliminated in grh-1 defective animals. The binding site of grh-1 is apparently known (Venkatesan, 2003), so is it possible to reduce the potential target list among the oscillating genes using a bioinformatics approach? This requires a substantial amount of work (2 to 3 months) but it would help tell a more complete story.
• It is striking that myrf-1 and nhr-23 RNAi display the same molting defects as grh-1 RNAi as show in figure 2B. Have the authors considered testing the genetic interactions of grh-1 with these two other candidates? Do they belong to the same GRN? Does grh-1 depletion impact the expression of the nhr-23 and myrf-1, or vice versa? Do they have the same target genes (Chip-seq data for nhr-23 are available)? This, again, would significantly strengthen the paper and would make the second part more complete. This alternative piece of work would require less experiments than the first suggestion but would be also of great interest. These two points are only suggestions as it represents a significant amount of work, and the paper could very well be published in its current form.

About the luciferase assay for grh-1, nhr-23 and myrf-1 RNAi, the authors observe "an apparent arrest in development or death following atypical molts". What do they mean by "atypical molt" at this stage of the paper? Indeed, for these candidates, the luminescence traces are highly perturbed after the second molt (for grh-1 and nrh-23 RNAi) or the third molt (for myrf-1), but these abnormal traces seem to reflect an arrest in development or larval lethality rather than an atypical molting. Can the authors clarify this point?

In the part on the phenotypic analysis of GRH-1 depleted animals, the authors conclude the paragraph with "GRH-1 is required for viability at least in part through its role in proper cuticle formation". This role in proper cuticle formation refers to the cuticle break in the head region as observed in time lapse. It would be useful to have a visual test of the cuticle permeability using an Hoechst staining.

The authors show GFP::GRH-1 pictures at different stages to describe a rhythmic protein accumulation (see also my minor comment on GFP picture quality). From the perspective of whether all tissues are oscillating, it would be interesting to see if all the cells they mention in the text are showing the same rhythmic fluorescence.

In relation to the previous comment, which tissue is responsible for the defects observed by the degradation of GRH-1? Is it possible to use a tissue-specific depletion of AID-tagged GRH-1 using Seam-cell specific, rectal cell specific, vulval precursors specific promoters, etc...?

In the last part of the results, the authors show that molting requires oscillatory GRH-1 activity by depleting GRH-1 at variable times in L2. It would be interesting to know what happens if a stable (non-oscillating) amount of GRH-1 protein is maintained over time in the worms (using a non-oscillating promoter).

Minor comments:

In figure 5 B, C, D, it seems that right before entering the M2, M3 and M4 respectively, there is a peak of luminescence (a thin bright line) and a strong luminescent signal is detected at the molt exit. Can the authors comment on that?

If I understand correctly, for the GRH-1 GFP CRISPR reporter (Fig S7, S8 and S9), the authors have imaged single worms in microchambers on a spinning disk microscope. I fail to see why they used such a sophisticated approach to describe the expression pattern of GRH-1. This imaging setup is ideal for timelapse. However, in the context of which cell express GRH-1, the resolution is not good enough to fully assess cell identity. Indeed, the GFP images are a bit blurry, and it is difficult to make out the difference between real GFP fluorescence and gut autofluorescence. It would be helpful to have better quality pictures with a more regular setup, i.e., a 2% agarose pad mounted on a regular microscope or confocal. For non-specialists of the C. elegans anatomy, small insets for each category of cells mentioned (seam cells, vulval precursors etc.) would be appreciated.

It would be easier to assess the GRH-1 expression decrease in adults if the pictures were shown in parallel with the larval pictures, with the same brightness/contrast correction (if any). Make insets to compare the same cells between different stages.

How can the authors quantify the duration of molts 3 and 4 in fig S4 when these molts are not seen in the luciferase assay in fig 2B? Can the authors clarify this point?

Writing/clarity: For non-specialists, mention why the authors used a PEST sequence in their constructs and explain what the eft-3 promoter is (they mention it in the Luciferase assay, but it is not clear enough).

In Fig S4, make clearer that EV = MOCK.

In the methods, the authors refer to the we146 mutant strain, but they neither use it nor mention it in the body of the text. Producing such a mutant strain is great and it should be mentioned in the results, with an explanation as to why they are dying. Otherwise, it should be removed from methods.

In the Methods, the genotype of the strains is misleading. For example HW1372 : EG6699; xeSi... is not the regular way to write a C. elegans genotype. It should be written as: HW1360 xeSi131[F58H1.2p::GFP::H2B::Pest::unc-54 3', unc-119+] II; unc-119(ed3) III as the strain used to generate the MosSci insertion is described in the paragraph on Transgenic reporter strain generation.

For the GFP CRISPR strain, the authors write either GFP::GRH-1 in Fig S7, S8 and S9, grh-1::gfp::3xflag in the methods or GRH-1-GFP fusion in the results. The authors should homogenize the way they write this reporter strain. Whether it is an Nterminal or a Cterminal fusion will determine how they should label it.

Enlarge the font size for Fig S8 and 9 for the scale bar.

Significance

The author's prior analysis (Meeuse, et al 2020) showed that mRNA oscillations are coupled with developmental processes, including molting. The present paper extends that finding by showing that oscillating transcript levels are directly linked to a rhythmic recruitment of the RNAPolII on their promoter. Then Meeuse and her colleagues use the molting as a model system to access the importance of oscillating TF for rhythmic processes. Through an RNAi screen, they have identified 6 candidates involved in the molting process. One of the candidates, grh-1, is characterized further. They combine a quantitative-based analysis (luciferase assay) with a time and dose-controlled degradation of GRH-1 to clearly describe the impact of grh-1 depletion on molting. This time and dose control is very smart and key to their study. Overall, the paper adds some interesting piece of information to the field of rhythmic control of molting cycles, as it shows that oscillating transcription factors provide a developmental clock in this process. But this notion is not completely new, as it has been shown in other developmental processes like the circadian clock. Moreover, how molting cycles are controlled by GRH-1 remains to be elucidated.

My field of expertise is GNRs studies, the genetic of C. elegans, embryonic and larval development in C.elegans, timelapse and confocal imaging. I do not have expertise in Chip-Seq analysis.

Note: This rebuttal was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

Reply to the reviewers

To reviewer #1

Reviewer #1 (Evidence, reproducibility and clarity (Required)): Please see combined review below in the next section, Reviewer #1 (Significance (Required)):

This is a descriptive manuscript providing a few new insights into a well-recognized and biologically important phenomenon - the lymphatic endothelial cells have heterogeneous origins in different organs. Overall, the idea of Islet1 lineage contributes to regional lymphatic vessel formation during a particular developmental stage is exciting and proven with detailed and careful lineage tracing. The first observation that Islet1 lineage gives rise to cardiac lymphatic vessels was published by the same group in Dev Bio in 2019 so the novelty here is dampened, although the pharyngeal lymphatics and the exact time of these non-venous origin lymphatic vessels arise were not previously characterized - so the current manuscript does provide new important insights. Both the data quality and manuscript layout need improvements, especially when it comes to defining where Islet1 is expressed at all the stages and statistics. The following suggestions will deepen the scope of the manuscript:

First of all, we would like to express our appreciation to the reviewer for all the constructive comments. We carefully read the reviewer’s comments and discussed it. We agree with the reviewer that our manuscript needs improvements with changes in layout several additional experiments. We have also included several description and new immunostaining data (e.g., Isl1,VEGFR3 and LYVE1 co-staining), to confirm our new findings and highlight the importance of the current manuscript beyond our previous one in Dev Biol in 2019. We also have included detailed quantification methods, single-channel images with improved data resolution, and improved clarity of the manuscript.

Specific points were addressed as follows:

COMMENTS BY THE REVIEWER

1. Whether Isl1 lineage is independent of venous-derived endothelial cells.

2. This point is very important: the manuscript does not actually show Isl1 expression through the stages they are inducing. I would want to be sure that lymphatic endothelial cells at this stage don't express Isl1. Another way to get at this is to maybe use other second-heart fields or even broader mesoderm drivers that are known to be never expressed in endothelial cells to confirm the findings.

