Reviewer #2 (Public review):

Summary:

The authors compiled 29 patients with various neurodevelopmental symptoms due to the ARID5B mutations. Although not directly, the mouse model demonstrated that the heterozygous mutant mouse showed mild behavioral problems. It would be interesting to see if the mice carry craniofacial features.

Strengths:

The HEK293T model showed that the mutant protein mis-localized, but did not show whether the mutation caused any changes in epigenetic status. Nevertheless, this paper delivers clear support for genotype-phenotype correlation.

Weaknesses:

(1) The paper would be improved by providing pedigrees of some of the patients with inherited variants.

(2) Figure 3d could provide more species for an accurate conservation assessment.

Reviewer #3 (Public review):

Summary:

In the present study, through international gene-matching efforts, the authors present 29 individuals with rare, heterozygous ARID5B variants and find that these individuals have a newly recognizable neurodevelopmental syndrome. A recurring clinical syndrome of developmental delay/intellectual disability, behavioral difficulties, renal malformation, and recurrent infections is described. 19 of these variants were confirmed to be de novo, and only one was inherited from an unaffected parent. 24/29 of these variants introduce premature termination codons in the final exon and are predicted to escape nonsense-mediated decay. The ARID5B p.Q522Ter variant was studied in a mouse heterozygous knock-in model, found to be associated with behavioral abnormalities. The well-described genetic and phenotypic data for this cohort provide convincing clinical evidence for a novel neurodevelopmental syndrome. The functional evidence provided is preliminary, and further studies are needed to understand disease mechanisms.

Strengths:

(1) The authors give a good description of a novel clinical syndrome manifesting as developmental delay/intellectual disability, facial dysmorphism, and behavioral challenges.

(2) The authors create a mouse model harboring an Arid5b(Q522*/+) variant and identify subtle behavioral changes.

(3) Attempts are made to functionally characterize a subset of ARID5B variants in human cell lines.

Weaknesses:

(1) The title - "ARID5B mutations cause a neurodevelopmental syndrome with neuroinflammation episodes" - should be revised. 2/29 individuals (7%) had CNS inflammation; this does not appear to be a core feature of the disease and should not be highlighted as such. If this is going to be a feature that is highlighted, then more details are needed. MRI images of cerebellitis and/or ADEM would be helpful, as well as lumbar puncture results and supplemental information detailing the treatment course.

(2) The abstract states that "Remarkably, 19 of 29 variants (66%) cluster within the first quarter of exon 10, are de novo, and escape nonsense-mediated mRNA decay (NMD), which we confirmed for two variants affecting seven individuals." The authors state in the Results that they "indeed found no signs of NMD". In Figure 3f, when assessing for transcript amount, there appears to be a great deal of variability. Three ARID5B variant lines are tested. Transcript amounts in two lines appear to be near control levels, but one LCL ARID5B Ile497AsnfsTer31 line appears to demonstrate significantly lower levels of transcript. The control lines also show a great deal of variability. No explanation is given for this large difference between LCL ARID5B Ile497AsnfsTer31 lines and for the variability in control lines, making these data uninterpretable. A major theme of the paper is that early truncating variants in exon 10 escape NMD and lead to the described phenotypes, so this is an important point that needs to be resolved, either by testing more patient-derived lines or knocking in these variants into cell lines.

(3) The Arid5b(Q522*/+) mice are not sufficiently molecularly characterized. Does the variant transcript escape NMD? What happens at the protein level? Is there mislocalization of the protein?

(4) For the HEK293T cell experiments, variants are overexpressed and compared to a control. These experiments appear to leave endogenous ARID5B intact. What might the authors expect to see if these variants were knocked in?

(5) The functional consequences of the missense variants are not tested. The authors suggest that missense variants may be more associated with macrocephaly and possibly ASD. Are these missense variants causing loss-of-function or gain-of-function? Is there preserved protein function?

(6) There are a number of functional assays performed, but it remains unclear if the tested variants are operating through a loss- or gain-of-function. Are truncating variants early in exon 10 leading to a partial loss-of-function? Or do they prevent the functioning of the other allele through a dominant negative mechanism? These possibilities are not directly tested.

Reviewer #1 (Public review):

Summary:

This is an excellent and strong paper. The authors not only show the mechanisms of action of destabilizing mutations in VHL, but notably, they also go on to computationally design and experimentally test an inhibitor that restores wild-type pVHL function, offering starting points for a new class of kidney cancer drugs. The approach that the authors take here can be used to target destabilizing mutations in repressor proteins, common in diseases, including cancer.

Strengths:

This paper is the culmination of an extraordinary amount of work, over years, including method development and testing by a broad range of tools and experiments. It is thorough and comprehensive. It is also well-written and easy to follow.

Reviewer #2 (Public review):

Summary:

Inactivating VHL mutations are common in clear cell renal cell carcinoma, and about half of those mutations unfold/destabilize the protein rather than directly interfering with critical protein-protein interactions. The authors identify a compound that can stabilize/refold mutant VHL and seemingly restore its ability to downregulate its major downstream targets.

Strengths:

The authors use a clever combination of virtual and cell-based screens, followed by suitable biophysical and cell-based validation assays, to arrive at a VHL refolder. This compound is suboptimal from an ADME point of view, but could be a starting point for further medicinal chemistry optimization. Success would have implications for other diseases linked to similar loss-of-function mutations.

Weaknesses:

The authors need to tighten up the evidence that the VHL refolder is downregulating HIF2 in a strictly "on-target" manner.

(1) In Figure 3C, the increase in VHL stability looks very modest. Taking into account the increased abundance of the VHL protein at time 0 in the presence of CP4 compared to control, it is not so clear that VHL is decaying much more slowly in the presence of CP4. I understand that the fact that the signal is low in the absence of CP4 at time 1 hour makes it hard to quantify the half-life of p30 in the absence of the drug (is a longer exposure needed?). However, perhaps the authors could try to quantify the p19 half-life.

(2) In going from CP4 to CP4.29 the authors screened based on downregulation of HIF. This is logical but also introduces the danger of identifying chemicals that can downregulate HIF in an "off-target" manner, i.e. non-specifically. It is therefore essential to clearly show that CP4.29 downregulates steady-state levels of HIF and HIF target genes in cells with suitable (hydrophobic core) VHL mutants but not in isogenic cells lacking VHL. Another prediction is that these chemicals should be inert in cells with VHL mutations that directly abrogate HIF binding. So Figure 4E (HIF2 target genes) needs the use of the isogenic VHL-/- cells described later in the paper. And the steady-state levels of HIF2 should be measured in the isogenic cells (mutant VHL vs -/-) with or without CP4.29. Figure 4G, as it is done now, is too indirect and not very compelling. I don't understand why the "time 0" HIF2 levels aren't lower in the presence of CP4.29, and I think the half-life differences with treatment are very subtle to my eyeball densitometer (although I applaud the authors' attempt to quantify), with the exception of I180N. I agree that Figure 4F is encouraging, but hypoxia has many effects, and this experiment is not as "clean" as the VHL-/- experiments. The same applies to some of the pharmacologic agents in Figure 5.

Reviewer #1 (Public review):

The current study is a follow-up to a previously published article in eLife in 2021, demonstrating that the transcription factor MYRF-1 interacts with the transmembrane protein PAN-1, which is required for the stability and targeting of MYRF-1 to the plasma membrane. There, MYRF-1 undergoes self-catalytic cleavage of its intracellular domain and translocates to the nucleus. Here, the authors analyze the activation of MYRF-1 during the larval development of C. elegans. They nicely show that MYRF-1 cleavage and nuclear translocation oscillate with larval stage transitions. They further identify two regions in MYRF-1 and PAN-1 that negatively regulate MYRF-1 cleavage and activation, and show that relief of this negative regulation causes premature lin-4 activation and overrides nutrient-responsive developmental checkpoints. The experiments are elegant and accurately support the conclusions raised. There are only minor comments and suggestions to improve the manuscript.

Reviewer #2 (Public review):

In this study, Xu et al. investigated the regulatory mechanisms controlling intramolecular cleavage of the transmembrane transcription factor MYRF-1, an important event that controls developmental progression in C. elegans.

The authors made important advances in several aspects:

(1) Through endogenous gene editing/tagging, further supported by western blots, the authors convincingly demonstrate the novel finding that the intramolecular cleavage and nuclear translocation of MYRF-1 is not static, but temporally controlled within each developmental stage: with nuclear translocation peaking at the late stage and then declining into lethargus/molts between developmental stages (Figure 1).

(2) They demonstrate that this cleavage and nuclear translocation is controlled by external stimuli, namely starvation.

(3) They reveal modes of regulation of the intramolecular cleavage that is mildly regulated by MYRF-1's own JM domain as well as the CCT tail of interacting partner PAN-1.

The conclusions of this paper are mostly well supported by data, but some aspects of the manuscript and conclusions should be clarified and extended to strengthen its findings.

(1) The authors concluded that the intramolecular cleavage and nuclear localization of MYRF-1 were similarly temporally-regulated in all tissue types. However, the data/image presented was limited to specific regions/cell types that were inconsistently chosen across developmental windows. For example, for the cleavage/nuclear translocation across L1 into lethargus (Figures 1B, E, F, G), the heads of the worm were shown to comprise mostly neurons and muscles. While across the rest of the larval stages, only mid-body pictures were shown, comprising mostly hypodermal and some intestinal cells. A complete coverage of all tissues across all time points would better support the author's conclusion that this temporal regulation occurs similarly in all tissue types. Additionally, the authors should clearly indicate which tissue/cell-types were used in the quantifications, as these were not done for several figure panels (including but not limited to Figure 1I and J).

(2) Related to point 1 above, this inconsistency in tissue assessment was also true for downstream experiments (Figures 2-6; e.g., starvation, JM, and CCT regulation, etc.). Broad tissue specific assessment for all downstream experiments would greatly enhance the strength and relevance of the findings. Judging by the current data presented (Figures 3, 5, 6), it seems to suggest that there are tissue/cell-type differences in the regulation of MYRF-1 nuclear translocation.

(3) Developmental progression was superficially and inconsistently assessed across the study. Developmental progression was mainly assessed by hypodermal (V-lineage) division patterns and worm length in this study. Several glaring omissions that should have been examined were the lengths of larval stages/lethargus and molting defects, as well as gonad development, to help identify which developmental landmarks were affected vs. not.

(4) The phosphorylation within MYRF-1's JM domain was insufficiently investigated. There were two serine phosphorylation sites that were discovered through mass spectrometry experiments, however the authors only investigated one of the serine (S623) residues without any justifications for the choice. Additional investigation of the other residues, as well as both together, would strengthen the relevance of these phosphorylation events to cleavage and nuclear translocation, especially considering the minimal effect observed with only mutating the one residue.

Reviewer #3 (Public review):

Summary:

In this paper, the authors identified dual inhibitory mechanisms, an intrinsic juxtamembrane (JM) region and an extrinsic cytoplasmic tail (CCT) domain in the binding protein PAN-1, that suppress MYRF-1 cleavage in C. elegans. The authors showed that MYRF-1 cleavage oscillates across larval stages, peaking in mid-to-late phases and being suppressed during molts. This oscillatory pattern is consistent with MYRF-1's role in promoting transitions of larval stages, particularly in late-L1 involving lin-4 activation and DD neuron remodeling.

Strengths:

This work generated several knock-in strains of fluorescent tags and mutations in the endogenous myrf-1 and pan-1gene loci, which will provide resources for future identification and characterization of the underlying molecular mechanisms regulating MYRF-1 cleavage inhibition.

The results presented in the paper are solid enough to support the paper's main conclusions.

This study is valuable for establishing MYRF-1 cleavage as a key gatekeeper of the C. elegans developmental timing. Findings from C. elegans MYRF-1 may provide insight into the regulation and function of mammalian MYRF.

Weaknesses:

The following points should be discussed to further support the authors' model that MYRF-1 cleavage is a key gatekeeper of developmental timing.

(1) Recent findings by Helge Großhans and Jordan Ward groups showed that KIN-20 (CK1δ) and LIN-42 (PERIOD) are required for proper molt timing in C. elegans, and that loss of LIN-42 binding or of the phosphorylated LIN-42 tail impairs nuclear accumulation of KIN-20, resulting in arrhythmic molts (EMBO J. 44, 6368-6396, 2025). In this paper, the authors concluded that PAN-1 promotes MYRF trafficking to the cell membrane, where MYRF-1 cleavage and nuclear translocation occur, and that oscillates with developmental molting cycles in C. elegans. It is unclear whether MYRF-1 and KIN-20 interact in the nucleus and, if so, how this interaction controls developmental timing.

(2) Separately, it was previously shown that the let-7 primary transcript (pri-let-7) exhibits oscillating, pulse-like expression that peaks during each larval stage, rather than a steady increase, and directly correlates with developmental molting cycles. It is unclear whether the nuclear-localized MYRF-1 fragment regulates the oscillatory primary let-7 expression during larval transition (McCulloch and Rougvie, 2014; Van Wynsberghe et al., 2011).

Reviewer #1 (Public review):

Summary:

This manuscript by Feng et al. uses mouse models to study the embryonic origins of HSPCs. Using multiple types of genetic lineage tracing, the authors aimed to identify whether BM-resident endothelial cells retain hematopoietic capacity in adult organisms. Through an important mix of various labeling methodologies (and various controls), they reach the conclusion that BM endothelial cells contribute up to 3-4% of hematopoietic cells in young mice.

Strengths:

The major strength of the paper lies in the combination of various labeling strategies, including multiple Cdh5-CreER transgenic lines, different CreER lines (col1a2), and different reporters (ZsGreen, mTmG), including a barcoding-type reporter (PolyLox). This makes it highly unlikely that the results are driven by a rare artifact due to one random Cre line, or one leaky reporter. The transplantation control (where the authors show no labeling of transplanted LSKs from the Cdh5 model) is also very supportive of their conclusions.