Response:

In our previous work (Maruyama et al, Dev Biol 452:134–143, 2019, Figure 2C), we demonstrated that Isl1+ lineages using Isl1-Cre mice did not contribute to endothelial cells in the cardinal vein and its branches (intersomitic vessels: ISVs), which had been thought to the primary and biggest sources of lymphatic endothelial cells (LECs). In this paper, we confirmed this finding using Isl1-MerCreMer mice with tamoxifen treatment at E8.5 (Figure 4J). We have scanned whole embryos and detected no eYFP+ cells in the cardinal vein or ISVs (the detailed quantification methods have been added in Methods section). Consistently, another group (Lioux et al., Dev Cell 52:350-363, 2020) re-evaluated this point using Isl1-Cre mice that the Isl1+ lineage contribute to endothelial cells of the cardinal vein only by less than 2%, which neither explains the abundant contribution of the Isl1+ lineage to coronary lymphatics (>50%) nor its restriction to the ventral heart. Based on these reports, we supposed that the Isl1+ lineage was independent of LECs derived from the cardinal vein and ISVs.

In the revised manuscript, we added new data showing thorough expression patterns of Isl1, Prox1, Flk1, and PECAM in the E9.0 to E11.5 pharyngeal arches and cardinal veins by immunostaining and presented them as Supplemental Figure 4. In these sections, we detected Isl1 and Prox1 expression with partial overlapping within the pharyngeal mesodermal core, whereas Isl1 was co-expressed with Flk1, or PECAM neither in vessel-like structures around the mesodermal core nor in the cardinal vein and their surrounding Prox1+/PECAM+ LECs (Supplemental Figure 4H’ and J’) confirming the independency. These findings have been described in the manuscript as follows:

To identify possible Isl1+ LEC progenitors, we investigated the expression patterns of Isl1, Prox1, and vascular endothelial markers (Flk1 and PECAM) by immunostaining sections of E9.0 to E11.5 pharyngeal arches and cardinal veins. Consistent with the previous report (Cai et al., 2003), Isl1 was abundantly expressed in the core mesoderm of the first and second pharyngeal arches corresponding to the CPM from E9.0 to E11.5 (Nathan et al., 2008), where Prox1+ cells also aggregated and partially overlapped with Isl1 signals (Supplemental Figure 4A, A’ C, C’ E, E’ G, G’ I, I’). By contrast, Flk1+ or PECAM+ cells were distributed mainly around the CPM and not expressed Isl1 (Supplemental Figure 4A, A’ C, C’ E, E’ G, G’ I, I’). Furthermore, Isl1 was expressed neither in the endothelial layer of the cardinal vein nor in surrounding Prox1+/PECAM+ LECs (Supplemental Figure 4B, B’ D, D’, F, F’, H, H’, J, and J’). Taken together with the result from Myf5-CreERT2 mice, these results indicate that Isl1+ non-myogenic CPM cells may serve as LEC progenitors independent of venous-derived LECs and the commitment to LEC differentiation occurs before E9.5 in the pharyngeal arch region. (Page 6-7, lines 187-200)

Regarding the use of other second-heart field drivers as the reviewer recommended,

Lioux et al., have already shown the contribution of Mef2c-AHF+ , which marked CPM-derived cells including second heart field, cranial musculatures and connective tissues(Adachi et al., 2020), to ventral cardiac lymphatics. We are also trying to introduce Mef2c-AHF-Cre mice, but it is unfortunately delayed due to the pandemic of COVID-19.

The author stated that Islet1 lineage gives rise to lymphatic endothelial cells via the Tie2 mechanism but did not elaborate on this part. What is the potential relationship between Islet1 and Tie2? Or Tie2 just serves as a pan-endothelial lineage marker here?

Response:

To clearly demonstrate the relationship between Isl1+ and Tie2+ lineages in facial lymphatics, we added schematic representation in Figure 6G, which showed the differential Tie2 expression in lymphatic vessels in the tongue and facial skin.

Related to point 1 and 2, it has been thought that almost all LECs are formed from cardinal vein-derived Tie2+ endothelial cells. However, we identified the presence of Isl1+/Tie2+ LECs in the tongue, which are apparently not originated from the cardinal vein. In previous reports using Tie2-GFP mice or in situ hybridization of Tie2, Tie2 was not detected in the developing LECs at E9.5, 11.5, 13.5, and E15.5(Motoike et al., 2000; Srinivasan et al., 2007). In adult mice, Tie2 expression in lymphatics was only observed in restricted regions (Morisada et al., 2005; Tammela et al., 2005). Taken together with our present data that the differentiation fate of Isl1+ CPM-derived LECs was determined between E6.5 and E9.5 (Figure 2-4, Supplemental Figure 3), Tie2 is supposed to be transiently expressed during LEC differentiation in the tongue from early Isl1+ CPM cells, although it remains difficult to identify the Tie2-expressing stage during non-venous LEC differentiation.

It will be an important future subject to identify the stage and implication of transient Tie2 expression in the lineage and, in this paper, we want to just note that the Tie2+ lineage does not always mean the derivation from cardinal vein endothelial cells.

This point has already been included in the manuscript as follows:

The present study further indicates that the LECs in the tongue are derived from Tie2-expressing cells among the Isl1+ lineage. Although it is unclear whether Isl1+-derived cells at the Tie2-expressing stage represent a venous endothelial identity, this result means that Tie2+ LECs are not equivalent to cardinal vein-derived LECs. (Page 10, 298-301)

The effect of lineage-specific Prox1 knockout is very descriptive, without any discussion of the potential biological function of such cellular origin heterogeneity. This part may be worth a few follow-up experiments in later embryonic stages or even in postnatal stages. The authors demonstrated that loss of Prox1 in Islet1 lineage decreases the number of lymphatic vessels and leads to lymphangiectasia, but whether this phenotype can be later compensated or shows any clinical impact was not proven. Therefore, the statement made in line 206 is questionable, and whether Islet1 lineage-derived lymphatic endothelial cells are dispensable/indispensable remains unclear.

Response:

We agree with the reviewer in that additional follow up experiments using later embryonic or postnatal stages will give an insight into the potential biological function of cellular origin heterogeneity. We are generating lineage-specific Prox1 knockout mice by treating Isl1-MerCreMer; Prox1fl/fl mice with tamoxifen at E8.5 to analyze phenotypes in the facial lymphatic vessels.

The layout of the manuscript needs to be reorganized: 1. Details in statistical methods and quantification logic were completely missing from the manuscript. For example, definitions of "a sample" (how many sections are taken from one biological sample and how many fields take from one section, etc.), "number of vessels per field", "diameters", and of what parameters the numbers were normalized to, etc. need to be described in the materials and methods section. For instance, it is not clear how "tomato+ lymphatic vessels per field/Vegfr3+ lymphatic vessels" was defined. First, what proportion of tomato+ cells need to colocalize with Vegfr3 expression cells in a specific vessel to make this vessel being determined as a "tomato+ lymphatic vessel"? Most data provided here are section immunostaining where "multiple vessels" are very likely coming from different cross-sections of one same vessel in the same field. Second, Vegfr3 can stain venous endothelial cells in earlier stages so the specificity of this marker can be controversial. These are some important technical aspects to include in the revised version. Figures needing more description in quantification methods include but are not limited to Fig 1H, 2H, 2P, 5K-R.

We have revised the statistical methods from the ratio of the count of the number to ratio of the area of lymphatic vessels in Figure 1H, 2H, P, and Supplemental Figure 3I to represent more precisely the contribution of Tomato+ cells to lymphatic vessels. We also added more detailed description of the quantification methods in ‘Materials and Methods’ section, as follows:

Quantification of the section and whole mount images

For the quantification of section immunostaining at E16.5 embryos, the average of two 16-μm-thick sections taken every 50 μm and 10x power field of views (0.42 mm2/field) for each anatomical part (the larynx, the skin of the lower jaw, the tongue, and the cardiac outflow tracts) were subjected to the analyses. In the facial skin, lymphatic vessels in superficial layers of dermis were subjected to the analyses. The middle sagittal sections, including the aorta, larynx, and tongue, which were selected as hall marks of midline, was chosen from created sections. The coronal sections, including both eyes, tongue, and olfactory lobes with left and right symmetrical features, was selected. For E12.5 embryos (Figure 4O), two 16-μm-thick sagittal sections taken every 50 μm, including the 1st and 2nd pharyngeal arches and outflow tracts, were subjected to analyses. The area and the number of cells were measured manually using ImageJ software. For the whole mount immunostaining of embryos and the heart, the whole samples were scanned every 20 μm and confirmed eYFP contribution to LECs (Figure 3) and cardinal veins (Figure 4J, and Supplemental Figure 4B, D, F, H, J). (Page 13, lines 402-416)

We also have tested expression patterns of VEGFR3 with Prox1 or LYVE1 as Supplemental Figure 1. At E14.5, VEGFR3 was widely co-expressed with Prox1 in the tongue, facial skin, and around the pulmonary artery (Supplemental Figure 1A-C’). At E16.5, VEGFR3 was co-expressed with LYVE1 in the tongue, facial skin, and around the pulmonary artery. Thus, we thought that VEGFR3 could be used as a marker of LECs in these cardiopharyngeal region.