Weaknesses:

While the updated manuscript now provides strong evidence for Cdh5-CreER+ cells as a source of myeloid-biased hematopoiesis, the true identity of these "adult EHT stem cells", their differentiation hierarchy, the kinetics, the EHT mechanism, and the physiological relevance of this process remain unaddressed.

Reviewer #2 (Public review):

Summary:

Feng, Jing-Xin et al. studied the hemogenic capacity of the endothelial cells in the adult mouse bone marrow. Using Cdh5-CreERT2 in vivo inducible system, though rare, they characterized a subset of endothelial cells expressing hematopoietic markers which was transplantable. They suggested that the endothelial cells need the support of stromal cells to acquire blood forming capacity ex vivo. This endothelial cells were transplantable and contribute to hematopoiesis with ca. 1% chimerism in a stress hematopoiesis condition (5-FU) and recruited to peritoneal cavity upon Thioglycolate treatment. Ultimately, the authors detailed the blood lineage generation of the adult endothelial cells in a single cell fashion suggesting a predominant HSPCs-independent blood formation by adult bone marrow endothelial cells, in addition to the discovery of Col1a2+ endothelial cells with blood forming potential corresponds to its high Runx1 expressing property.

Comments on revised version:

Overall, the authors have addressed our main concerns, and the revised manuscript is improved. The new data now more strongly supports long-term multilineage reconstitution of the adult hemogenic ECs. However, critical data, especially regarding the ECs' hematopoietic identity and functional capacity remains insufficient, which limits the strength of hemogenic claim, especially to assert that these adult hemogenic ECs generate bona fide HSCs.

Points that are sufficiently addressed:

(1) Exclusion of the potential contamination during cell sorting for the ex vivo CD45- ZsGreen+ fraction culture has been explicitly shown to be of a high purity in Fig. 2B.

(2). The pre-cultured ZsG+ fraction is shown to having a long-term multilineage reconstituting capacity (10 months chimerism, Fig. 2J), which increases confidence that the fraction is not limited to short-lived progenitors.

Points that are insufficiently addressed:

(1) As noted in the "Limitation of Study", the absence of LT-HSC phenotyping and/or secondary transplantation data of ZsG+ donor limits confidence in concluding that the adult hemogenic BM-ECs generate HSPCs.

(2) The lack of early to the end of reconstitution kinetics in Fig 2E-2J restricts interpretation of whether the donor fraction contains rapid reconstituting transient progenitor versus sustained repopulating HSCs.

Reviewer #1 (Public review):

Summary:

The goal of this paper was to determine whether the T cell receptor (TCR) repertoire differs between a male or female human. To address this, this group sequenced TCRs from double-positive and single-positive thymocytes in male and female humans of various ages. Such an analysis on sorted thymocyte subsets has not been performed in the past. The only comparable dataset is a pediatric thymocyte dataset where total thymocytes were sorted.

They report on participant ages and sexes, but not on ethnicity, race, nor provide information about HLA typing of individuals. The experiments are heroic, yet do represent a relatively small sampling of diverse humans. They observed no differences in TCRbeta or TCRalpha usage, combinational diversity, or differences in the length of the CDR3 region, or amino acid usage in the CD3aa region between males or females. Though they observed some TCRbeta CD3aa sequence motifs that differed between males and females, these findings could not be replicated using an external dataset and therefore were not generalizable to the human population.

They also compared TCRbeta sequences against those identified in the past databases using computational approaches to recognize cancer-, bacterial-, viral-, or autoimmune-antigens. They found little overlap of their sequences with these annotated sequences (depending on the individual, ranged from 0.82-3.58% of sequences). Within the sequences that were in overlap, they found that certain sequences against autoimmune or bacterial antigens were significantly over-represented in female versus male CD8 SP cells. Since no other comparable dataset is available, they could not conclude whether this is a generalizable finding in the human population.

Strengths:

It is a novel dataset that attempts to understand sex differences in the T cell repertoire in humans. Overall, the methodologies are sound and are the current state-of-the-art. There was an attempt to replicate their findings in cases where an appropriate dataset was available. I agree that there are no gross differences in TCR diversity between males and females. This is an important negative result.

Weaknesses:

Overall, the sample size is small given that it is an outbred population. This reviewer recognizes the difficulty in obtaining samples for this experiment (which were from deceased donors), and this limitation was appropriately discussed. Their analysis was limited by the current availability of other TCR sequences. These weaknesses were appropriately discussed and considered.

Reviewer #2 (Public review):

Summary:

This study addresses the hypothesis that the strikingly higher prevalence of autoimmune diseases in women could be the result of biased thymic generation or selection of TCR repertoires. The biological question is important and the hypothesis is valuable. Although the topic is conceptually interesting and the dataset is rich, the study has a number of major issues. In particular, the majority of "autoimmunity-related TCRs" considered in this study are in fact specific to type 1 diabetes (T1D). Notably, T1D incidence is higher in males, which directly contradicts the stated objective of the study - to explain the higher prevalence of autoimmune diseases in women. Given this conceptual inconsistency, the evidence presented does not support the authors' conclusions.

Strengths:

The key strength of this work is the newly generated dataset of TCR repertoires from sorted thymocyte subsets (DP and SP populations). This approach enables the authors to distinguish between biases in TCR generation (DP) and thymic selection (SP). Bulk TCR sequencing allows deeper repertoire coverage than single-cell approaches, which is valuable here, although the absence of TRA-TRB pairing and HLA context limits the interpretability of antigen specificity analyses. Importantly, this dataset represents a valuable community resource and should be openly deposited rather than being "available upon request."

Weaknesses:

I thank the authors for their detailed responses to my previous comments. Several concerns were addressed satisfactorily; however, important issues remain unresolved, and a new major concern has emerged from the revised manuscript.

Major concerns:

(1) Autoimmune specificity is dominated by T1D, contradicting the study's premise. Newly added supplementary Table 3 shows that the authors considered only 14 autoimmune-related epitopes, of which 12 are associated with type 1 diabetes (T1D) and 2 with celiac disease (CeD). (I guess this is because identification of particular peptide autoantigens is an extremely difficult task and was only successful in T1D and CeD.) Thus conclusions of this work mostly relate to T1D. However, the incidence of T1D is higher in males than in females (e.g. doi:10.1111/j.1365-2796.2007.01896.x; doi:10.25646/11439.2). This directly contradicts the stated objective of the study - to explain the higher prevalence of autoimmune diseases in women. As a result, the authors' conclusions (a) cannot be generalized to autoimmune disease as a whole as the authors only considered T1D and CeD antigens and (b) are internally inconsistent with the stated objective of the study.

(2) By contrast, CeD does show a female bias (~60/40 female/male; doi:10.1016/j.cgh.2018.11.013). However, the manuscript does not allow evaluation of how much the reported "autoimmune TCR enrichment" derives from T1D versus CeD. Despite my previous request, the authors did not provide per-donor and per-epitope distributions of autoimmune-specific TCR matches. I therefore explicitly request a table in which: each row corresponds to a specific autoimmune antigen; each column corresponds to a donor (with metadata available including sex); each cell reports the number of unique TCRs specific to that antigen in that donor. Without such data, the conclusions cannot be evaluated.

(3) It is scientifically inappropriate to generalize findings to "autoimmune diseases" when only T1D and CeD were analyzed. Moreover, given that T1D and CeD show opposite directions of sex bias, combining them into a single "AID" category is misleading. All analyses presented in Figure 8 and Supplementary Figure 16 should be repeated and shown separately for T1D and CeD, rather than combined.

(4) The McPAS database contains TCRs associated with other autoimmune diseases (e.g., multiple sclerosis, rheumatoid arthritis), although the exact autoantigens in these contexts are unknown. Why didn't the authors perform the search for such TCRs? I believe disease association even without particular known antigen could still be insightful.

(5) Misuse of the concept of polyspecificity. I appreciate the authors' reference to Don Mason's work; however, the concept of polyspecificity discussed there is fundamentally different from the authors' usage. Mason, Sewell (doi:10.1074/jbc.M111.289488), Garcia (doi:10.1016/j.cell.2014.03.047), and others demonstrated that individual TCRs can recognize multiple peptides, possibly around 1 million. But importantly these peptides are not random but share some sequence motif. This is a general feature of TCRs, i.e. 100% of TCRs are polyspecific in this sense.<br /> In contrast, the authors define polyspecificity as TRB sequences annotated as specific to unrelated epitopes in TCR databases such as VDJdb. These databases are well known to contain substantial numbers of false-positive annotations (see, e.g., Ton Schumacher's preprint https://www.biorxiv.org/content/10.1101/2025.04.28.651095.abstract). The authors acknowledge that, under their definition, polyspecificity has been experimentally validated for only one (!) TCR (Quiniou et al.). In the absence of robust experimental validation, use of the term "polyspecificity" in this context is misleading. I strongly recommend removing all analyses and conclusions related to polyspecificity from the manuscript unless supported by independent functional validation.

(6) I agree that comparing specificity enrichment between sexes is meaningful. However, enrichment relative to the database composition itself is not biologically interpretable, as acknowledged by the authors in their response. I therefore recommend removing Supplementary Figure 15, which is potentially misleading.

(7) In contrast, Supplementary Figure 16 represents the most convincing result of the study (keeping in mind that the AID group should be splitted to T1D and CeD with T1D and that T1D and CeD have opposing directions of sex biases) and should be shown as a main figure, replacing Figure 8A-B which is less convincing as it doesn't show per-donor distribution.

(8) The authors argue that applying mixed-effects modeling to Rényi entropy would require assuming a common sex effect across subsets. I do not find this assumption unreasonable. For example, if sex effects are mediated through AIRE-dependent negative selection, one would indeed expect a consistent direction of effect across subsets. The lack of statistical significance in Figure 3 may reflect limited sample size rather than true absence of the difference. Moreover, the title's phrasing "comparable TCR repertoire diversity" is vague: what is the statistical definition of "comparable"?

Reviewer #1 (Public review):

Summary:

This paper tackles an important question: What drives the predictability of pre-stimulus brain activity? The authors challenge the claim that "pre-onset" encoding effects in naturalistic language data have to reflect the brain predicting the upcoming word. They lay out an alternative explanation: because language has statistical structure and dependencies, the "pre-onset" effect might arise from these dependencies, instead of active prediction. The authors analyze two MEG datasets with naturalistic data.

Strengths:

The paper proposes a very interesting alternative hypothesis for claims in prior work (e.g., Goldstein et al., 2022). In contrast to claims in prior work, the current paper convincingly demonstrates that prior results can be explained by inherent stimulus dependencies in natural language, as opposed to the brain actively predicting future linguistic content.

Two independent datasets are analyzed. The analyses with the most and least predictive words is clever, and is nicely complementing the more naturalistic analyses. The work emphasizes how claims about linguistic prediction cannot be trivially drawn using encoding models in naturalistic designs.

Reviewer #2 (Public review):

Summary:

At a high-level, the reviewers demonstrate that there is a explanation for pre-word-onset predictivity in neural responses that does not invoke a theory of predictive coding or processing. The paper does this by demonstrating that this predictivity can be explained solely as a property of the local mutual information statistics of natural language. That is, the reason that pre-word onset predictivity exist could simply boil down to the common prevalence of redundant bigram or skip-gram information in natural language.

Strengths:

The paper addresses a problem of significance and uses methods from modern NeuroAI encoding model literature to do so. The arguments, both around stimulus dependencies and the problems of residualization, are compellingly motivated and point out major holes in the reasoning behind several influential papers in the field, most notably Goldstein et al. This result, together with other papers that have pointed out other serious problems in this body of work, should provoke a reconsideration of papers from encoding model literature that have promoted predictive coding. The paper also brings to the forefront issues in extremely common methods like residualization that are good to raise for those who might be tempted to use or interpret these methods incorrectly.

Weaknesses:

After author revision, I see no major weaknesses in the underlying arguments or data processing steps.

Reviewer #3 (Public review):

Summary:

The study by Schönmann et al. presents compelling analyses based on two MEG datasets, offering strong evidence that the pre-onset response observed in a highly influential study (Goldstein et al., 2022) can be attributed to stimulus dependencies-specifically, the auto-correlation in the stimuli-rather than to predictive processing in the brain. Given that both the pre-onset response and the encoding model are central to the landmark study, and that similar approaches have been adopted in several influential works, this manuscript is likely to be of high interest to the field. Overall, this study encourages more cautious interpretation of pre-onset responses in neural data, and the paper is well written and clearly structured.

Strengths:

• The authors provide clear and convincing evidence that inherent dependencies in word embeddings can lead to pre-activation of upcoming words, previously interpreted as neural predictive processing in many influential studies.

• They demonstrate that dependencies across representational domains (word embeddings and acoustic features) can explain the pre-onset response, and that these effects are not eliminated by regressing out neighboring word embeddings-an approach used in prior work.

• The study is based on two large MEG datasets and one ECoG dataset, showing that results previously observed in ECoG data can be replicated in MEG. Moreover, the stimulus dependencies appear to be consistent across the three datasets.

Weaknesses:

• While this study shows that stimulus dependency can account for pre-onset responses, it remains unclear whether this fully explains them, or whether predictive processing still plays a role. The more important question is whether pre-activation remains after accounting for these confounds.

Comments on revisions:

I appreciate the added analyses. This study raises an important methodological concern regarding an influential paper and will certainly have a high impact on our field.

Reviewer #1 (Public review):

Summary:

The authors investigated the population structure of the invasive weed Lantana camara from 36 localities in India using 19,008 genome-wide SNPs obtained through ddRAD sequencing.