This point has been included in the manuscript as follows:

Co-immunostaining of platelet endothelial cell adhesion molecule (PECAM) and vascular endothelial growth factor receptor 3 (VEGFR3), which we confirmed its co-localization with lymphatic vessel endothelial hyaluronan receptor 1 (LYVE1) at E14.5 and E16.5 (Supplemental Figure 1), revealed tdTomato+ LECs in and around the larynx, the skin of the lower jaw, the tongue, and the cardiac outflow tracts, at various frequencies, whereas no such cells were found on the dorsal side of the ventricles, which agrees with our previous study (Maruyama et al., 2019). (Page 4, lines 112-119)

Data resolution needs to be improved. The magnification of the figures in Fig 1-4 is not sufficient to demonstrate the marker colocalization as described in the texts. Single-channel images (such as the ones shown in Fig 5-6 but in higher magnifications) are also necessary to show the co-expression of markers.

Response:

There was a limit on the data capacity when submitting the manuscript. We were therefore obliged to reduce the quality of images and file size. We have revised the figures to add several higher magnification and single-channel images with improved data resolution throughout Figure 1-4.

The experimental design is not well-elaborated in the context. For example, the scientific logic of choosing a particular time point/stage for lineage-knockout induction or sample collection needs to be justified. Also, it seems that the authors are using fl/+ as control littermates in most of the experiments. Any specific reason favors using fl/+ heterozygous instead of fl/fl littermates without cre exposure, which is the more commonly used control sample in this kind of comparison, should be addressed.

Response:

Knockdown of Prox1 in the Tie2+ lineage has shown to cause an initial failure in specification of LECs at E14.5 with no appearance of lymphatics even at E17.5(Klotz et al., 2015; Lioux et al., 2020; Maruyama et al., 2019), indicating that the effect on lymphatic vessels would not be compensated even at E16.5. In addition, the systemic lymphatic network formation is almost completed at E16.5(Srinivasan et al., 2007), and the lineage trace was also evaluated at this stage. Thus, it was reasonable to compare the phenotype at E16.5.

This point has been addressed in the text as follows:

When Prox1 is knocked down in the Tie2+ lineage, an initial failure in specification of LECs was confirmed at E14.5 with a lack of LECs even at E17.5(Klotz et al., 2015; Lioux et al., 2020; Maruyama et al., 2019). Therefore, we compared lymphatic vessel phenotypes at E16.5, by which systemic lymphatics formation is normally completed(Srinivasan et al., 2007). (Page 7, lines 208-212)

In Prox1-flox(Prox1fl/+) mice, recombinant cells were labeled with EGFP(Iwano et al., 2012), as already described in the manuscript (Page 7, lines206-208). Therefore, the recombined cells can be visualized by EGFP expression in both heterozygous (fl/+) and homozygous (fl/fl) mice, which enables phenotype analysis referring to the recombined (knocked-out in fl/fl) cells. Importantly, these mice showed no specific phenotypes(Klotz et al., 2015; Maruyama et al., 2019). It is therefore reasonable to use heterozygous mice as controls to compare the phenotype appropriately. Although fl/fl littermates without cre exposure could usually serve as controls, they do not express EGFP in the Prox1 lineage, detracting from their utility(Klotz et al., 2015; Maruyama et al., 2019).

Some of the phrases are not clear in the text- either because of the writing style or because the corresponding figures failed to support the statements. These include but are not limited to lines 104-106, 122, 206, 226, and 228-233.

104-106: we crossed Isl1-Cre mice, which express Cre recombinase under the control of the Isl1 promoter and in which second heart field-derivatives are effectively labeled, with the transgenic reporter line R26R-tdTomato at E16.5.

Response:

We have re-phrased this sentence as follows:

we crossed Isl1-Cre mice, which express Cre recombinase under the control of the Isl1 promoter and in which second heart field-derivatives are effectively labeled, with the transgenic reporter line R26R-tdTomato and analyzed at E16.5, when lymphatic networks are distributed throughout the whole body. (Pages 4, lines 109-112)

122: After tamoxifen was administered at E8.5, tdTomato+ cells were broadly detected in the muscle in the head and neck regions at E16.5, indicating effective Cre-mediated recombination of the target gene.

Response;

We have re-phrased this sentence as follows:

After tamoxifen was administered at E8.5, tdTomato+ cells were broadly detected in the skeletal muscle in the head and neck regions at E16.5, indicating effective Cre recombination in CPM-derived musculatures. (Page 4, lines 130-132)

We have also included red arrowheads, indicating CPM-derived musculatures in Supplemental Figure 1.

206: These results suggested that defects in LEC differentiation and/or maintenance due to Prox1 deletion in the Isl1+ lineage were compensated for by other cell sources, probably of venous origin, in facial skin, but not in the tongue, resulting in impaired lymphatic vessel formation in the tongue.

Response:

We have re-phrased this sentence as follows:

These results suggested that defects in LEC differentiation and/or maintenance due to Prox1 deletion in the Isl1+ lineage were compensated for by LECs from other cell sources, probably of venous origin, in facial skin, but not in the tongue. (Page 7-8, lines 231-233)

226:　 Almost all of the LYVE1+/PECAM+ lymphatic vessels in the tongue were positive for eGFP in the Tie2-Cre;Prox1fl/+ heterozygous mice (Figure 6A and Supplemental Figure 3D), indicating that the majority of LECs derived from Isl1+ CPM cells developed through Tie2 expression in the tongue.

Response:

We have added new cartoon in Figure 6G to more clearly show the relation of Tie2 expression in Isl1+ lineages. Previous reports have used Tie2-Cre mice to show the vein-derived LECs (Klotz et al., 2015; Srinivasan et al., 2007) , because most of cardinal vein endothelium were composed of Tie2+ lineages. In our present study, in the tongue, most of the LECs were derived from Isl1+/Tie2+ lineages (Figure 1D, H, Figure 2D, Figure 4G, Figure 5B, N, Figure 6A and Supplemental Figure 3F, I). These data suggested that there was a group of Tie2+ lineages even though they are derived from non-venous Isl1+ lineages.

Reference needed for Myf5-Cre as a driver for Myogenic CPM in the results section. Response:

We have included several reference, as shown below:

(Harel et al., 2012, 2009; Heude et al., 2018)

1. Harel et al., Dev cell, 2009
2. Harel et al., PNAS, 2012
3. Heude et al., eLife, 2018

4. In discussion the reference to Pitx2-driven mesenteric lymphatic heterogeneity (Mahadevan et al 2014) is missing yet Islet1 has been shown downstream of Pitx2 (Davis et al 2008). The authors should discuss their findings of gut lymphatic heterogeneity in this context, considering that mediastinum is mesentery-derived.

Response:

Isl1+ CPM-derived LECs have been distributed to the anterior mediastinum and their relationship to mesenteric lymphatic vessels, which continuous with the thoracic duct in the posterior mediastinum, is currently unclear. However, since this paper is valuable for understanding the heterogeneity of the origins of LECs, we have included the indicated paper (Mahadevan et al., 2014) in ‘Introduction’ section to show gut lymphatic heterogeneity. (Page 3, line 67)

To reviewer #2

Reviewer #2 (Evidence, reproducibility and clarity (Required)):

The manuscript entitled "The cardiopharyngeal mesoderm contributes to lymphatic vessel development" identified a novel non-venous origin of craniofacial and cardiac LECs using genetic lineage tracing. Their results also revealed the spatiotemporal difference between CPM- and venous-derived LECs. Overall, the paper is well-organized and has certain implications for understanding lymphatic development. However, some issues still need to be improved:

First of all, we would like to express our appreciation to the reviewer for all the constructive comments. We carefully read the reviewer’s comments and discussed it. We agree with the reviewer’s comments to make the text easier to understand and emphasize what we really want to say.

Specific points were addressed as follows:

(1). Clearly, the introduction needs to be more concise and focused on the main questions you propose to answer and why these questions are important.

Response:

We have revised introduction section to be more concise and focus on the developmental process of lymphatic vessels and its relation to CPM. (Page 2-4, lines 41-103)

(2). In the discussion section, you should focus on how the questions have been answered and what they mean. And it would be rash to infer the role of LECs in lymphatic malformation. It would be helpful to validate the changes of CPM-derived LECs in LM patient samples.