Strengths:

The manuscript is well-written, the analyses are sound, and the figures are of great quality.

Weaknesses:

The narrative almost completely ignores the fact that this plant is popular in horticultural trade and the different color morphs that form genetic populations are most likely the result of artificial selection by humans for certain colors for trade, and not the result of natural selfing. Although it may be possible that the genetic clustering of color morphs is maintained in the wild through selfing, there is no evidence in this study to support that. The high levels of homozygosity are more likely explained as a result of artificial selection in horticulture and relatively recent introductions in India. Therefore, the claim of the title that "the population structure.. is shaped by its mating system" is in part moot, because any population structure is in large part shaped by the mating system of the organism, but further misleading because it is much more likely artificial selection that caused the patterns observed.

Update after manuscript was revised by authors:

The authors added a selfing experiment, showing that the wild plants are selfing and not outcrossing, which limits the genetic exchange. This supports their claims, but a link with the horticultural industry is still lacking in the study, and conclusions should still be viewed in the regional context of India rather than globally.

Reviewer #2 (Public review):

Summary:

The authors performed a series of population genetic analyses in Lantana camara using 19,008 genome-wide SNPs data from 359 individuals in India. They found clear population structure that did not show a geographical pattern, and flower color was rather associated with population structure. Excess of homozygosity indicate high selfing rate, which may lead to fixation of alleles in local populations and explain the presence of population structure without a clear geographic pattern. Authors also performed a forward simulation analysis, theoretically confirming that selfing promotes fixation of alleles (higher Fst) and reduction in genetic diversity (lower heterozygosity).

Strengths:

Biological invasion is a critical driver of biodiversity loss, and it is important to understand how invasive species adapt to novel environments despite limited genetic diversity (genetic paradox of biological invasion). Lantana camara is one of the hundred most invasive species in the world (IUCN 2000), and the authors collected 359 plants from a wide geographical range in India, where L. camara has invaded. The scale of the dataset and the importance of the target species are the strength of the present study. Coalescent-based analysis nicely supports the authors' claim that multiple introductions may have contributed the population structure of this species.

Weaknesses:

The main findings of the SLiM-based simulation were that inbreeding promotes fixation of alleles and reduction in genetic diversity. These are theoretically well known, and such findings themselves are not novel, although it may have become interesting if these findings are quantitatively integrated with their empirical findings in the studied species.

Reviewer #1 (Public review):

Summary:

In this paper, Chen et al. identified a role for the circadian photoreceptor CRYPTOCHROME (CRY) in promoting wakefulness under short photoperiods. This research is potentially important as hypersomnolence is often seen in patients suffering from SAD during winter times. The mechanisms underlying these sleep effects are poorly known.

Strengths:

The authors clearly demonstrated that mutations in cry lead to elevated sleep under 4:20 Light-Dark (LD) cycles. Furthermore, using RNAi, they identified GABAergic neurons as a primary site of CRY action to promote wakefulness under short photoperiods. They then provide genetic and pharmacological evidence demonstrating that CRY acts on GABAergic transmission to modulate sleep under such conditions.

Weaknesses:

The authors then went on to identify the neuronal location of this CRY action on sleep. This is where this reviewer is much more circumspect about the data provided. The authors hypothesize that the l-LNvs which are known to be arousal promoting may be involved in the phenotypes they are observing. To investigate this, they undertook several imaging and genetic experiments.

While the authors have made improvements in this resubmitted manuscript, there are still multiple concerns about the paper. I think the authors provide enough evidence suggesting that CRY plays a role in sleep under short photoperiod. The data also supports that CRY acts in GABAergic neurons. However, there are still major issues with the quality of the confocal images presented throughout the paper. In many cases it appears that the images are oversaturated with poor resolution, making it hard to understand what is going on. In addition, none of the drivers used in this study are specific to the neurons the authors aim to manipulate. Therefore, the identity of the GABAergic neurons involved in this CRY dependent sleep mechanism remains unclear. Similarly, whether l-LNvs are the target of this GABA mediated sleep regulation under short photoperiod is not fully demonstrated. The data presented suggests that but does not prove it.

Major concerns:

(1) While the authors provided sleep parameters like consolidation or waking activity for some experiments. These measurements are still not shown for several experiments (for example Figures 2E, 3, 4, 5, and 6). These data are essential, these metrics must be reported for all sleep experiments.

(2) Line 144 "We fed flies with agonists of GABA-A (THIP) and GABA-B receptor (SKF-97541) (Ki and Lim, 2019; Matsuda et al., 1996; Mezler et al., 2001). Both drugs enhance sleep in WT," The proper citation is needed here, Dissel et al., 2015 PMID:25913403. Both THIP and SKF-97541 were used in that paper.

(3) Figure 2C and 2F: it appears that the control data is the same in both panels. That is not acceptable.

(4) Figure 4A: With the quality of the images, it is impossible to assess whether GABA levels are increased at the l-LNvs soma.

(5) Fig 4 S1A shows colabeling of l-LNvs and Gad1-Gal4 expressing neurons. They are almost 100% overlapping signals. This would indicate that the l-LNvs are GABAergic themselves, or that there is a problem with this experiment.

(6) Fig 4 S1B: Again, I can see colabelling of the GFP and PDF staining, suggesting that Gad1-Gal4 expresses in l-LNvs.

(7) Line 184: "Consistently, knocking down Rdl in the l-LNvs rescues the long sleep phenotype of cry mutants (Figure 4-figure supplement 1D)." This statement is incorrect as the driver used for this experiment, 78G01-GAL4 is not specific to the l-LNvs, so it is possible that the phenotypes observed are not coming from these neurons.

(8) Figure 4G-K: None of these manipulations are specific to the l-LNvs. The authors describe 10H10-GAL4 and 78G01-GAL4 as l-LNvs specific tools, but this is not the case. Why not use the SS00681 Split-GAL4 line described in Liang et al., 2017 PMID: 28552314? It is possible that some of the effects reported in this manuscript are not caused by manipulating the l-LNvs.

(9) Similarly for the manipulation of s-LNvs, the authors cannot rule out effect that are coming from other cells as R6-GAL4 is not specific to s-LNvs.

(10) The staining presented in Fig 5 S1 is not very convincing. Difficult to see whether Gad1-GAL4 only expresses in the s-LNvs.

Reviewer #3 (Public review):

Summary:

In humans, short photoperiods are associated with hypersomnolence. The mechanisms underlying these effects is however, unknown. Chen et al. use the fly Drosophila to determine the mechanisms regulating sleep under short photoperiods. They find that mutations in the circadian photoreceptor cryptochrome (cry) increase sleep specifically under short photoperiods (e.g. 4h light : 20 h dark). They go on to show that cry is required in GABAergic neurons and that the effects of the cry mutation on sleep are mediated by alterations in GABA signalling. Further, they suggest that the relevant subset of GABAergic neurons are the well-studied small ventral lateral neurons that they suggest inhibit the arousal promoting large ventral neurons via GABA signalling

Strengths:

Genetic analysis to show that cryptochrome (but not other core clock genes) mediates the increase in sleep in short photoperiods, and circuit analysis to localise cry function to GABAergic neurons.

Weaknesses:

The authors' have substantially revised their manuscript, and the manuscript is better for the revisions. However, the conclusion that the sLNvs are GABAergic is unfortunately still not well supported by the data. A key sticking point remains the anti GABA immunostaining, and specific driver lines for sLNvs and lLNvs.

The authors should tone down their conclusions to reflect the fact that their data, as presented, does not support the model that cry acts in sLNvs to modulate GABA signalling onto lLNvs and thus modulate sleep.

Reviewer #4 (Public review):

Summary:

Short photoperiod is an important experimental manipulation in neurobiology, endocrinology, and metabolism studies. However, the molecular mechanisms by which short photoperiod gives rise to behavioral phenotypes that are seen in seasonal affective disorders remain unknown. Using the classic circadian model organism Drosophila, this study examines short photoperiod-induced hypersomnolence and identifies the circadian photoreceptor cryptochrome as a regulator of GABAergic tone within the clock neural circuit to promote wakefulness under short photoperiod conditions. The discovery has broad implications for understanding how short photoperiod modulates neural inhibition in circadian circuits in regulating sleep.

Strengths:

The Drosophila model provided a powerful platform to dissect the molecular mechanisms underlying short photoperiod-induced hypersomnolence. A battery of behavioral, imaging, circuit-manipulation approaches was employed to test the novel hypothesis that the circadian photoreceptor cryptochrome modulates GABAergic tone within the clock neural circuit to promote wakefulness under short photoperiod conditions.

Weaknesses:

The current model proposed by the authors suggests that the small ventral lateral neurons of the Drosophila clock circuit are GABAergic; however, this remains unclear. At present, the field lacks sufficient data and validated reagents to definitively establish the GABAergic identity of these neuropeptidergic neurons.

Reviewer #1 (Public review):

Summary:

The authors developed a new autofocusing method, LUNA (Locking Under Nanoscale Accuracy), to address severe focus drift-a major challenge in time-lapse microscopy. Using this method, they tackle a fundamental question in bacterial cold shock response: whether cells halt growth and division following an abrupt temperature downshift. Overall, the experimental design, modeling, and data analysis are solid and well executed. However, several points require clarification or further support to fully substantiate the authors' conclusions.

Strengths:

(1) The LUNA method outperforms existing autofocusing systems with nanoscale precision over a large focusing range. The focusing time is reasonable for the presented experiments, and the authors note potential improvements by using faster motors and optimized control algorithms, suggesting broad applicability. The theoretical simulations and experimental validation provide solid support for the robustness of the method.

(2) Using LUNA, the authors address a long-standing question in bacterial physiology: whether cells arrest growth and division after an abrupt cold shock. Single-cell analyses monitoring the entire course of cold adaptation and steady-state growth reveal features that are obscured in bulk-culture studies: cells continue to grow at reduced rates with smaller cell sizes, resulting in an apparently unchanged population-level OD. The experiments are well designed and analyses are generally solid and largely support the authors' conclusions.

(3) The authors also propose a model describing how population-level OD measurements depend on cell dry mass density, volume, and concentration. This provides a valuable conceptual contribution to the interpretation of OD-based growth measurements, which remain a gold-standard method in microbiology.

Weaknesses:

(1) It is unclear whether the author's model explaining the population-level OD during acclimation is broadly applicable. Most analyses focus on a shift from 37˚C to 14˚C, where the model agrees well with experimental data. However, in the 37˚C to 12˚C experiment, OD600 decreases after cold shock (Fig. 5e), and the computed OD does not match the experimental measurements (Fig. S16a). Although the authors attribute this discrepancy to a "complicated interplay," no further explanation is provided, which limits confidence in the model's general applicability.

(2) The manuscript proposes that cell-cycle progression becomes synchronized across the population after cold shock, but the supporting evidence is not fully convincing. If synchronization refers primarily to the uniform reduction in growth rate following cold shock, this could plausibly arise from global translation inhibition affecting all cells. However, the additional claim that "cells encountering a relatively late CSR will accelerate division to maintain synchronization" is not strongly supported by the presented data.

(3) Several technical terms used in the method development section are not clearly defined and may be unfamiliar to a broad readership, which makes it difficult to fully understand the methodology and evaluate its performance. Examples include depth of focus, focusing precision, focusing time, focusing frequency, and drift threshold value. In addition, the reported average focusing time per location (~0.6 s) lacks sufficient context, limiting the reader's ability to assess its significance relative to existing autofocusing methods.

Reviewer #2 (Public review):

Summary:

This study presents LUNA, an autofocus method that compensates for focus drift during rapid temperature changes. Using this approach, the authors show that E. coli cells continue to grow and divide during cold shock, revealing a coordinated, multi-phase adaptation process that could not be deduced from traditional population measurements. They propose a scattering-theory-based model that reconciles the paradox between growth differences of the bacteria at the single-cell level vs population level.

Strengths:

(1) The LUNA approach is pretty creative, turning coma aberration from what is normally a nuisance into an exploit. LUNA enabled long-term single-cell imaging during rapid temperature downshifts.

(2) The authors show that the long-assumed growth arrest during cold shock from population-level measurements is misleading. At the single-cell level, bacteria do not stop growing or dividing but undergo a continuous, three-phase adaptation process. Importantly, this behavior is highly synchronized across the population and not based on bet-hedging.

(3) Finally, the authors propose a model to resolve a long-standing paradox between single-cell vs population behavior: if cells keep growing, why does optical density (OD) of the culture stop increasing? Using light-scattering theory, they show that OD depends not only on cell number but also on cell volume, which decreases after cold shock. As a result, OD can remain flat, or even decrease, despite continued biomass accumulation. This demonstrates that OD is not a reliable proxy for growth under non-steady conditions.

Weaknesses:

(1) While the authors theoretically explain the advantages of LUNA over existing autofocus methods, it is unclear whether practical head-to-head comparisons have been performed, apart from the comparison to Nikon PFS shown in Video S1. As written, the manuscript gives the impression that only LUNA can solve this problem, but such a claim would require more systematic and rigorous benchmarking against alternative approaches.

(2) No mutants/inhibitors used to test and challenge the proposed model.

(3) Cells display a high degree of synchronization, but they are grown in confined microfluidic channels under highly uniform conditions. It is unclear to what extent this synchrony reflects intrinsic biology versus effects imposed by the microfluidic environment.

(4) To further test and generalize the model, it would be informative to also examine bacterial responses at intermediate temperatures rather than focusing primarily on a single cold-shock condition.

Reviewer #1 (Public review):

Summary

The manuscript by Peden-Asarch et al. introduces MPS, a new open-source software package for processing miniscope data. The authors aim to provide a fast, end-to-end analysis pipeline tailored to miniscope users with minimal experience in coding or version control. The work addresses an important practical barrier in the field by focusing on usability and accessibility.

Strengths

The authors identify a clear and well-motivated need within the miniscope community. Existing pipelines for miniscope data analysis are often complex, difficult to install, and challenging to maintain. In addition, users frequently encounter technical limitations such as out-of-memory errors, reflecting the substantial computational demands of these workflows-resources that are not always available in many laboratories. MPS is presented as an attempt to alleviate these issues by offering a more streamlined, accessible, and robust processing framework.

Weaknesses

The authors state that "MPS is the first implementation of Constrained Non-negative Matrix Factorization (CNMF) with Nonnegative Double Singular Value Decomposition (NNDSVD) initialization." However, NNDSVD initialization is the default method in scikit-learn's NMF implementation and is also used in CaIMAN. I recommend rephrasing this claim in the abstract to more accurately reflect MPS's novelty, which appears to lie in the specific combination of constrained NMF with NNDSVD initialization, rather than being the first use of NNDSVD initialization itself.

At present, there are practical issues that limit the usability of the software. The link to the macOS installer on the documentation website is not functional. Furthermore, installation on a MacBook Pro was unsuccessful, producing the following error:<br /> "rsync(95755): error: ... Permission denied ... unexpected end of file."

For the purposes of this review, resolving this issue would significantly improve the evaluation of the software and its accessibility to users.

More broadly, the authors propose self-contained installers as a solution to the "package-management burden" commonly associated with scientific software. While this approach is appealing and potentially useful for novice users, current best practices in software development increasingly rely on continuous integration and continuous deployment (CI/CD) pipelines to ensure reproducibility, testing, and long-term maintenance. In this context, it has become standard for Python packages to be distributed via PyPI or Conda. Without dismissing the value of standalone installers, the overall quality and sustainability of MPS would be greatly enhanced by also supporting conventional environment-based installations.

Reviewer #2 (Public review):

Summary:

This manuscript introduces Miniscope Processing Suite (MPS), a novel no-code GUI-based pipeline built to easily process long-duration one-photon calcium imaging data from head-mounted Miniscopes. MPS aims to address two large problems that persist despite the rapid proliferation of Miniscope use across the field. The first issue is concerned with the high technical barrier to using existing pipelines (e.g., CaImAn, MIN1PIPE, Minian, CaliAli) that require users to have coding skills to analyze data. The second problem addressed is the intense memory limitations of these pipelines, which can prevent analysis of long-duration (multi-hour) recordings without state-of-the-art hardware. The MPS toolbox takes inspiration from what existing pipelines do well, innovates new modules like Window Cropping, NNDSVD initialization, Watershed-based segmentation, and improves the user experience to improve access to calcium imaging analysis without the need for new training in new coding languages. In many ways, MPS achieves this aim, and thus will be of interest to a growing, broad audience of new calcium imagers.

There are, however, some concerns with the current manuscript and pipeline that, if addressed, would greatly improve the impact of this work. Currently, the manuscript provides insufficient evidence that MPS can generate good results efficiently on various data sets, and it is not properly benchmarked against other established packages. Additionally, considering the goal of MPS is to attract novices to attempt Miniscope analysis, better tutorials, documentation, and walkthroughs of expected vs inaccurate results should be provided so that it is clear when the user can trust the output. Otherwise, this simplified approach may end up leading new users to erroneous results.

Strengths:

The manuscript itself is well-organized, clear, and easy to follow. MPS is clearly designed to remove the computational barrier for entry for a broad neuroscience community to record and analyze calcium data. The development of several well-detailed algorithmic innovations merits recognition. Firstly, MPS is extremely easy to install, keep updated, and step through. Having each step save every output automatically is a well-thought-out feature that will allow users to enter back into the pipeline at any step and compare results.

The implementation of an erroneous frame identifier and remover during preprocessing is an important new feature that is typically done offline with custom-built code. Interactive ROI cropping early in the pipeline is an efficient way to lower pixel load, and NNDSVD initialization is a new way to provide nonnegative, biologically interpretable starting spatial and temporal factors for later CNMF iterations. Parallel temporal-first update ordering cuts down dramatically on later computational load. Together, all these features, neatly packaged into a no-code GUI like the Data Explorer for manual curation, are practical additions that will benefit end users.

Weaknesses:

A major limitation of this manuscript is that the authors don't validate the accuracy of their source extraction using ground-truth data or any benchmark against existing pipelines. The paper uses their own analysis of processing speeds, component counts, signal-to-noise ratio improvements, and morphological characteristics of detected cells, but it needs to be reworked to include some combination of validation against manually annotated ground truth data sets, simulated data with known cell locations and activity patterns, or cross-validation with established pipelines on identical datasets. Without this kind of validation, it is impossible to truly determine whether MPS produces biologically acceptable results that help distinguish it from what is currently already available. For example, line 57 refers to the CaImAn pipeline having near-human efficiency (Figures 3-5 and Tables 1 and 2 of the CaImAn paper), but no specific examples for MPS performance benchmarks are made. Figure 15 of the Minian paper provides other examples of how to show this.

Considering one of the main benefits of MPS is its low memory demand and ability to run on unsophisticated hardware, the authors should include a figure that shows how processing times and memory usage scale with dataset sizes (FOV, number of frames and/or neurons, sparsity of cells) and differing pipelines. Figure 8 of the CaImAn paper and Figure 18 of the Minian paper show this quite nicely. Table 1 currently references how "traditional approaches" differ methodologically from MPS innovations, but runtime comparisons on identical datasets processed through MPS, CaImAn, Minian, or CaliAli would be necessary to substantiate performance claims of MPS being "10-20X faster". Additionally, while the paper does mention the type of hardware used by the experimenters, a table with a full breakdown of components may be useful for reproducibility. As well as the minimum requirements for smooth processing.

The current datasets used for validating MPS are not described in the manuscript. The manuscript appears to have 28 sessions of calcium imaging, but it is unclear if this is a single cohort or even animal, or whether these data are all from the same brain region. Importantly, the generalizability of parameter choices and performance could vary for others based on brain region differences, use of alternative calcium indicators (anything other than GCaMP8f used in the paper), etc. This leads to another limitation of the paper in its current form. While MPS is aimed at eliminating the need to code, users should not be expected to blindly trust default or suggested parameter selections. Instead, users need guidance on what each modifiable parameter does to their data and how each step analysis output should be interpreted. Perhaps including a tutorial with sample test data for parameter investigation and exploration, like many other existing pipelines do, is warranted. This would also increase the transparency and reproducibility of this work.

Currently, the documentation and FAQ website linked to MPS installation does not do an adequate job of describing parameters or their optimization. The main GitHub repository does contain better stepwise explanations, but there needs to be a centralized location for all this information. Additionally, a lack of documentation on the graphs created by each analysis step makes it hard for a true novice to interpret whether their own data is appropriately optimized for the pipeline. Greater detail on this would greatly improve the quality and impact of MPS.

Reviewer #2 (Public review):

Summary:

In this work, the authors investigate the role of fluid flow in shaping the colony size of a freshwater cyanobacterium Microcystis. To do so, they have created a novel assay by combining a rheometer with a bright field microscope. This allows them to exert precise shear forces on cyanobacterial cultures and field samples, and then quantify the effect of these shear forces on the colony size distribution. Shear force can affect the colony size in two ways: reducing size by fragmentation and increasing size by aggregation. They find limited aggregation at low shear rates, but high shear forces can create erosion-type fragmentation: colonies do not break in large pieces, but many small colonies are sheared off the large colonies. Overall, bacterial colonies from field samples seem to be more inert to shear than laboratory cultures, which the authors explain in terms of enhanced intercellular adhesion mediated by secreted polysaccharides.

Strengths:

-This study is timely, as cyanobacterial blooms are an increasing problem in freshwater lakes. They are expected to increase in frequency and severeness because of rising temperatures, and it is worthwhile learning how these blooms are formed. More generally, how physical aspects such as flow and shear influence colony formation is often overlooked, at least in part because of experimental challenges. Therefore, the method developed by the authors is useful and innovative, and I expect applications beyond the presented system here.

-A strong feature of this paper is the highly quantitative approach, combining theory with experiments, and the combination of laboratory experiments and field samples.

Weaknesses:

This study has no major weaknesses. Although the initial part of the introduction seems to imply that fluid flow is the predominant factor in shaping cyanobacterial colony (de)formation, the ensuing discussion is sufficiently nuanced for the reader to understand that the multicellular lifestyle of cyanobacterium Microcystis is shaped by multiple effects, that include bacterial behavior (e.g. which and how much EPS is produced), environmental variables that control cellular aggregation or adhesion and, indeed, fluid flow.

Joint Public Review:

Summary:

The authors state the study's goal clearly: "The goal of our study was to understand to what extent animal individuality is influenced by situational changes in the environment, i.e., how much of an animal's individuality remains after one or more environmental features change." They use visually guided behavioral features to examine the extent of correlation over time and in a variety of contexts. They develop new behavioral instrumentation and software to measure behavior in Buridan's paradigm (and variations thereof), the Y-maze, and a flight simulator. Using these assays, they examine the correlations between conditions for a panel of locomotion parameters. They propose that inter-assay correlations will determine the persistence of locomotion individuality.

Comments from the editors on the latest version:

In the latest communication, the authors were asked to (i) justify their selection of metrics (i.e. why these specific five behavioural metrics were chosen from the many recorded), (ii) discuss the variation in ICCs, and (iii) in light of this variation and the reliance on a few selected behavioural parameters, tone down the general claim so as not to overstate that individuality persists across all behaviours.

We note that the justification for choosing the five metrics and the discussion of ICC variation are purely qualitative, and, despite the edits, the manuscript continues to frame individual behaviours as broadly stable.

Reviewer #1 (Public review):

Summary:

The authors present a compelling case for the necessity of age-specific templates in functional hyperalignment. Given that the brain undergoes substantial developmental, structural, and functional changes across the lifespan, a 'one-size-fits-all' canonical template is often insufficient. This study effectively demonstrates that incorporating age-congruent features significantly enhances the performance and sensitivity of hyperalignment models. By validating these findings across two independent datasets (Cam-CAN and DLBS), the paper provides robust evidence that accounting for age-related functional organization is a critical prerequisite for accurate functional alignment in lifespan research.

Strengths:

(1) The authors used three metrics to evaluate performance. Across all metrics, they found that age-congruent templates outperformed age-incongruent templates, suggesting that age-specific templates can improve alignment.

(2) These findings highlight the superiority of age-congruent templates for hyperalignment. This work underscores the importance of age-matching in cross-subject functional mapping and represents a vital step forward for the methodology.

Weaknesses:

(1) Participant Demographics and Group Separation:

The study defines the 'older' cohort as 65-90 years and the 'younger' cohort as 18-45 years. While this 20-year gap (ages 46-64) effectively maximizes the contrast between groups, the results in Figure 4a suggest that the predicted individualized connectomes follow a continuous distribution. Given this continuity, could the authors provide the average median trends for Figures 2a and 2b to illustrate how the model behaves across the missing age range?

(2) Request for Implementation:

I have been unable to locate the source code associated with this publication. Could the authors please provide a link to the repository or clarify if the implementation is available for reproduction?

(3) Analysis of Prediction Performance and Distribution:

While Figures 3b and 5b clearly demonstrate that the congruent template improves correlation, Figure 4a shows a distinct shift in the scatter distribution. Could the authors provide a detailed explanation of the prediction performance metrics used? Specifically, I would like to understand how the underlying method accounts for the distribution differences observed when applying the congruent template.

Reviewer #2 (Public review):

Summary:

In this study, Zhang and colleagues examine the role of participant selection in creating and using functional templates to improve analyses using hyperalignment. Hyperalignment aligns participants' functional MRI data to a shared functional template, analogous to the anatomical templates used to bring anatomical MRI data into a shared space (e.g., MNI152). The question of appropriate template creation is especially pressing for population-level analyses, where a large number of demographic groups (e.g., different age ranges, clinical statuses) may be included in the same analysis. These different demographic groups may have differences in their functional organization that complicate the creation of a single study-specific functional template.

To provide an initial investigation of the potential effect of demographic-specific templates, the authors use the publicly available Cam-CAN dataset, which contains participants from 18 to 87 years of age. They define a young adult (< 45 years of age) and an older adult group (> 65 years of age) from this dataset with approximately the same number of participants. They investigate whether "age-congruent" templates (i.e. defined in the same age group they are used) improve three analyses where hyperalignment has been previously shown to boost performance: inter-subject correlation, predicting individual connectomes, and predicting individual functional responses. Using the Cam-CAN-derived older adult template, they then replicate the ISC analyses using the publicly available Dallas Lifespan Brain Study (DLBS).

Overall, the presented results are highly suggestive that age-congruent templates consistently improve performance, though the absolute effects are small.

Strengths:

The use of a separate validation sample, reusing the same template calculated with Cam-CAN, highlights the potential of developing independent templates for individual demographic groups and then distributing these for wider use, analogous to the MNI templates that are widely used throughout the field of neuroimaging. This suggests that the potential impact of this framework is significant.

Weaknesses:

While the authors appropriately highlight the potential applications of this result (e.g., to different clinical statuses), it is not apparent how to appropriately extend this methodology to many common experimental paradigms. For example, in case-control studies (where researchers are interested in comparing clinical and non-clinical participants) the use of two different functional templates may complicate rather than ease analyses. Providing this as a potential limitation of the current template construction method, or providing recommendations to researchers interested in comparing across groups, would help to increase the impact of this work.