Response:

We have revised the discussion section to be more concise. To demonstrate our findings more clearly, we have also revised and added some cartoons in Figure 6G and Figure 7.

(3). For the statistical analysis, all the quantitative data should be tested for statistical significance. There are several bar charts lacking P values.

Response:

We have included P values in the Figure legends.

Reviewer #2 (Significance (Required)):

This study enriched the contribution of CPMs to broader regions of the facial, cardiac and laryngeal lymphatic network and revealed the spatiotemporally difference between CPM- and venous-derived LECs, which provided some basic reference for understanding lymphatic vessel development.

To reviewer #3

Reviewer #3 (Evidence, reproducibility and clarity (Required)):

Short summary of the findings and key conclusions:

The work from Murayama and colleagues traces the ontogenetic origin of the endothelial cells of the lymphatic vessels in the head and neck region. Using the Cre-lox-based mouse genetics approach, they conclude that the lymphatic endothelial cells (LECs) in this region have mixed origin, with contributions from both the cardiopharyngeal mesoderm (CPM) as well as from cardinal vein. The lineage tracing study is buttressed by assaying LEC formation following selective deletion of the key LEC regulator Prox1 in CPM lineage.

First of all, we would like to express our appreciation to the reviewer for all the constructive comments. We carefully read the reviewer’s comments and discussed it.

Specific points were addressed as follows:

Major comments: 1) The key conclusions: LECs in the head and neck region derive from CPM. LECs in this region have mixed developmental origins. Both these conclusions are convincingly supported by the study. However, the work would greatly be strengthened by Pax3-Cre lineage tracing. This would complement the Isl1-Cre lineage tracing. As the authors observe, the LEC descendants of Isl1+ cells also appear to go through Tie2+ state. Therefore, Tie2-Cre study has not helped to delineate the LECs of CPM and cardinal vein origins. In this context, tracing with Pax3-Cre is likely to give a very clear picture of LEC origins.

Response:

We agree with the reviewer in that the data using Pax3-Cre mice will strengthen our manuscripts. Unfortunately, we could not find out researchers who had this line in our society in Japan. For using this line, we need to get cryo-recovered mice from Jaxon laboratory. It will take at least several months. Therefore it is not realistic for us to use Pax3-Cre mice in this work because of time limitation. Instead, we addressed this issue by rewriting the discussion on the possible complementation with the Pax3-Cre lineage by citing (Lupu et al., 2022; Stone and Stainier, 2019).

This point has been addressed in the text as follows:

A recent study has suggested that Pax3+ paraxial mesoderm-derived cells contribute to the cardinal vein and therefore venous-derived LECs originate from the Pax3+ lineage (Stone and Stainier, 2019). The same group has further argued that the Pax3+ lineage gives rise to lymphatic vessels on the trunk side through lymphangiogenesis(Lupu et al., 2022). Therefore, the Isl1+ and Pax3+ lineages may complement each other to form systemic lymphatic vessels. (Page 10, lines 314-319)

2) In addition, the article should be revised to include the number of sections and the number of cells counted per embryo in the Figure legend in each case. This will help assess how robust and reliable are the measurements.

Response:

We have revised the statistical methods from the ratio of the count of the number to the area in Figure 1H, 2H, P, and Supplemental Figure 3I to demonstrate more precisely the contribution of Tomato+ cells in lymphatic vessels. We also added more detailed description of the quantification methods in ‘Materials and Methods’ section, as follows:

Quantification of the section and whole mount images

For the quantification of section immunostaining at E16.5 embryos, the average of two 16-μm-thick sections taken every 50 μm and 10x power field of views (0.42 mm2) for each anatomical part (the larynx, the skin of the lower jaw, the tongue, and the cardiac* outflow tracts) were subjected to the analyses. In the facial skin, lymphatic vessels in superficial layers of dermis were subjected to the analyses. The middle sagittal sections, including the aorta, larynx, and tongue, which were hall marks of midline, was selected from created sections. The coronal sections, including both eyes, tongue, and olfactory lobes with left and right symmetrical features, was selected. For E12.5 embryos (Figure 4O), two 16-μm-thick sagittal sections taken every 50 μm, including the 1st and 2nd pharyngeal arches and outflow tracts, were subjected to analyses. The area and the number of Prox1+ cells were measured manually using ImageJ software. For the whole mount immunostaining of embryos and the heart, the whole samples were scanned every 20 μm and confirmed eYFP contribution to LECs (Figure 3) and cardinal veins (Figure 4J, and Supplemental Figure 4B, D, F, H, J). * (Pages 13, lines 402-416)

We have also included the number of eYFP+/Prox1+ cells among Prox1+ cells in the first and second pharyngeal in the Figure 4O legends as follows;

• (the number of eYFP+/Prox1+ cells (10.83 (mean) ± 1.249 (SEM)): Prox1+ cells (30.83 ± 4.549)) or E9.5 (the number of eYFP+/Prox1+ cells (2.833 ± 1.108): Prox1+ cells (35.50　±　5.847)). (Page 23, lines 684-686)*

Minor comments: 1) Several groups have contributed to the CPM literature. The citation of seminal works from Tzahor and Kelly groups is good, however, work from other groups has not been cited. For example, reports such as Heude et al and Grimaldi et al from Tajbakhsh group are very relevant to this work.

Response:

According to reviewer’s suggestion, we have included following references in the introduction section for the explanation of CPM derivatives. (P3, line 70)

1. (Heude et al., 2018)
2. (Grimaldi et al., 2022)

   2) It would help the reader if the authors explain the reasons for selecting specific regions, such as the tongue, and the skin of the lower jaw, for the study.

Response:

This is because many lymphatic vessels are distributed in these cardiopharyngeal area and these area is well known as anatomical parts where lymphatic malformation most often occurs. This has been mentioned in the manuscript as follows:

From a clinical viewpoint, head and neck regions contributed by the CPM are the most common sites of lymphatic malformations (LMs) (Page 3-4, lines 99-100)

3) The authors should consider presenting the wholemount images, such as those in Figures 3A and 3E for Figures 5 and 6. This would help assess the lymphatic vessel development in a holistic manner.

Response:

Although we tried to do the whole mount images of facial and tongue lymphatics, we could not succeed. Antibodies did not penetrate well on the tongue and, as for lymphatics of facial skin, their complicated morphology prevented clear visualization. Whole-mount imaging of the entire head was difficult for the same reason. In our experience, the antibody was useful for immunostaining of the early-stage embryos (up to E11.5) and the surface area of the heart, where lymphatic vessels were distributed on the epicardium. Even in the whole-mount heart, we have not succeeded in clear and estimable imaging of the vascular structure in the myocardium. Instead, we improved the quality of images and statistical comparisons in the revised manuscript, which we believe makes it more convincing.

Reviewer #3 (Significance (Required)):

The nature and significance of the advance for the field & the work in the context of the existing literature: Groups working in the domain of cardiopharyngeal mesoderm (CPM) have focussed on skeletal muscle and heart development. This pool is also known to give rise to skeletal tissues as well as blood vessel endothelium. A recent work Nomaru et al. (Morrow group, Nat Commun 2021) has identified a multi-lineage primed population in the cardiopharyngeal field. In this context, the work from Maruyama and colleagues highlights the versatility of CPM by providing evidence for the emergence of LEC from this multipotent pool. This complex developmental potential of CPM has implications to understand the evolutionary origin of CPM itself.

The connective tissues in the head/neck have mixed origins (Heude et al, 2018 and Grimaldi et al 2022 from Tajbakhsh group)- from CPM as well as neural crest. This work shows mixed origin for LECs. These works begin to put together the pieces of the puzzle of vertebrate head evolution. Jacob proposed evolution is tinkering. This appears to be true both at the molecular level as well as the cellular level. Head tissues appear to have been put together by exploiting varied sources.

The study is of broad interest to developmental biologists.

Reviewer: A developmental biologist with an interest in understanding the axial patterning of mesoderm early during mammalian development. Not an expert in lymphatic vasculature development.