Reviewer #1 (Public review):

Summary:

The manuscript examines the factors that restrict the induction of IL-17-producing T cells during Mycobacterium tuberculosis (Mtb) infection. The authors show that neither the infectious route nor the duration of infection is responsible. But they do show that mice that lack the Th1-defining transcription factor, a finding consistent with prior reports in the field of immunology. They also show that 2 highly attenuated Mtb mutants in ESX-1 and PDIM, two well-known Mtb virulence factors, do induce IL-17-producing T cells. In contrast, Mtb mutants in mmpl4 are also similarly attenuated, but do not induce IL-17-producing T cells, suggesting that this property is not simply a result of attenuation but due to specific properties of ESX-1 and PDIM-deficient mutants.

Strengths:

(1) It is interesting that mice infected with ESX-1 and PDIM mutants have increased induction of Th17 cells.

(2) The data are solid and convincing throughout.

Weaknesses:

There are two main criticisms:

(1) It is not clear how much the factors uncovered here are true beyond B6 mice. B6 mice, compared to humans, are known to be very Th1-skewed, and Tbet is a strong inhibitor of Th17-specific T cells. Many people make IL-17-producing T cells in response to Mtb infection.

(2) Very few novel insights are mechanistically revealed about how Th17 induction is restricted by Mtb. Tbet induction is known to restrict Th17 development, and this is a T-cell intrinsic mechanism. In contrast, the IL-23 association revealed seems to be extrinsic to T cells and to act on T cells. How, if at all, are these factors related to each other in restricting Th17 induction? Also, the conclusion that it is not a result of attenuation is not completely convincing.

Other points:

(1) The authors show that mice infected with a deficiency in ESX-1 have more IL-17-producing CD4 T cells in response to stimulation with an ESAT-6 peptide pool (Figure 3B). Because ESAT-6 is encoded by ESX-1, why do mice infected with this Mtb mutant have any ESAT-6-specific T cells? Is it an incomplete knockdown?

(2) The manuscript states, "Under the conditions where Th17s are highly induced, mice infected with either ΔESX-1 or PDIM lacking Mtb, the Il17a-/- mice had ~3-5 fold higher CFU than WT mice (Figures 3F-G). These results indicate that the induction of Th17s is not dependent on the attenuation of Mtb in general, but instead Mtb utilizes ESX-1 and PDIM to suppress the induction of a Th17 response that enhances protection against Mtb infection." I don't think the last sentence is necessarily true. I can imagine a scenario in which the induction of the Th17s is, in fact, due to the attenuation, and the Th17 induction still contributes to protection.

(3) ESX-1, PDIM, and mmpl4 mutants all have similarly reduced CFUs in the lung, but what about the LN? The bacterial burden in the LN may be more important for regulating T-bet, IL-23, and Th17 differentiation, since the LN is where T cell priming occurs, than the CFU in the lung. Perhaps ESX-1 and PDIM mutants have reduced CFU in the LN, but mmpl4 does not. This difference in LN burdens may be the primary driver of Th17 priming, as high avidity interactions are thought to be an important driver of T-bet induction.

(4) Do LN cDC1 and high levels of IL-12 p35 in mice infected with the mmpl4 mutant? Likewise, LN cDC2's express low levels of IL-12 p19 (akin to those infected with WT Mtb)? If these observations for ESX-1 and PDIM mutants are mechanistically linked to the increased numbers of Th17 cells, then you would expect mice infected with mmpl4 mutants to be more like those infected with WT Mtb than those infected with ESX-1 and PDIM mutants.

(5) ESX-1 and PDIM are very different virulence factors - a protein secretory pathway and cell wall lipid, respectively? Mechanistically, how would mutants in these pathways give very similar outcomes regarding Th17 cells unless it was simply as an aspect of their attenuation? Perhaps, mmpl4 mutants simply differ in some aspects of their attenuation, such as bacterial burdens in LNs, or their interaction with cDCs?

Reviewer #2 (Public review):

Summary:

In this manuscript, the authors tackle an important question of why IL-17 production and TH17 responses are lower than expected during Mtb infection. The authors identify an axis of cross-regulation between TH1 and TH17 cells and provide data to support roles for Mtb virulence factors ESX1 and PDIM in promoting TH1 responses and/or suppressing TH17 responses.

Strengths:

The strengths include the significance of the work, the combination of host and Mtb genetic models to dissect the mechanistic basis for regulation of IL-17 production from T cells during infection, and the rigor of the experiments. There are a number of exciting findings from the work, including the cross-talk between T cell responses and the impact of ESX1 and PDIM on these responses.

Weaknesses:

The following conclusions and interpretations should be revisited, rephrased, and re-evaluated:

(1) The manuscript neglects to analyze T cell responses in the dLN, which is the critical site where these responses are initiated (only DC cytokine production is measured in the dLN). The differences in the lungs could reflect trafficking of T cells to the lungs, local lung T cell responses, or durability of the T cell responses in the lungs. The authors state in the last results section that "These results indicate that the ESX-1 and PDIM virulence factors impact naïve T cell differentiation at the draining mediastinal lymph node..." but T cell responses are never measured in the dLN.

(2) Figure 2: The authors state that "Importantly, IFN-γ deficient mice did not exhibit elevated levels of IL-17A producing CD4 T cells demonstrating that IFN-γ production is not the mechanism by which Th1 T cells limit a Th17 response during Mtb infection", but the difference is significantly different and even more obvious in Panel B. In fact, if the Panel D y-axis was on a log scale, the Ifng-/- would likely look more like Tbet-/- than WT. Based on this data, it seems like IFNg is having an effect and should not be completely discounted. Does the deletion of Ifng affect the number of Tbet+ T cells?

In addition, the deletion of Tbet results in an increased number of IFNg+IL-17+ double positive T cells (Figure 2B), in addition to a sizable IFNg single positive T cell population maintained in the Tbet-/- mice (10x the negative control of Ifng-/-). Is this why Tbet deletion is not as severe as Ifng deletion, because T cells are still making IFNg?

Along these lines, the statement in the text that, "Tbet-/-Il17a-/- mice completely lacked both IFN-γ producing...." T cells is not supported by the data in Figure 2C. Tbet-/-Il17a-/- mice look to have more gamma-producing T cells than Tbet-/- mice (which is already 10x the negative control of Ifng-/- in panel 2B if one includes the gamma single positive and IFNg/IL-17 double positive).

(3) In the Results sections describing Figures 3, 4, and 5, the authors equate IL-17 production by T cells with TH17 responses and IFNg expression with TH1, but Tbet and RORgt expression in the T cells should be measured to make conclusions about TH1 and TH17. Or the authors can rephrase their findings to specifically state the observations as IFNg or IL-17 expressing CD4+ T cells.

(4) Conceptually, do the authors think that ESX1/PDIM promotes TH1 responses and this blocks TH17 or are ESX1/PDIM blocking TH17 responses directly, allowing for increased TH1 responses? It would be helpful to clarify the model in this regard, describe how the data supports one model or the other, and then make sure the language is consistent throughout. Can these effects on T cell responses be tested and recapitulated in vitro using infected APC and T cell co-cultures?

Reviewer #3 (Public review):

Summary:

The manuscript by Zilinskas et al seeks to understand the mechanisms underlying the ability of Mtb to suppress Th17 differentiation. As Th17 responses are needed for protective immunity against TB, this is an important topic of investigation. They use Mtb mutants that lack eccC1 (from the ESX-1 locus) and fadD28 (encoding PDIM) and implicate a Tbet-dependent pathway by which Mtb modulates Th17 differentiation. The mechanism by which ESX-1/PDIM function to impact Th17 differentiation is, however, unclear, which limits the novelty of the results.

Strengths:

Understanding how Mtb limits Th17 differentiation has implications for vaccine development. Comparative study of KO mice and Mtb mutants is a strength.

Weaknesses:

(1) The authors should acknowledge and reference key findings from the literature that have identified suppression of Th17 differentiation as an Mtb virulence mechanism, e.g., the role of the Hip1 protease and CD40 signaling (Madan-Lala JI 2014, Sia Plos Path 2017, Enriquez iScience 2022) and Khader JI 2005, showing the requirement of IL-23 for Th17 responses in vivo in a TB mouse model.

(2) Addressing several questions related to the Tbet KO mouse experiments would strengthen the study. Do the Tbet KO mice have elevated IL-4/5/13 (which has been previously reported in non-TB studies) in addition to IL-17? The lack of Th17 cells in the IFNg KO compared to the Tbet KO may be due to a difference in timing, since only 3-week data are shown; earlier and later time points would provide better interpretation. The authors do not present any data on neutrophil infiltration in WT vs Tbet KO vs IFNg KO mice. Since IL-17 is known to be important for recruiting neutrophils to the lung, data on neutrophils are important for clarifying the mechanism for the CFU outcomes.

(3) While IL-23 is important for sustaining IL-17 production, IL-6, TGF-b and/or IL-1β are necessary for Th17 polarization. What were the levels of these cytokines in DCs in the lung? (Figure 5). Additionally, Tbet-deficient DCs exhibit impaired activation of antigen-specific Th1 cells and have reduced IL-12 production. Given the data showing higher IL-17 levels in Tbet KO mice, the authors should provide information on the DC phenotype (IL-23, IL-6, etc.) in the Tbet KO experiments.

(4) The mechanism by which ESX-1/PDIM function to impact Th17 differentiation is not clear. While data showing a role for ESX-1 and PDIMs in inhibiting Th17 responses is interesting, there is no insight into the potential mechanism of action. Figure 3 showing reduction in IFNg+ CD4 T cells after infection with eccC1 and fadD28 mutants suggests that this outcome is due to a lower bacterial load relative to WT Mtb at the 3-week time point. Since IFNg is known to suppress IL-17, the higher levels of Th17 cells could be due to the reduction in IFNg due to the attenuated growth of the mutants. Additionally, what was the level of Type I IFNs elicited by these mutants?

(5) Since macrophages have been implicated in the reduced cytokines seen in the ESX-1 mutant, IL-23 and other cytokine data on lung macrophages would complement the DC data.

(6) Figure 5. There are many fewer DCs overall in the eccC1 and fadD28 mutant groups, which could account for the increased % IL-23p19 in DCs (5D). What were the levels of IL-23 in DC1s?

Reviewer #1 (Public review):

Summary:

In this study, the authors examine the effect of Chlorpyrifos (CPF) exposure on zebrafish social development. They expose larval zebrafish to CPF (0 - 3 dpf), and report social deficits at juvenile stages. They show that the gut microbial metabolite butyrate can rescue these social deficits, proposing that butyrate acts as a histone deacetylase (HDAC) inhibitor, given that inhibition of some HDACs can also rescue social deficits. They also show that CPF changes neuronal gene expression, and butyrate partially rescues these changes. Finally, they demonstrate changes in gut microbiome and metabolome composition, pointing to potential modulation of nitrogen metabolism pathways. They then hypothesise that NO can modulate HDAC activity and attempt to link the NO pathway to social behavior.

Strengths:

The authors demonstrate an interesting link between early Chlorpyrifos (CPF) exposure and later-life social deficits, such as changes in neuronal gene expression, including some autism-related genes, and provide solid evidence that butyrate and epigenetic modulation (histone deacetylase inhibition) may be involved.

They also comprehensively characterise the microbiome and metabolome of CPF-exposed zebrafish, providing a useful resource for further investigation into its gut-brain mechanisms.

They are cautious in framing some of their conclusions as a hypothesis and provide some suggestions for future analyses.

Weaknesses:

The claim that butyrate's effects on CPF-induced social deficits and neuron activity changes are mediated by histone deacetylase inhibition is lacking some additional controls and, hence, is not completely supported.

Details on the social behavior assay performed and other potential morphological or behavioral changes were not provided.

Claims on the mechanism of action of CPF are inconclusive. The causal role of the gut microbiome is not established, especially since gut microbial dysbiosis may also be a downstream consequence of direct effects of CPF on the host, such as changes in host gut gene expression. Evidence for the role of nitrogen metabolism is also incomplete, and the authors have not discussed or ruled out the potential alternative mechanism of reduced butyrate production due to gut microbiome changes.

Reviewer #2 (Public review):

Summary:

This paper by Diaz et al. uses the zebrafish model to examine how early embryonic exposure to Chlorpyrifos (CPF), a widely used organophosphate pesticide, induces social behavior deficits later in life. This paper combined behavioral testing, pharmaceutical treatment, genetic manipulation, and multi-omics to test the hypothesis that early CPF increases the abundance of denitrifying bacteria, Pseudomonas, which, in turn, enhances nitric oxide production and induces selective inhibition of HDAC8 and abnormal gene expression in the brain.

Strengths:

(1) The observation that early embryonic CPF exposure causes behavior deficits in juvenile zebrafish is very intriguing. It is especially exciting to see that CPF-induced behavior deficits can be reversed by overnight treatment with butyrate or HDAC1 inhibitors in juvenile zebrafish. In humans, CPF exposure during pregnancy causes brain abnormalities and neurological disorders such as Autism. Though it is far away from the zebrafish experimental study to human application, the experimental effects reported in the paper are still quite thought-provoking.

(2) The authors performed RNA sequencing experiments on control zebrafish, CPF-exposed zebrafish, and CPF-exposed zebrafish that were treated with Butyrate. The data not only showed large-scale transcriptomic changes in the juvenile zebrafish brain in response to embryonic CPF exposure but also showed that many CPF-induced genetic alterations can be alleviated by butyrate exposure later in life.

(3) The authors also performed untargeted metabolomics on zebrafish gut and metagenomic analysis in zebrafish feces samples. The results are interesting and support the conclusion that increased Intestinal Nitric oxide metabolism and the abundance of denitrifying bacteria, such as Pseudomonas, are associated with CPF exposure.

(4) The large datasets presented in the paper will be useful to other researchers interested in understanding how CPF or butyrate alters brain and gut function. It might be useful to generate new hypotheses to power other research lines.

(5) The social preferences, behavior testing, and experimental paradigm used by the paper may also be used by other researchers to investigate the interaction among gene, environmental factors, and brain function.