References for the revision

Adachi N, Bilio M, Baldini A, Kelly RG. 2020. Cardiopharyngeal mesoderm origins of musculoskeletal and connective tissues in the mammalian pharynx. Development 147:dev185256. doi:10.1242/dev.185256

Cai C-L, Liang X, Shi Y, Chu P-H, Pfaff SL, Chen J, Evans S. 2003. Isl1 Identifies a Cardiac Progenitor Population that Proliferates Prior to Differentiation and Contributes a Majority of Cells to the Heart. Dev Cell 5:877–889. doi:10.1016/s1534-5807(03)00363-0

Grimaldi A, Comai G, Mella S, Tajbakhsh S. 2022. Identification of bipotent progenitors that give rise to myogenic and connective tissues in mouse. Elife 11:e70235. doi:10.7554/elife.70235

Harel I, Maezawa Y, Avraham R, Rinon A, Ma H-Y, Cross JW, Leviatan N, Hegesh J, Roy A, Jacob-Hirsch J, Rechavi G, Carvajal J, Tole S, Kioussi C, Quaggin S, Tzahor E. 2012. Pharyngeal mesoderm regulatory network controls cardiac and head muscle morphogenesis. Proc National Acad Sci 109:18839–18844. doi:10.1073/pnas.1208690109

Harel I, Nathan E, Tirosh-Finkel L, Zigdon H, Guimarães-Camboa N, Evans SM, Tzahor E. 2009. Distinct Origins and Genetic Programs of Head Muscle Satellite Cells. Dev Cell 16:822–832. doi:10.1016/j.devcel.2009.05.007

Heude E, Tesarova M, Sefton EM, Jullian E, Adachi N, Grimaldi A, Zikmund T, Kaiser J, Kardon G, Kelly RG, Tajbakhsh S. 2018. Unique morphogenetic signatures define mammalian neck muscles and associated connective tissues. Elife 7:e40179. doi:10.7554/elife.40179

Klotz L, Norman S, Vieira JM, Masters M, Rohling M, Dubé KN, Bollini S, Matsuzaki F, Carr CA, Riley PR. 2015. Cardiac lymphatics are heterogeneous in origin and respond to injury. Nature 522:62–67. doi:10.1038/nature14483

Lioux G, Liu X, Temiño S, Oxendine M, Ayala E, Ortega S, Kelly RG, Oliver G, Torres M. 2020. A Second Heart Field-Derived Vasculogenic Niche Contributes to Cardiac Lymphatics. Dev Cell 52:350–363. doi:10.1016/j.devcel.2019.12.006

Lupu I-E, Kirschnick N, Weischer S, Martinez-Corral I, Forrow A, Lahmann I, Riley PR, Zobel T, Makinen T, Kiefer F, Stone OA. 2022. Direct specification of lymphatic endothelium from non-venous angioblasts. Biorxiv 2022.05.11.491403. doi:10.1101/2022.05.11.491403

Mahadevan A, Welsh IC, Sivakumar A, Gludish DW, Shilvock AR, Noden DM, Huss D, Lansford R, Kurpios NA. 2014. The Left-Right Pitx2 Pathway Drives Organ-Specific Arterial and Lymphatic Development in the Intestine. Dev Cell 31:690–706. doi:10.1016/j.devcel.2014.11.002

Maruyama K, Miyagawa-Tomita S, Mizukami K, Matsuzaki F, Kurihara H. 2019. Isl1-expressing non-venous cell lineage contributes to cardiac lymphatic vessel development. Dev Biol 452:134–143. doi:10.1016/j.ydbio.2019.05.002

Morisada T, Oike Y, Yamada Y, Urano T, Akao M, Kubota Y, Maekawa H, Kimura Y, Ohmura M, Miyamoto T, Nozawa S, Koh GY, Alitalo K, Suda T. 2005. Angiopoietin-1 promotes LYVE-1-positive lymphatic vessel formation. Blood 105:4649–4656. doi:10.1182/blood-2004-08-3382

Motoike T, Loughna S, Perens E, Roman BL, Liao W, Chau TC, Richardson CD, Kawate T, Kuno J, Weinstein BM, Stainier DYR, Sato TN. 2000. Universal GFP reporter for the study of vascular development. Genesis 28:75–81. doi:10.1002/1526-968x(200010)28:23.0.co;2-s

Nathan E, Monovich A, Tirosh-Finkel L, Harrelson Z, Rousso T, Rinon A, Harel I, Evans SM, Tzahor E. 2008. The contribution of Islet1-expressing splanchnic mesoderm cells to distinct branchiomeric muscles reveals significant heterogeneity in head muscle development. Development 135:647–57. doi:10.1242/dev.007989

Srinivasan RS, Dillard ME, Lagutin OV, Lin F-J, Tsai S, Tsai M-J, Samokhvalov IM, Oliver G. 2007. Lineage tracing demonstrates the venous origin of the mammalian lymphatic vasculature. Gene Dev 21:2422–2432. doi:10.1101/gad.1588407

Stone OA, Stainier DYR. 2019. Paraxial Mesoderm Is the Major Source of Lymphatic Endothelium. Dev Cell 50:247-255.e3. doi:10.1016/j.devcel.2019.04.034

Tammela T, Saaristo A, Lohela M, Morisada T, Tornberg J, Norrmén C, Oike Y, Pajusola K, Thurston G, Suda T, Yla-Herttuala S, Alitalo K. 2005. Angiopoietin-1 promotes lymphatic sprouting and hyperplasia. Blood 105:4642–4648. doi:10.1182/blood-2004-08-3327

Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

Learn more at Review Commons

Referee #3

Evidence, reproducibility and clarity

Short summary of the findings and key conclusions:

The work from Murayama and colleagues traces the ontogenetic origin of the endothelial cells of the lymphatic vessels in the head and neck region. Using the Cre-lox-based mouse genetics approach, they conclude that the lymphatic endothelial cells (LECs) in this region have mixed origin, with contributions from both the cardiopharyngeal mesoderm (CPM) as well as from cardinal vein. The lineage tracing study is buttressed by assaying LEC formation following selective deletion of the key LEC regulator Prox1 in CPM lineage.

Major comments:

1. The key conclusions: LECs in the head and neck region derive from CPM. LECs in this region have mixed developmental origins. Both these conclusions are convincingly supported by the study. However, the work would greatly be strengthened by Pax3-Cre lineage tracing. This would complement the Isl1-Cre lineage tracing. As the authors observe, the LEC descendants of Isl1+ cells also appear to go through Tie2+ state. Therefore, Tie2-Cre study has not helped to delineate the LECs of CPM and cardinal vein origins. In this context, tracing with Pax3-Cre is likely to give a very clear picture of LEC origins.
2. In addition, the article should be revised to include the number of sections and the number of cells counted per embryo in the Figure legend in each case. This will help assess how robust and reliable are the measurements.

Minor comments:

1. Several groups have contributed to the CPM literature. The citation of seminal works from Tzahor and Kelly groups is good, however, work from other groups has not been cited. For example, reports such as Heude et al and Grimaldi et al from Tajbakhsh group are very relevant to this work.
2. It would help the reader if the authors explain the reasons for selecting specific regions, such as the tongue, and the skin of the lower jaw, for the study.
3. The authors should consider presenting the wholemount images, such as those in Figures 3A and 3E for Figures 5 and 6. This would help assess the lymphatic vessel development in a holistic manner.

Significance

The nature and significance of the advance for the field & the work in the context of the existing literature:

Groups working in the domain of cardiopharyngeal mesoderm (CPM) have focussed on skeletal muscle and heart development. This pool is also known to give rise to skeletal tissues as well as blood vessel endothelium. A recent work Nomaru et al. (Morrow group, Nat Commun 2021) has identified a multi-lineage primed population in the cardiopharyngeal field. In this context, the work from Maruyama and colleagues highlights the versatility of CPM by providing evidence for the emergence of LEC from this multipotent pool. This complex developmental potential of CPM has implications to understand the evolutionary origin of CPM itself.

The connective tissues in the head/neck have mixed origins (Heude et al, 2018 and Grimaldi et al 2022 from Tajbakhsh group)- from CPM as well as neural crest. This work shows mixed origin for LECs. These works begin to put together the pieces of the puzzle of vertebrate head evolution. Jacob proposed evolution is tinkering. This appears to be true both at the molecular level as well as the cellular level. Head tissues appear to have been put together by exploiting varied sources.

The study is of broad interest to developmental biologists.

Reviewer: A developmental biologist with an interest in understanding the axial patterning of mesoderm early during mammalian development. Not an expert in lymphatic vasculature development.

The first thing that popped into my head was a clickbait title or more like a pull you in strategy. A very interesting question that can get lots of replies that can have a variety of answers to. But trying to trust something so big and so secretive can be very hard for some. How can one trust the source?