Weaknesses:

(1) The presented link between gut microbiome and CPF-induced behavior and genetic alteration is an association, but not causation. Although the research data align with the hypothesis, the hypothesis is not fully supported or tested by the data presented in the paper in the current state.

(2) The authors performed several large omic studies. However, some of the presented analyses are relatively simple and incomplete. For example, the authors performed shotgun metagenomic analysis on zebrafish feces. However, the paper only displayed the bacterial taxa differences. Are there any differences in bacterial genetic pathways, especially the pathways associated with microbial nitrogen metabolism? What is the alpha and beta diversity looking like when comparing different experimental groups?

Reviewer #1 (Public review):

Summary:

In this manuscript, the authors investigate how the anterior claustrum may integrate temporally separated task-relevant signals to guide behavior in a delayed escape paradigm. Because in vivo neural recordings from claustrum during this task are extremely limited - comprising single-trial data with small neuronal samples - the authors adopt a modeling-driven approach. They train recurrent neural networks (RNNs) using only behavioral data (escape latency) to reproduce task performance and then analyze the internal dynamics of the trained networks. Within these networks, they identify a subset of units whose activity exhibits persistent responses and strong correlations with behavior, which the authors label as "claustrum-like." Using dimensionality reduction, decoding, and information-theoretic analyses, they argue that these units dynamically integrate conditioned stimulus (CS) and door-opening signals via nonlinear, trajectory-based population dynamics rather than fixed-point attractor states.

To bridge model predictions and biology, the authors complement the modeling with in vitro slice experiments demonstrating recurrent excitatory connectivity and prolonged activity in the anterior claustrum that depends on glutamatergic transmission. They further compare latent neural trajectories derived from previously published in vivo claustrum recordings to those observed in the RNN, reporting qualitative similarities. Based on these results, the authors propose that the claustrum implements temporal signal integration through recurrent excitatory circuitry and dynamic population trajectories, potentially supporting broader theories of integrative brain function.

Strengths:

This study addresses an important and challenging problem: how to infer population-level computation in a brain structure for which in vivo data are sparse and experimentally constrained. The authors are commendably transparent about these limitations and seek to overcome them through a principled modeling framework. The integration of behavioral modeling, RNN analysis, and slice electrophysiology is ambitious and technically sophisticated.

Several aspects stand out as strengths. First, the behavioral RNN is carefully trained and interrogated using a rich set of modern analytical tools, including cross-temporal decoding, trajectory analysis, and partial information decomposition, providing multiple complementary views of network dynamics. Second, the slice experiments convincingly demonstrate recurrent excitatory connectivity in the anterior claustrum, lending biological plausibility to the model's reliance on recurrent dynamics. Third, the manuscript is clearly written, logically organized, and conceptually engaging, and it offers a coherent mechanistic hypothesis that could guide future large-scale recording experiments.

Importantly, the work has significant heuristic value: rather than merely fitting data, it attempts to generate testable computational ideas about claustral function in a regime where direct empirical access is currently limited.

Weaknesses:

Despite these strengths, the manuscript suffers from a recurring and substantial conceptual issue: systematic over-interpretation of model-data correspondence. While the modeling results are potentially insightful, the extent to which they are presented as recapitulating real claustral neural mechanisms goes beyond what the available data can support.

A fundamental limitation is that the RNN is trained solely on behavioral output, without being constrained by neural data at either single-unit or population levels. As a result, the internal network dynamics are underdetermined and non-unique. Many distinct internal solutions could plausibly generate identical behavior. However, the manuscript frequently treats the specific internal solution discovered in the RNN as if it were a close approximation of the actual claustrum circuit.

This issue is compounded by the sparse nature of the in vivo data used for comparison. The GPFA-based trajectory analyses rely on pseudo-populations and single-trial recordings, yet are interpreted as evidence for robust population-level dynamics. Because neurons were not recorded simultaneously, the inferred trajectories necessarily lack true population covariance and shared trial-to-trial variability, limiting their interpretability as genuine population dynamics. Similarly, conclusions about trajectory-based versus attractor-based computation are drawn almost exclusively from model analyses and then generalized to the biological system.

Overall, while the modeling framework is appropriate as a hypothesis-generating tool, the manuscript repeatedly crosses the line from proposing plausible mechanisms to asserting explanatory or even causal equivalence between the model and the brain. This undermines the otherwise strong contributions of the work.

Below are several specific points that warrant further clarification or revision:

(1) Tone of model-data correspondence

Numerous statements describe the RNN as "closely mimicking," "recapitulating," or being "nearly identical" to claustral neural dynamics, sometimes extending to claims about causal relationships between neural activity and behavior. Given that neural data were not used to train the model, and that only a small subset of trained networks showed the reported dynamics, these statements should be substantially softened throughout the manuscript. The RNN should be framed as providing one possible computational realization consistent with existing data, not as a close instantiation of the biological circuit

(2) Non-uniqueness of RNN solutions

The fact that only a small fraction of trained networks exhibited "claustrum-like" clusters deserves deeper discussion. This observation raises the possibility that the identified solution is fragile or highly specific rather than canonical. The authors should explicitly discuss the non-uniqueness of internal solutions in behavior-trained RNNs, including the range of alternative network dynamics that can reproduce the same behavior. In particular, it should be clarified why the specific network exhibiting "claustrum-like" clusters is informative about claustral computation, rather than representing one arbitrary solution among many.

(3) GPFA trajectory comparisons

The qualitative similarity between RNN trajectories and GPFA-derived trajectories from sparse in vivo data is interesting but insufficient to support claims of robustness or population-level structure. Statements suggesting that these patterns are unlikely to arise from noise or random fluctuations are not justified, given the single-trial, pseudo-population nature of the data. Either additional quantitative controls should be added, or the interpretation should be substantially tempered.

(4) Scope of functional claims

The discussion connecting the findings to broad theories of claustral function, global workspace, or consciousness extends well beyond the data presented. These speculative links should be clearly labeled as such and significantly reduced in strength and prominence.

(5) Comment on Conceptual Interpretation of the Behavioral Paradigm:

The manuscript repeatedly describes the delayed escape task as an "inference-based behavioral paradigm" and states that animals "infer that a value-neutral alternative space is likely to be safer" when the CS is presented in a novel environment. While I appreciate that the US-CS association was established in a different context and that the CS is then presented in a new environment, I am not convinced that the current behavioral evidence uniquely supports an inference interpretation.

First, it is not clear that this task is widely recognized in the literature as a canonical inference task, in the sense of, for example, sensory preconditioning, transitive inference, or model-based inference paradigms. Rather, the observed effect-that CS animals escape faster to a neutral compartment than neutral-CS controls-can be parsimoniously interpreted in terms of generalized threat value, heightened fear/anxiety, or a bias toward avoidance/escape under elevated threat, without requiring an explicit inferential step about the specific safety of the alternative compartment. The fact that no prior training is needed is compatible with flexible generalization, but does not by itself demonstrate inference in a more formal computational sense.

Second, the inference claim becomes central to the manuscript's conceptual framing (e.g., the idea that rsCla supports "inference-based escape"), yet the behavioral analyses presented here and in the cited prior work do not clearly rule out simpler accounts. Clarifying this distinction would help avoid overstating both the inferential nature of the behavior and the specific role of rsCla and the RNN's "claustrum-like" cluster in supporting inference per se, as opposed to more general integration of threat-related signals with an opportunity for escape.

Overall Assessment:

This manuscript presents an interesting and potentially valuable modeling-based framework for thinking about temporal integration in the claustrum, supported by solid slice physiology. However, in its current form, it overstates the degree to which the proposed RNN dynamics reflect actual claustral neural mechanisms. With substantial revision - especially a more cautious interpretation of model-data similarity and a clearer articulation of modeling limitations - the study could make a meaningful contribution as a hypothesis-generating work rather than a definitive mechanistic account.

Reviewer #2 (Public review):

This manuscript reports the behavior of a computational model of rat claustral neurons during the performance of a behavioral task known as the delayed escape task (in this reviewer's understanding, this behavioral task was created and implemented by this group only). These authors have argued in a prior manuscript (Han et al.) that a group of neurons located "rostral to striatum" is part of the claustrum. The group names the region the "rostral to striatum claustrum." Additionally, in the Han et al. paper, the authors argue that these cells are responsible for maintaining a signal that lasts through the delay period.

The main findings of the current paper are:

(1) The authors have built a model network that was trained to show firing similar to what was reported for rats in their prior paper.

(2) The authors' analysis of model behavior is used to suggest that the model network recapitulates biological activity, including the existence of a cluster of cells mainly responsible for the delay period firing.

(3) The authors offer evidence from patch clamp recordings for excitatory interconnections among claustral neurons that are an essential feature of the model network.

A major value of the computational network is that "trials" of the network can be performed. In experiments on animals, only single trials can be used.

Concerns:

(1) This paper is based on behavioral results and neural recordings from their prior paper (Han et al.), but data, e.g., in Figure 1, are not clearly identified as new or as coming from that source. Figure 1A, for example, appears to be taken directly from Han et al. No methods are given in this manuscript for the behavioral testing or the in vivo electrophysiology.

(2) Many other details are unclear. Examples include model training, the weight matrices and how these changed with training (p. 13), equations 2 and 3 (p. 13), the sources for the constants in the equations (p. 14), the methods (anesthesia, stereotaxic coordinates, injection specifics and details for "sparse expression") for the ChrimsonR injections.

(3) The explorations of model behavior are a catalog of everything tried rather than an organized demonstration of what the model can and cannot do. The figures could be reduced in number to emphasize the key comparisons of the different clusters and the model's behavior under different conditions, intended to "test" the model.

(4) On page 6, the E-E connectivity is argued from Shelton et al. (2025) and against Kim et al. (2016), but ignores Orman (2015), which, to this reviewer's knowledge, was the first to demonstrate such connectivity, including the long-duration events and impact of planes of section.

(5) Whereas the authors are entitled to their own opinion of prior work (references 3-8), it is inappropriate to misrepresent prior work as only demonstrating a "limited function" of claustum. Additional papers by Mathur's group and Citri's group are ignored.

In summary, the authors have made a computational model that recapitulates the firing of a subset of potentially claustral neurons during a particular behavioral task (delayed escape is certainly not the only behavior that involves claustrum - see e.g., attention, salience, sleep). If the conclusion is that excitatory claustral cells must be connected to other excitatory claustral cells, such a conclusion is not new, and the electrophysiological E-E metrics are not well quantified (e.g., connectivity frequency, strength of connection). If the model is intended to predict how the claustrum might accomplish any other task, there is insufficient detail to evaluate the model beyond the evidence that the model creates a subset of cells that can sustain firing during the delay period in the delayed escape task.

All relevant work must be appropriately cited throughout the manuscript.

Reviewer #2 (Public review):

Summary:

Missed diagnosis of myocardial ischemia (MI) is more common in women, and treatment is typically less aggressive. This diagnosis stems from the fact that women's ECGs commonly exhibit 12 lead ECG biomarkers that are less likely to fall within the traditional diagnostic criteria. Namely, women have shorter QRS durations and lower ST junction and T wave amplitudes, but longer QT intervals, than men. To study the impact, this study aims to quantify sex differences in heart-torso anatomy and ECG biomarkers, as well as their relative associations, in both pre- and post-MI populations. A novel computational pipeline was constructed to generate torso-ventricular geometries from cardiac magnetic resonance imaging. The pipeline was used to build models for 425 post-myocardial infarction subjects and 1051 healthy controls from UK Biobank clinical images to generate the population.

This study has a strength in that it utilizes a large patient population from the UK Biobank (425 post-MI and 1051 healthy controls) to analyze sex-based differences. The computational pipeline is state-of-the-art for constructing torso-ventricular geometries from cardiac MR and is clinically viable. It draws on novel machine learning techniques for segmentation, contour extraction, and shape modeling. This pipeline is publicly available and can help in the large-scale generation of anatomies for other studies. The study then deploys a linear regression model to relate the level of influence of various factors to ECG-based changes. This allows computation of various anatomical factors (torso volume, cavity volume, etc), and subsequent linear regression analysis on how these factors are altered before and after MI from the 12-lead ECG.

A major weakness is that a linear additive model may not adequately capture how anatomy and electrophysiology interact. Myocardial infarction dramatically alters both anatomy and electrophysiology in ways that are not easily separable and could be considered non-linear. As such, the electrophysiological factors in the model may still include factors that have an anatomical basis (i.e. the formation of scar) that were not accounted for during model generation. However, the technique remains useful for dissecting large factors beyond anatomy, as demonstrated in this study.

Reviewer #1 (Public review):

Summary:

The electrocardiogram (ECG) is routinely used to diagnose and assess cardiovascular risk. However, its interpretation can be complicated by sex-based and anatomical variations in heart and torso structure. To quantify these relationships, Dr. Smith and colleagues developed computational tools to automatically reconstruct 3D heart and torso anatomies from UK Biobank data. Their regression analysis identified key sex differences in anatomical parameters and their associations with ECG features, particularly post-myocardial infarction (MI). This work provides valuable quantitative insights into how sex and anatomy influence ECG metrics, potentially improving future ECG interpretation protocols by accounting for these factors.

Strengths:

• The study introduces an automated pipeline to reconstruct heart and torso anatomies from a large cohort (1,476 subjects, including healthy and post-MI individuals). • The 3-stage reconstruction achieved high accuracy (validated via Dice coefficient and error distances). • Extracted anatomical features enabled novel analyses of disease-dependent relationships between sex, anatomy, and ECG metrics. • Open-source code for the pipeline and analyses enhances reproducibility.

Weaknesses:

• The study attributes residual ECG differences to sex/MI status after controlling for anatomical variables. However, regression model errors could distort these estimates. A rigorous evaluation of potential deviations (e.g., variance inflation factors or alternative methods like ridge regression) would strengthen the conclusions.