Reviewer #2 (Public Review):

This work provides tools for the acquisition and analysis of human brain MRI data at the mesoscopic scale as demonstrated in the visual and auditory cortex. Magnetic resonance imaging is a key tool for noninvasive evaluation of human brain structure and function but has been traditionally hampered by low sensitivity and spatial resolution. This paper provides acquisition strategies to surmount several barriers to achieving mesoscopic-scale MRI data, including motion secondary to blood flow in small vessels, and analysis tools to characterize changes in MRI contrast along the complex surface of the cerebral cortex. These approaches provide a framework for more robust acquisition and analysis of mesoscopic MRI data in the living human brain, particularly at ultra-high field, and serve as useful tools for advancing human neuroscience with more detailed characterization of human brain structure and function.

Strengths:

The averaging of multi-echo gradient echo MRI data with orthogonal phase-encoding directions using minimum intensity images voxels provides a clever approach to leveraging information across multiple image acquisitions and offers a viable approach to mitigating spatial misregistration due to blood flow motion while boosting signal-to-noise ratio, which proves to be important at mesoscopic spatial resolution.

A major strength of this paper is the exquisite image quality and high spatial resolution attained in the living human cortex. The high quality of the 0.35 mm isotropic spatial resolution images with detailed segmentations of the cortex surrounding the calcarine sulcus and Heschl's gyrus, including demonstration of fine-scale cortical substructures such as the stria of Gennari and intracortical vessels, demonstrates the promise of such technology in characterizing cortical laminar architecture in the living human brain.

The comparison of T2* variations at different cortical depths in visual and auditory cortex provides a sound validation of the acquisition and analysis methods in reproducing known trends in anatomy in these different cortical regions.

The availability of the analysis tools and data through open-source software and data-sharing enables the widespread dissemination of such mesoscopic imaging data, which is difficult to acquire and not readily accessible on standard scanners.

Weaknesses:

Some of the methods demonstrated in the manuscript are not fully discussed or characterized in-depth, leading to a lack of clarity regarding how to place the technical advances in the context of existing methods. For example, the authors describe how the cortical patch flattening method has desirable distortion characteristics compared to a specific triangular mesh-based tool developed by (Kay et al., 2019), yet they do not demonstrate systematically how the local distortions induced by flattening a folded cortex may impact the representation of cortical metrics, particularly as a function of spatial resolution.

The authors claim that the acquisition and analysis methods developed in this paper represent a significant advance toward demonstrating mesoscopic scale imaging in the living human brain, yet they confine their analyses to cortical regions with well-defined differences in laminar architecture. The paper thus reads as a confirmation of largely well-known areal differences in myeloarchitecture in the human cortex, as opposed to the intended application of mesoscopic scale imaging toward uncovering subtle differences and cyto- and myeloarchitecture in areas of cortex where the laminar architectonic boundaries are less well-delineated and understood.

The paper focuses on mitigating motion artifact related to blood flow while largely glossing over the challenge of mitigating bulk head motion, which is a greater source of error for the long acquisitions required for mesoscopic-scale imaging. It would be valuable for the authors to provide more detailed information and insight into how bulk motion was mitigated in the presented data.

Reviewer #2 (Public Review):

Wang et al. elegantly exploit single-cell RNA-seq datasets to question the putative involvement of lncRNAs in human germ cell development. In the first part of the study, the authors use computational approaches to identify and characterize, from existing data, lncRNAs expressed in the germline. Of note, the scRNA-seq data used were generated from polyA+ RNAs, and thus non-polyadenylated lncRNAs could not be retrieved. Most of the lncRNAs identified in the germ cells and in the somatic cells of the gonads were previously unannotated. While this increases the catalog of lncRNA genes in the human genome, further characterization is needed to determine which fraction of these newly identified lncRNAs represent bona fide transcripts or transcriptional noise.

Differential expression analysis between developmental stages, sexes, or cell types led to several observations: (i) whatever the stage of development, the number of expressed lncRNAs is higher in fetal germ cells compared to gonadal somatic cells; (ii) there is a continuous increase in the number of expressed lncRNA during the development of the germline; of note, a similar, although the more subtle trend is observed for protein-coding genes; (iii) the developmental stage at which there is the highest number of lncRNA expressed differs between male and female germ cells. While convincing, the significance of these observations is difficult to assess. However, the authors remain prudent with their conclusion and are not over-interpreting their findings.

Interestingly, integrating lncRNA expression to classify cell types led to the identification of a novel population of cells in the female germline that had not been revealed by protein-coding gene only-based classification. The biological relevance of this population, which cluster with mitotic populations, remains to be demonstrated. Finally, by examining lncRNA biotype, the authors could demonstrate an enrichment, in the germ cells, of the antisense head-to-head organization (in relation to the nearby protein-coding gene) compared to other biotypes. Whether this is different from the general distribution of lncRNA should be discussed.

In the second part of the manuscript, Wang et al focus on one pair of divergent lncRNA-protein coding genes (LNC1845-LHX8). To document the choice of this particular pair, it would be informative to have its correlation score indicated in Figure 3C. The existence of this transcript was validated using female fetal ovaries, and its function was addressed in late primordial germ cells like cells (PGCLC) derived from human embryonic stem cells (hESCs). The authors have used an admirable set of orthogonal approaches that led them to conclude as to a role for LNC1845 in regulating in cis the nearby gene LHX8. They further went on to identify the underlying mechanisms, which involve modification of the chromatin landscape through direct interaction of LNC1845 with a histone modifier. Among the different strategies used (KO, stop transcription, overexpression), the shRNA-mediated knock-down is the only one to specifically address the function of the transcript itself, as opposed to the active transcription. The result of this experiment led the authors to conclude that the LNC1845 RNA is functional, a conclusion that is reinforced by the demonstration of physical interaction between the LNC1845 RNA and WDR5, a component of MLL methyltransferase complexes. The result of the KD experiment is however puzzling as RNAi has been shown not to be the method of choice for targeting nuclear lncRNAs (Lennox et al. NAR 2016).

Overall the functional investigation is convincing and strengthened by the inclusion of multiple clones for each approach, and by the convergence in the outcome of each individual approach. The depth of characterization is also remarkable. The analyses of the mechanisms at stake are somehow less solid, as there is less evidence demonstrating the involvement of the LNC1845 RNA and its interaction with WDR5.

Altogether, this study provides a convincing demonstration of the role of a lncRNA on the regulation of a nearby gene in the context of the germline. However, to have a better understanding of the functionality of lncRNA genes in general, it would be interesting to know whether other pairs of lncRNA-PC genes have been functionally investigated in this context, where no function for the lncRNA gene could be demonstrated. Negative results are highly informative and if so, these could be included in the manuscript.

Reviewer #2 (Public Review):

The authors use a combination of optogenetics and calcium imaging to assess the contribution of cortical areas (posterior parietal cortex, retrosplenial cortex, S1/V1) on a visual-place discrimination task. Head-fixed mice were trained on a simple version of the task where they were required to turn left or right depending on the visual cue that was present (e.g. X = go left; Y = go right). In a more complex version of the task the configurations were either switched during training or the stimuli were only presented at the beginning of the trial (delay).

The authors found that inhibiting the posterior parietal cortex and retrosplenial cortex affected performance, particularly on the complex tasks. However, previous training on the complex tasks resulted in more pronounced impairments on the simple task than when behaviourally naïve animals were trained/tested on a simple task. This suggests that the more complex tasks recruit these cortical areas to a greater degree, potentially due to increased attention required during the tasks. When animals then perform the simple version of the task their previous experience of the complex tasks is transferred to the simple task resulting in a different pattern of impairments compared to that found in behaviorally naïve animals.

The calcium imaging data showed a similar pattern of findings to the optogenetic study. There was overall increased activity in the switching tasks compared to the simple tasks consistent with the greater task demands. There was also greater trial-type selectivity in the switching task compared to the simple task. This increased trial-type selectivity in the switching tasks was subsequently carried forward to the simple task so that activity patterns were different when animals performed the simple task after experiencing the complex task compared to when they were trained on the simple task alone

Strengths:

The use of optogenetics and calcium-imaging enables the authors to look at the requirement of these brain structures both in terms of necessity for the task when disrupted as well as their contribution when intact.

The use of the same experimental set up and stimuli can provide a nice comparison across tasks and trials.

The study nicely shows that the contribution of cortical regions varies with task demands and that longer-term changes in neuronal responses c can transfer across tasks.

The study highlights the importance of considering previous experience and exposure when understanding behavioural data and the contribution of different regions.

The authors include a number of important controls that help with the interpretation of the findings.

Weaknesses:

There are some experimental details that need to be clarified to help with understanding the paper in terms of behavior and the areas under investigation.

The use of the same stimuli throughout is beneficial as it allows direct comparisons with animals experiencing the same visual cues. However, it does limit the extent to which you can extrapolate the findings. It is perhaps unsurprising to find that learning about specific visual cues affects subsequent learning and use of those specific cues. What would be interesting to know is how much of what is being shown is cue specific learning or whether it reflects something more general, for example schema learning which could be generalised to other learning situations. If animals were then trained on a different discrimination with different stimuli would this previous training modify behavior and neural activity in that instance. This would perhaps be more reflective of the types of typical laboratory experiments where you may find an impairment on a more complex task and then go on to rule out more simple discrimination impairments. However, this would typically be done with slightly different stimuli so you don't introduce transfer effects.

It is not clear whether length of training has been taken into account for the calcium imaging study given the slow development of neural representations when animals acquire spatial tasks.

The authors are presenting the study in terms of decision-making, however, it is unclear from the data as presented whether the findings specifically relate to decision making. I'm not sure the authors are demonstrating differential effects at specific decision points.

Reviewer #3 (Public Review):

Previous research has sought to understand the correlation between neuronal activity and decision-making in different regions of the cortex, and a plethora of cortical regions have been tested for necessity in a variety of decision-making tasks. For example, the posterior parietal cortex (PPC) has been shown to be necessary for visual decision-making tasks, whereas the retrosplenial cortex (RSC) is important in navigational planning. Although the necessity of different cortical regions in completion of certain discrimination tasks has yielded insights into these brain regions' roles, the previous experiences of each individual animal has not been explored. This raises the possibility that the previous experience and learned associations may affect how cortical areas process subsequent decision-making tasks.

To test this hypothesis, the authors used in vivo optogenetic activation of GABAergic interneurons to silence excitatory activity in the PPC, RSC, and S1 (a control, as S1 has not been shown to be involved in visual decision-making). They also employed 2-photon in vivo calcium imaging in head-fixed mice. A virtual Y-maze enabled the mice to decide to turn left or right to receive a reward. In the "simple task," mice had one rule - horizontal or vertical bars indicating whether the mouse should turn left or right to receive the reward. In the "complex task," there were two rules, A and B. Rule A was the exact same parameters as the simple task, whereas Rule B switched the left/right reward association with the horizontal/vertical bars. In some cases, there was an additional "complex task" where there was a delay between cue onset and decision making.

Inhibition of the PPC or RSC during the simple task resulted in small decreases in performance. Interestingly, inhibition of the PPC or RSC during more complex tasks, such as the delay task or switching task, resulted in much greater decrease in correct decisions. S1 inhibition decreased performance in the complex tasks, but not the simple task. Interestingly, when the animal was trained on a complex task (either the delay task or switching task) prior to testing in the simple task, there was a greater decrease in performance upon inhibition of the PPC or RSC compared to mice that had only undergone the simple task, implying that the prior experience of the complex task altered the cortical requirements for performing the simple task.

The next question was whether the neural activity was different in these cortical regions between tasks and between mice with and without previous experience. Neural activity in both the PPC and RSC was greater in mice that had solely undergone the switching task compared to mice that had only undergone the simple task. The trial-type selectivity of neurons was also higher in the switching task in both the PPC and RSC, and it took fewer cells to decode a trial accurately in mice from the switching task. Interestingly, compared to mice who had only experienced the simple task, animals that had prior experience with the switching task showed greater neuronal activity and neuronal selectivity in the PPC and RSC while performing the simple task.

i feel like this is kind of a hurtful term to use because the chimera in mythology is a creature with a variation of lion's head, goats body (and sometimes head), and serpent's tail because it's basically calling one a freak of nature

except Mondays when the theatre is dark.

Author Response

Reviewer #1 (Public Review):

In their paper, titled ‘Group II truncated haemoglobin YjbI prevents reactive oxygene species-induced protein aggregation in Bacillus subtilis’, Imai et al., suggest that the protein YjbI acts as a hydroperoxide peroxidase and therefore it may protect cell-surface cells from oxidation. Using AFM and contact angle measurements they show that yjbI mutants lead to changes in cell surface properties as well as to the formation of more hydrophilic biofilms, relative to the wild-type (WT) strain. Since both tasA and yjbI mutants experienced a similar phenotypic behaviour, the authors linked between the two proteins, TasA and YjbI, and in a series of biophysical and biochemical tests they tried to establish this link. This study touches upon an important question, how do biofilms protect themselves from reactive oxygene species (ROIs), that is nicely described in the introduction; The link between the above proteins in very interesting and relevant to the main question proposed in the study. However, the experiments presented does not always directly support the conclusions made.

The points that I find necessary to clarify/extend:

1) A major claim in the paper is that biofilms that do not harbour the tasA gene (tasA-) are flat, and therefore their contact angle is low, indicating that they are less hydrophobic than WT strains. However, the phenotype of biofilms of tasA mutants are normally not that flat (see for example Romero et al., PNAS 2010; Vlamakis et al., Genes and Development, 2008; Erskine et al., Molecular Microbiology 2018). As a matter of fact, even the WT biofilms that are used as a control in this study are much more flat than the biofilms that serve as standards in the papers referenced above.

We appreciate the reviewer’s comment. As we explained above (answer to Essential Revisions, point 4), there are differences in the morphology of colonies between the 168 and NCBI3610 strains of B. subtilis, as previously pointed out in the literature (Arnaouteli et al. Nat. Rev. Microbiol., 19:600-614, 2021; Mielich-Süss and Lopez, Environ. Microbiol., 17:555-565, 2014). We employed B. subtilis strain 168 because this strain is a close representative of B. subtilis, as described by Zeigler et al. (J. Bacteriol., 190:6983-6995, 2008), and serves as a model organism for wider aspects of basic research, including oxidative damage responses.

To clarify this point, we have added the following text to the revised manuscript in lines 269–277: “Most studies on biofilm formation in B. subtilis use the B. subtilis NCBI3610 strain as a model bacterium because of its ability to form well-structured three-dimensional biofilms (Arnaouteli et al., 2021, Mielich-Süss et al., 2014). The biofilms of the wild-type and tasA mutant strains of the B. subtilis 168 strain are known to be morphologically different from those of the B. subtilis NCBI3610 strain (Romero et al., 2010, Vlamakis et al., 2008, Erskine et al., 2018). In this study, the B. subtilis 168 strain was used because it is the most representative of B. subtilis and serves as a model organism for a wider range of research aspects (Zeigler et al., 2008) as we were not only interested in evaluating biofilm formation but also in more general aspects of oxidative damage responses in bacteria.”

2) Figure 1. The authors use AFM phase imaging to probe differences in cellular stiffness. This AFM mode is not quantitative and the differences presented could also result from differences in adhesion between the tip and the sample. A more quantitative means to evaluate stiffness is a direct measurement of moduli in Force mode, a standard AFM module.

Thank you for your comment. As mentioned above (answer to Essential Revisions, point 3), the AFM data have been removed.

3) Line 147. The authors link between the lack of monomeric TasA in YjbI mutants and the formation of covalent cross linking in TasA aggregates. This is a strong statement that unfortunately is not supported by any of the experiments described in the manuscript.

As mentioned above (answer to Essential Revisions, point 5), we have removed the statement regarding the lack of monomeric TasA in the mutant. The following has been included to highlight the potential involvement of covalent bonds in the TasA aggregate formation in lines 126–129 in the revised manuscript: “No monomeric TasA was detected in the insoluble fraction of the yjbI-deficient mutant strain. An aggregate of TasA was observed under strong reducing and heat-denaturing conditions in SDS sample buffer, suggesting that covalent bonds may be involved in aggregate formation.”

4) The authors seek to make a connection between YjbI and TasA. However, this link is either not well established or only hinted indirectly in this manuscript, through precipitation assays, contact angle measurements and growth curves. To establish such a link, a more molecular approach is advised. Experiments that would provide a direct link between the two proteins and mark specific molecular changes of the proteins include for example titration NMR studies of labelled proteins (at least one of the proteins). In cases where the authors need to show protein localization to the cell surface, it would be of help to use TEM or high-end fluorescence microscopy.

We thank the reviewer for this valuable advice. In response to this comment, we carried out additional experiments, as described above (answer to Essential Revisions, point 1). We will consider the suggested studies, mainly with a molecular approach including titration NMR and TEM, for future studies, as facilities for these specific studies are currently not available.