Reviewer #1 (Public review):

Summary:

The electrocardiogram (ECG) is routinely used to diagnose and assess cardiovascular risk. However, its interpretation can be complicated by sex-based and anatomical variations in heart and torso structure. To quantify these relationships, Dr. Smith and colleagues developed computational tools to automatically reconstruct 3D heart and torso anatomies from UK Biobank data. Their regression analysis identified key sex differences in anatomical parameters and their associations with ECG features, particularly post-myocardial infarction (MI). This work provides valuable quantitative insights into how sex and anatomy influence ECG metrics, potentially improving future ECG interpretation protocols by accounting for these factors.

Strengths:

(1) The study introduces an automated pipeline to reconstruct heart and torso anatomies from a large cohort (1,476 subjects, including healthy and post-MI individuals).

(2) The 3-stage reconstruction achieved high accuracy (validated via Dice coefficient and error distances).

(3) Extracted anatomical features enabled novel analyses of disease-dependent relationships between sex, anatomy, and ECG metrics.

(4) Open-source code for the pipeline and analyses enhances reproducibility.

Weaknesses:

(1) The linear regression approach, while useful, may not fully address collinearity among parameters (e.g., cardiac size, torso volume, heart position). Although left ventricular mass or cavity volume was selected to mitigate collinearity, other parameters (e.g., heart center coordinates) could still introduce bias.

(2) The study attributes residual ECG differences to sex/MI status after controlling for anatomical variables. However, regression model errors could distort these estimates. A rigorous evaluation of potential deviations (e.g., variance inflation factors or alternative methods like ridge regression) would strengthen the conclusions.

(3) The manuscript's highly quantitative presentation may hinder readability. Simplifying technical descriptions and improving figure clarity (e.g., separating superimposed bar plots in Figures 2-4) would aid comprehension.

(4) Given established sex differences in QTc intervals, applying the same analytical framework to explore QTc's dependence on sex and anatomy could have provided additional clinically relevant insights.

Reviewer #2 (Public review):

Summary:

Missed diagnosis of myocardial ischemia (MI) is more common in women, and treatment is typically less aggressive. This diagnosis stems from the fact that women's ECGs commonly exhibit 12 lead ECG biomarkers that are less likely to fall within the traditional diagnostic criteria. Namely, women have shorter QRS durations and lower ST junction and T wave amplitudes, but longer QT intervals, than men. To study the impact, this study aims to quantify sex differences in heart-torso anatomy and ECG biomarkers, as well as their relative associations, in both pre- and post-MI populations. A novel computational pipeline was constructed to generate torso-ventricular geometries from cardiac magnetic resonance imaging. The pipeline was used to build models for 425 post-myocardial infarction subjects and 1051 healthy controls from UK Biobank clinical images to generate the population.

Strengths:

This study has a strength in that it utilizes a large patient population from the UK Biobank (425 post-MI and 1051 healthy controls) to analyze sex-based differences. The computational pipeline is state-of-the-art for constructing torso-ventricular geometries from cardiac MR and is clinically viable. It draws on novel machine learning techniques for segmentation, contour extraction, and shape modeling. This pipeline is publicly available and can help in the large-scale generation of anatomies for other studies. This allows computation of various anatomical factors (torso volume, cavity volume, etc), and subsequent regression analysis on how these factors are altered before and after MI from the 12-lead ECG.

Weaknesses:

Major weaknesses stem from the fact that, while electrophysiological factors appear to play a role across many leads, both post-MI and healthy, the electrophysiological factors are not stated or discussed. The computational modeling pipeline is validated for reconstructing torso contours; however, potential registration errors stemming from ventricular-torso construction are not addressed within the context of anatomical factors, such as the tilt and rotation of the heart. This should be discussed as the paper's claims are based on these results. Further analysis and explanation are needed to understand how these sex-specific results impact the ECG-based diagnosis of MI in men and women, as stated as the primary reason for the study at the beginning of the paper. This would provide a broader impact within the clinical community. Claims about demographics do not appear to be supported within the main manuscript but are provided in the supplements. Reformatting the paper's structure is required to efficiently and effectively present and support the findings and outcomes of this work.

Reviewer #1 (Public review):

Summary:

In this study, the authors trained rats on a "figure 8" go/no-go odor discrimination task. Six odor cues (3 rewarded and 3 non-rewarded) were presented in a fixed temporal order and arranged into two alternating sequences that partially overlap (Sequence #1: 5⁺-0^--1^--2⁺; Sequence #2: 3⁺-0^--1^--4⁺) --forming an abstract figure-8 structure of looping odor cues.

This task is particularly well-suited for probing representations of hidden states, defined here as the animal's position within the task structure beyond superficial sensory features. Although the task can be solved without explicit sequence tracking, it affords the opportunity to generalize across functionally equivalent trials (or "positions") in different sequences, allowing the authors to examine how OFC representations collapse across latent task structure.

Rats were first trained to criterion on the task and then underwent 15 days of self-administration of either intravenous cocaine (3 h/day) or sucrose. Following self-administration, electrodes were implanted in lateral OFC, and single-unit activity was recorded while rats performed the figure-8 task.

Across a series of complementary analyses, the authors report several notable findings. In control animals, lOFC neurons exhibit representational compression across corresponding positions in the two sequences. This compression is observed not only in trial/positions involving overlapping odor (e.g., Position 3 = odor 1 in sequence 1 vs sequence 2), but also in trials/positions involving distinct, sequence-specific odors (e.g., Position 4: odor 2 vs odor 4) --indicating generalization across functionally equivalent task states. Ensemble decoding confirms that sequence identity is weakly decodable at these positions, consistent with the idea that OFC representations collapse incidental differences in sensory information into a common latent or hidden state representation. In contrast, cocaine-experienced rats show persistently stronger differentiation between sequences, including at overlapping odor positions.

Strengths:

- Elegant behavioral design that affords the detection of hidden-state representations.<br /> - Sophisticated and complementary analytical approaches (single-unit activity, population decoding, and tensor component analysis).

Weaknesses:

-The number of subjects is small --can't fully rule out idiosyncratic, animal-specific effects.

Comments on revisions:

The authors have thoroughly addressed all of my previous comments. Congratulations on an excellent paper!

Reviewer #2 (Public review):

In the current study, the authors use an odor-guided sequence learning task described as a "figure 8" task to probe neuronal differences in latent state encoding within the orbitofrontal cortex after cocaine (n = 3) vs sucrose (n = 3) self-administration. The task uses six unique odors which are divided into two sequences that run in series. For both sequences, the 2nd and 3rd odors are the same and predict reward is not available at the reward port. The 1st and 4th odors are unique, and are followed by reward. Animals are well-trained before undergoing electrode implant and catheterization, and then retrained for two weeks prior to recording. The hypothesis under test is that cocaine-experienced animals will be less able to use the latent task structure to perform the task, and instead encode information about each unique sequence that is largely irrelevant. Behaviorally, both cocaine and sucrose-experienced rats show high levels of accuracy on task, with some group differences noted. When comparing reaction times and poke latencies between sequences, more variability was observed in the cocaine-treated group, implying animals treated these sequences somewhat differently. Analyses done at the single unit and ensemble level suggests that cocaine self-administration had increased the encoding of sequence-specific information, but decreased generalization across sequences. For example, the ability to decode odor position and sequence from neuronal firing in cocaine-treated animals was greater than controls. This pattern resembles that observed within the OFC of animals that had fewer training sessions. The authors then conducted tensor component analysis (TCA) to enable a more "hypothesis agnostic" evaluation of their data.

Overall, the paper is well written and the authors do a good job of explaining quite complicated analyses so that the reader can follow their reasoning. The findings are important, and the results are compelling. The introduction and discussion contextualize the experiments in the context of the literature, and explain the novelty and significance of the current findings. Specifically, the observation that cocaine self-administration impairs generalization across task sequences at the single unit level builds on previous observations of aberrant neuronal activity within the OFC in animals with a history of cocaine self-administration. These new data point to a neurophysiological mechanism that could explain why drug-seeking is so context dependent, and hard to ameliorate with therapeutic strategies that take place within a clinical setting.

The authors clearly acknowledge the major limitations of this work, namely that the sample size is restricted due to the technical challenges of performing in vivo electrophysiology recordings combined with self-administration, and that animals of only one sex were used. Importantly, the data from all rats within each group was remarkably homogeneous, increasing confidence in the conclusions drawn.

Reviewer #1 (Public review):

The manuscript "Heterozygote advantage cannot explain MHC diversity, but MHC diversity can explain heterozygote advantage" explores two topics. First, it is claimed that the recently published by Mattias Siljestam and Claus Rueffler conclusion (in the following referred to as [SR] for brevity) that heterozygote advantage explains MHC diversity does not withstand an even very slight change in ecological parameters. Second, a modified model that allows an expansion of MHC gene family shows that homozygotes outperform heterozygotes. This is an important topic and could be of potential interest to the readership of eLife if the conclusions are valid and non-trivial.

The resubmitted manuscript addresses several questions from my previous review. In particular, there is a more detailed description of how the code of Siljestam and Rueffler ([SR]) was used for the simulations and the calculation of the factor 2.7 x 10^43 that is the key to the alleged breakdown of the numerical reasoning presented by in [SR].

Yet I think that important aspects of my critique of the first statement of the manuscript about the flaws of [SR] model remain unanswered. I guess the discussion becomes rather general about the universality and robustness of various types of models to parameter changes. My point is that none of the models is totally universal. The model in [SR] is not phenomenological as none of the parameters or functional forms were derived empirically. Instead, it is a proof of principle demonstration that inevitably grossly simplifies the actual immune response. The choice of constants and functions used in Eqs. (1-5) is dictated by the mathematical convenience and works in a limited range of parameter values. It is shown in [SR] that for 3 pathogens and reasonable "virulence " \nu, the alleles branch. These conclusions are supported by the analytically derived Adaptive Dynamics branching criteria (7), which, contrary to the statement is the cover letter (" It is clear from Fig. 4 of Siljestam and Rueffler that the branching condition is far from sufficient for high MHC diversity.") is perfectly confirmed by the simulation data shown in Fig. 4.

The mathematical simplicity of the [SR] model generates various artifacts, such as the mentioned by the Author reduction of the "condition" by an enormous factor 2.7 x 10^43 and the resulting decrease in the "survival" induced by the addition of a new pathogen. This occurs at the very large value of \nu=20, whose effect is enormous due to the Gaussian form of (1), which, once again, was chosen for the mathematical convenience. In reality, a new pathogen cannot reduce the "survival" by such a factor as it would wipe out any resident population. So to compensate for such an artifact, the additional factor c_max was introduced to buffer such an excess. There is no reason to fix c_max once for an arbitrary number of pathogens, because varying c_max basically reflects the observation that a well-adapted individual must have a reasonable survival probability. At the same time, there are many ways in which the numerical simulation may break down when the survival rates become of the order of 10^(-43) instead of one, so it comes to no surprise that the diversification, predicted by the adaptive dynamics, does not readily occur in the scenario with an addition or removal of the 8th pathogen with a very high virulence \nu=20.

I have doubts that the reported breakdown of the [SR] model with fixed c_max remains observable with less extreme values of m and \nu (say, for \nu=7 and m=3 plus or minus 1 used in Fig. 3 in the manuscript).

So I still find the claim that " the phenomenon that leads to high diversity in the simulations of Siljestam and Rueffler depends on finely tuned parameter values" is not well substantiated.

Reviewer #2 (Public review):

Summary:

This study addresses the population genetic underpinnings of the extraordinary diversity of genes in the MHC, which is widespread among jawed vertebrates. This topic has been widely discussed and studied, and several hypotheses have been suggested to explain this diversity. One of them is based on the idea that heterozygote genotypes have an advantage over homozygotes. While this hypothesis lost early on support, a reason study claimed that there is good support for this idea. The current study highlights an important aspect that allows us to see results presented in the earlier published paper in a different light, changing strongly the conclusions of the earlier study, i.e., there is no support for a heterozygote advantage. This is a very important contribution to the field. Furthermore, this new study presents an alternative hypothesis to explain the maintenance of MHC diversity, which is based on the idea that gene duplications can create diversity without heterozygosity being important. This is an interesting idea, but not entirely new.

Strength:

(1) A careful re-evaluation of a published model, questioning a major assumption made by a previous study.

(2) A convincing reanalysis of a model that, in the light of the re-analysis-loses all support.

(3) A convincing suggestion for an alternative hypothesis.

Weakness:

(1) The title of the study is catchy, but it is explained only in the very end of the paper.

Reviewer #1 (Public review):

Summary:

This study examined whether infraslow fluctuations in noradrenaline and in heart rate are coupled and how they are affected by sleep transitions. The authors used the fluorescent NA biosensor GRAB-NE2m in the medial prefrontal cortex of mice to record extracellular NA while also recording EEG and EMG during sleep-wake episodes. They also analyzed previously published human data to reproduce relationships they found between sigma power and RR intervals in mice.

Strengths:

This is an impressive study with significant strengths, as it involves a rich set of data that includes not only observations of associations between heart rate and noradrenergic dynamics but also optogenetic manipulation of the locus coeruleus. Human data is presented to show parallels in the association between sigma power during sleep and phasic heart-rate bursts.

Weaknesses:

(1) Language could be clearer and more precise. As detailed below, in both the introduction and the discussion, the way the hypotheses and study objectives are described could use some revision to be more precise and accurate.

1A) In the introduction on p. 4: The overarching question is framed as "could the peripheral autonomous systems be a read-out of the central LC-NE system and thus be a biomarker of memory consolidation and LC dysfunction?" This gives the impression that the LC function would be the main influence on peripheral autonomous systems. There are, of course, many influences on peripheral autonomous systems, so it would be advisable for the authors to be more specific here about what signal(s) in particular would be predicted to be sensitive markers of LC function.