5) This paper suggests that the protein YjbI acts as an electron donor. Given that there are other proteins with a similar role (in other organisms), it would be nice to show whether there is any homology (by sequence and/or structure) to these proteins.

Thank you for your comments. We have added a description of animal peroxiredoxins and selenomethionine (with GSH or a thioredoxin system) that have been shown to scavenge protein hydroperoxides to the revised manuscript. We also added a description of how YjbI differs from peroxiredoxin and selenomethionine.

The corresponding sentences have been added to the revised manuscript in lines 254–268: “Peroxiredoxins have been reported to repair intracellular protein peroxidation in mammals (Peskin et al., 2010). However, YjbI is distinct from peroxiredoxins in that it is a haem protein with no significant sequence homology (<15%). The second-order rate constants (M-1·s -1) for the reactions of mammalian peroxiredoxins 2 and 3 with BSA-OOH are 160 and 360, respectively, and have been shown to reduce protein peroxides more efficiently than GSH under physiological conditions (Peskin et al., 2010). Although direct comparison is difficult due to different experimental conditions, YjbI and peroxiredoxins are likely to have a similar catalytic rate, as both proteins can reduce BSA-OOH in the order of several mM in roughly 5 min at similar protein concentrations (Fig. 3e) (Peskin et al., 2010). Interestingly, selenomethionine can catalyze the removal of hydroperoxides from proteins in the presence of GSH or a thioredoxin system (Rahmanto & Davies, 2011). However, this system, as well as peroxiredoxins, localises in the cytoplasm of cells, which is a significant difference between YjbI and these proteins. Moreover, whether bacteria utilize peroxiredoxins and the selenomethionine system to remove hydroperoxides from proteins remains unclear.”

6) (Minor point). The use of Pymol to demonstrate that the YjbI's pocket could serve as a binding site for haem molecule is nice, but using Molecular Dynamics (or any other calculation) would be more quantitative and convincing of the specificity of the interaction.

We appreciate your comment regarding this point. However, we believe that analysis using molecular dynamics (or other calculations) is largely difficult because the structure of the hydroperoxidised protein substrates is not available. Further, the degree of similarity between the structure of TasA or BSA and the hydroperoxidised form is unclear. A calculation analysis with a small model substrate can be adopted in future work. Therefore, we only showed the surface opening of the YjbI structure, which is potentially relevant for binding to a hydroperoxidised protein substrate.

Reviewer #2 (Public Review):

In this study, Imai et al. uncover a role for the truncated haemoglobin protein YjbI in biofilm formation by the model bacterium B. subtilis. They show that yjbI gene disruption results in altered biofilms, with increased wettability and different matrix stiffness relative to cells. The absence of YjbI activity results in aggregation of the amyloid-like TasA matrix protein, and the biofilm wettability defect of the yjbI mutant can be recapitulated by anti-YjbI immune serum, suggesting that YjbI is located on the cell surface. Absence of YjbI also modestly increases the sensitivity of cells growing on agar plates to the oxidant AAPH. Using the model protein substrate BSA, purified YjbI can at least partially reverse oxidant-induced BSA aggregation in vitro, convincingly showing the YjbI has protein hydroperoxide peroxidase activity, which is evidently an unusual enzymatic activity. Finally, the authors examine lipid peroxidation and conclude that YjbI is not involved. The results are interesting in that they connect YjbI to a biofilm phenotype and convincingly show protein hydroperoxide peroxidase activity by a truncated haemoglobin protein, an activity not previously attributed to this class of proteins.

The experiments are largely well done, but some of the corresponding conclusions are overinterpreted, connecting ideas without experimental support. Moreover, the yjbI mutant has a narrow and relatively mild phenotype.

1) The paper identifies two separate properties of YjbI: its mutant phenotype with respect to biofilm formation, and its peroxidase activity against oxidant-induced aggregation of TasA and BSA. The authors conclude that these properties are connected, but this is not formally tested. While purified YjbI can reverse hydrogen peroxide-induced aggregation of purified TasA in vitro, and the yjbI mutant shows more TasA in the insoluble fraction of B. subtilis pellicle lysates, these experiments do not show that the TasA aggregates in pellicle lysates are caused by peroxidation, nor do they show that TasA aggregation is normally kept at bay by YjbI peroxidase activity (it is possible that YjbI has a separate role in biofilm integrity). Some experiments that might lend support to this connection include examining the biofilm phenotype of a catalytically dead point mutant of YjbI (perhaps Y25 or Y63, l. 298, or other residues informed by the crystal structure of Giangiacomo et al.) to establish whether peroxidase activity is important for biofilm formation. Such a mutant would be particularly valuable, as it could also be used to test whether inactivation of enzyme activity affects other phenotypes (cell stiffness, for example). Another approach would be to use a soluble antioxidant molecule, purified YjbI, or another peroxidase to see if the yjbI biofilm can be rescued.

We greatly appreciate this comment, which is critical for improving our manuscript. To address this issue, we performed additional experiments using the Y25F, Y63F, and W69F variants of YjbI. The introduction of the Y63F variant gene into the yjbI-deficient strain failed to complement the defective phenotype of the yjbI-deficient strain in biofilm repellency (revised Fig. 1b). We found that the purified Y63F lost its hydroperoxide peroxidase activity (revised Fig. 3g). These results show a connection between the protein hydroperoxide peroxidase activity of YjbI and the abnormal biofilm phenotype of the yjbI-deficient strain. Accordingly, Figs. 1b and 3g have been added to the revised manuscript and figure descriptions have been included in lines 220–226, 322–324, and 327–328 (as explained above in the answer to Essential Revision, Point 2).

2) The authors conclude on the basis of the AFM data in Fig. 1 that yjbI mutant cells are less stiff than WT cells, but the data only show relative stiffness. It is also unclear why a change in cell envelope stiffness would relate to biofilm wettability (ll. 130-131). If there truly is a change in cell envelope stiffness, a high-resolution, head-to-head AFM comparison of planktonically grown cells would be informative.

We appreciate the reviewer’s comment on this point. As mentioned above (answer to Essential Revision, points 3 and 6), we realized that our interpretations of the AFM data were not appropriate and not relevant to biofilm repellency. Accordingly, the AFM data were removed.

3) The data in Fig. 2F showing hypersensitivity of yjbI mutant cells to AAPH were generated in an unusual way: stationary-phase liquid culture was spotted on an LB plate, and the colonies were "fractionated" at the noted intervals and resuspended in saline for OD measurement. Measuring sensitivity to AAPH just in shaking liquid planktonic culture would make this phenotype more convincing. Under non-biofilm forming conditions, is a surface-associated peroxidase important for cell growth or survival under oxidant challenge?

We appreciate your comment regarding this point and apologize for the error in the description “'Planktonically grown B. subtilis strains under AAPH-induced oxidative stress'” in the Methods section. No solid medium was used in the experiments. The description in the Figure legend of Fig. 2f is correct. The sentence in the text has been rewritten in the revised manuscript in lines 813.

I agree with Minha's assessment of the project. Her research question is phrased perfectly for the overall topic of these combined videos. I can't stop, and I think I won't stop thinking about what it truly means for me to age. Each voice represents a background that provides a resource for both the voice owner and the audience to answer this question. Aging for me means being more cautious with words and actions. I consciously do this because I see everyone around me go through this process and talk about it. Aging for me means looking at my grandparents and and thinking what I will do and what I will look like when I reach their age. I thought about this question a few times when I was much younger, then there was a long period of me not worrying about it at all, and in college, the question came back to me at higher rate of frequencies. I often ask myself if my future kids/grandkids (if I ever have them) would care about me and life after death was something that seems to be in my head for the longest time. Aging for me means carrying new responsibilities. I know that there are things that was acceptable when I was one year younger and became inapplicable for me the year after, and vice versa. "What it means to age?" is repeatedly asked throughout the video, motivating us to give it a try and craft our own response. This research question has well summarized for the bigger and better understanding of the purpose that these 'storytellers' and collaborators embed in this project. Same with taylortots, I may revisit this project from time to time with newer perspectives about the definition of growing old. Thank you for the insightful post!

Letting your past control your future - because people form their reference frames based on their experiences (anecdotal evidence) and this means that people have very different perceptions of the same event. Your past shapes the future you see

live out the consequences of your taste

self-condemned to an appalling nightmare

you won't be getting it all the subtle stuff

can't nonsense foolishness