1B) In the discussion on p. 12: "In this study, we leveraged real-time measurements of mPFC NE levels and HR measurements from EMG recordings in mice to investigate the causal link between the two variables with high temporal resolution in freely moving sleeping mice, with similar inspection in humans." To test the causal link between mPFC NA levels and HR measures, the study would manipulate NA levels just in the mPFC and not elsewhere in the brain. However, in this study, the manipulation occurred in the LC, and so there would be broad cortical changes in NA levels. Thus, it could be that LC activity causes HR changes via a non-PFC pathway.

(2) Comparisons with the control condition need further development.

2A) While the authors did include a key YFP control condition, in the main text no direct statistical comparison between the closed-loop optogenetic stimulation (ChR2) condition and the YFP control condition was reported. (It was reported in Supplementary Figure 2c-d.) Instead, in the main text, the authors only reported that the effects of stimulation were significant in the closed-loop condition and not in the control. However, that is not the same as demonstrating that the two conditions significantly differed from each other, and it is the direct test that is important for the conclusions, so it seems important to include this result in the main presentation.

2B) In addition, the authors should address the issue that the pre-stimulation NE was consistently significantly lower in the YFP condition than in the ChR2 condition (see Supplementary Figure 2c), which is a potential confound.

2C) Direct comparison of the strengths of correlations shown in Figure 2h vs. Supplementary Figure 2f should be included. Currently, we see relatively weak correlations in both ChR2 and YFP conditions, and it is not clear if the relationships differ in the control. It seems they are still present in the control condition but weaker, which would contradict the apparently broad claim on p. 7 that "No such effects were present in the control condition" (it is not entirely clear whether this claim refers to all effects discussed in the figure or just a subset - this language should be clarified).

2D) Did the YFP controls vs. ChR2 animals show any differences in the number of NA states that triggered stimulation in the closed-loop system? With ChR2 animals, stimulation changes NA, which could change future triggering. In YFP animals, nothing changes NA (other than natural fluctuations), so the dynamics of stimulation timing could diverge between groups in a way that complicates interpretation. Specifically, if ChR2 stimulation raises NA and prevents future threshold crossings, ChR2 animals may end up receiving fewer subsequent stimulations than YFP animals (or a different temporal clustering). If the number or pattern of stimulation differed in two groups, it would be important to have a yoked control where matched animals get the same stimulation pattern but not triggered by their own NA.

(3) Some more discussion/explanation of the rationale for the closed-loop approach and how it influences how we should interpret the results could be useful. For instance, currently, it is not clear whether LC stimulation needs to be timed after an NA dip to yield the effects seen.

(4) The section on heart rate decelerations is hard to follow. In particular, I was not sure how to interpret Figure 3f-j. For Figure 3f, what does the middle line represent? The laser onset or the max RR value after laser onset? What is the baseline that is used to correct the values to obtain amplitudes? If it is the whole period before the maximal RR value or the laser onset, wouldn't baseline values differ significantly across conditions and so potentially account for differences seen between conditions in the reported HR decelerations? Larger HR decelerations may be seen in conditions with higher HR simply as a regression to the mean phenomenon.

(5) The findings regarding LC suppression could be further clarified.

5A) Page 8: "observed a response in NE decline" - please be more precise. Did NE decline more or less?

5B) It would be helpful to also show the correlation between NE and RR in the control (YFP) condition and whether there were any differences between YFP and Arch conditions (Figure 4e).

5C) This sentence took me multiple readings to understand - it would be helpful to rewrite to make it clearer: "indicating that, while HR generally did not respond strongly to LC suppression, the variability in RR responses was dependent on NE changes to the suppression (Figure 4e)."

5D) The two colors in Figure 4 are similar and hard to distinguish.

5E) The correlations shown in Figure 4j seem to be driven by just two of the cases. Are the effects significant when outliers are removed?

5D) Page 10: Were there any differences in memory performance between the Arch and YFP conditions?

5E) Page 10: "We found a correlation between RR responses to LC suppression and sigma power, suggesting that a stronger HR reduction response is linked to higher spindle power." It should be noted in the text that the correlation was not specific to sigma (it was also seen for theta and beta, Figure 4i).

(6) It is not clear which of the sigma power and RR interval findings do/do not exactly line up between the mice and humans. It could be helpful to have a table comparing them. For instance, was the finding in humans that pre-HRB sigma power was positively associated with slowing in heart rate after the HRB also seen in mice? Was there evidence in mice (as seen in the human sample) that sleep-dependent memory improvement was associated with pre-HRB sigma power?

(7) Page 18: It is not clear if the sex of mice was balanced across controls and optogenetics groups.

Reviewer #2 (Public review):

Summary:

The major part of this study reproduces previously published findings in both mice and humans and provides incremental analyses on these findings. In essence, the work reaffirms the presence of coordinated infraslow fluctuations in sigma power and heart rate during NREM sleep. It further confirms previous findings that coordination depends on noradrenaline-releasing neurons in the locus coeruleus. Also supporting previously published work in mice and humans, the authors describe a link between the strength of these infraslow fluctuations and memory consolidation in mice and humans.

Strengths:

The authors successfully replicate key previously reported phenomena across both mice and humans. Confirmatory studies and demonstrations of reproducibility are essential for progress in neuroscience. To maximize their value, such studies should clearly acknowledge their confirmatory nature and carefully situate what, in their view, are novel results, going beyond existing literature.

Weaknesses:

The authors' interpretation of their data needs to be revised. Many of their claims regarding the mechanistic basis of their findings and the predictive value of their correlative datasets are not supported by the available evidence.

In the present manuscript, several citations of literature on the work they reproduce lack precision or completeness, which reduces transparency and obscures how the reported findings relate to previously established results.

Reviewer #1 (Public review):

Summary:

Liao et al. performed a large-scale integrative analysis to explore the function of two cancer genes (BRCA1 and BRCA2) in lung cancer, which is one of the cancers with an extremely high mortality rate. The detailed genetic analysis demonstrated new roles of BRCA1/2 in causing the tumor microenvironment in lung cancer. In particular, the discovery of different mechanisms of BRCA1 and BRCA2 provides an essential foundation for developing drugs that target BRCA1 or BRCA2 in lung cancer therapy.

Strengths:

(1) This study leveraged large-scale genomic and transcriptomic datasets to investigate the prognostic implications of BRCA1/2 mutations in LUAD patients (~2,000 samples). The datasets range from genomics to single-cell RNA-seq to scTCR-seq.

(2) In particular, the scTCR-seq offers a powerful approach for understanding T cell diversity, clonal expansion, and antigen-specific immune responses. Leveraging these data, this study found that BRCA1 mutations were associated with CD8+ Trm expansion, whereas BRCA2 mutations were linked to tumor CD4+ Trm expansion and peripheral T/NK cell cytotoxicity.

(3) This study also performed a comprehensive analysis of genomic variation, gene expression, and clinical data from the TCGA program, which provides an independent validation of the findings from LUAD patients newly collected in this study.

(4) This study provides an exemplary integration analysis using both computational biology and wet bench experiments. The experimental testing in the A549 cell line further supports the robustness of the computational analysis.

(5) The findings of this study offer a comprehensive view of the molecular mechanisms underlying BRCA1 and BRCA2 mutations in LUAD. BRCA1 and BRCA2 are two well-known cancer-related genes in multiple cancers. However, their role in shaping the tumor microenvironment, particularly in lung cancer, is largely unknown.

(6) By focusing on PD-L1-negative LUAD patients, this study demonstrated the molecular mechanisms underlying resistance to immune therapy. These new insights highlight new opportunities for personalized therapeutic strategies to BRCA-driven tumors. For example, they found histone deacetylase (HDAC) inhibitors consistently downregulated 4-R genes in A549 cells.

(7) The deposition of raw single-cell sequencing (including scRNA-seq and scTCR-seq) data will provide an essential data resource for further discovery in this field.

Weaknesses:

(1) The finding of histone deacetylase (HDAC) inhibitors suggests the potential roles of epigenetic regulation in lung cancer. It would be interesting to explore epigenetic changes in LUAD patients in the future.

(2) For some methods, more detailed information is needed.

(3) There are grammar issues in the text that need to be fixed.

(2) Some text in the figures is not labeled well.

Reviewer #2 (Public review):

Summary:

This study investigates the impact of BRCA1/2 mutations on immunotherapy in lung adenocarcinoma using multi-omics approaches. The work highlights distinct roles of BRCA1 and BRCA2 mutations in shaping immune-related processes, and is logically structured with clearly presented analyses. However, the conclusions rely primarily on descriptive computational analyses and would benefit from additional immunological validation.

Strengths:

By integrating public datasets with in-house data, this study examines the impact of BRCA1/2 mutations on immunotherapy in lung adenocarcinoma from multiple perspectives using multi-omics approaches. The analyses are diverse in scope, with a clear overall logic and a well-organized structure.

Weaknesses:

The study is largely descriptive and would benefit from additional immunological experiments or validation using in vivo models. The fact that the BRCA1 and BRCA2 samples were each derived from a single patient also limits the robustness of the conclusions.

Reviewer #1 (Public review):

Summary:

Small molecule therapeutics for snakebite have received a lot of attention for their potential to close the gap between bite and treatment, where antivenom is not immediately available.

Strengths:

There has been a lot of focus on Africa, Asia, and India, but very little work related to neotropical regions. The authors seek to begin filling this gap in the preclinical literature. The authors use well-developed methods for preclinical assessment.

Weaknesses:

A clearer and more focused discussion of the limitations of the overall present work would be desirable (e.g. protection vs. rescue, why marimastat over prinomastat for in vivo assays when both have been through clinical trials for other indications; real-world feasibility of nafamostat, which has a half-life of 1-2 minutes compared to camostat, which has a half-life of hours). All of this could be be improved in a revision.

Reviewer #2 (Public review):

Summary:

The authors set out to test whether a defined set of small molecules can lessen damaging effects caused by venoms from several Bothrops species, and whether these effects are consistent enough to suggest a broadly applicable approach. They present a cross-venom dataset spanning in-vitro activity readouts and blood-based functional outcomes, and include a chicken embryo model to explore whether venom inhibition can translate into improved survival. The central message is that certain small molecules can reduce specific venom-driven effects across multiple samples, providing a comparative resource for the field and a basis for prioritizing future validation.

Strengths:

The main value of this work is the breadth and structure of the dataset, which places multiple venoms and multiple readouts into a single, comparable framework that should be useful for readers evaluating patterns across samples. The experimental flow is generally coherent, moving from activity measurements to functional outcomes and then to an in-vivo test, which helps the reader understand how the authors link mechanism-oriented assays to more integrated endpoints. The manuscript also provides practical information for the community by highlighting which readouts appear most consistently affected across venoms, which can help guide hypothesis generation and study design in follow-up work.

Weaknesses:

Several aspects of the study design and framing reduce the confidence with which readers can translate the findings beyond the specific experimental context presented. The evidence base is strongest in controlled in-vitro settings, while the bridge to real-world effectiveness remains limited, particularly for understanding performance under conditions that better reflect delayed treatment and systemic exposure. As a result, the manuscript is best interpreted as a well-organized comparative screening study with promising signals, rather than a definitive demonstration of a broadly effective, deployable intervention.

Reviewer #3 (Public review):

In this work, the authors wanted to evaluate repurposed small molecule inhibitors for the treatment of envenomation by snakes of the Bothrops genus; one of the most medically relevant in the Americas. I believe the objectives of the research were clearly achieved, and compelling evidence for the ability of these molecules to neutralize enzymatic and toxic activities of metalloproteinases and phospholipases in all the tested venoms is provided. Furthermore, the work highlights the limited efficacy of the tested serineprotease inhibitor, suggesting a need for drug discovery campaigns to address toxicity caused by this protein family. The methods are well designed and performed, and the use of both in vitro and in vivo methodologies makes this a thorough and robust work.

These results are extremely relevant, since they take us one step further to a potential orally administered snakebite treatment. The existence of such a treatment could improve the outcomes for thousands of snakebite victims worldwide. I have a few comments and questions that I hope will be useful to the authors:

During the introduction, the authors mention that small-molecule inhibitors can neutralize the localized tissue damage via cytotoxicity of some venoms, and cite PLA2s, SVMPs and/or cytotoxic 3FTxs as the main causing agents of this pathology. I am not aware of any direct effect described by small molecule inhibitors on cytotoxic 3FTxs alone. Has this been observed at all? Or is it more likely that the small molecule inhibitors act on the enzymatic toxins only, preventing synergistic effects with 3FTxs?

I think it would be relevant to address the effects of non-enzymatic PLA2s, such as myotoxin II, which have been described in detail within Bothrops venoms. I believe there is some evidence of Varespladib also having a neutralizing effect on the myotoxicity caused by these non-enzymatic PLA2s. I suggest adding a comment about the contribution of these toxins in the discussion or in the section where PLA2 activity of the venoms is compared. In my opinion, right now it seems like these were overlooked.

Regarding Marimastat and the other MP inhibitors, are there any studies showing that they don't have an effect on endogenous MPs? I understand they have been approved for human use before, but is there any indication that they would not have an effect at the doses that would be required to treat envenomation?

Regarding the quenched fluorescence substrate used for enzymatic activity. Is there a possibility that some of the SVMPs would not act on this substrate, and therefore their activity or neutralization is not observed? Would it be relevant to test other substrates, such as gelatin, collagen, or even specific clotting factors?

Finally, could the authors comment or provide some bibliography regarding the translatability of the chicken embryo model in the context of envenomation?

Reviewer #1 (Public review):

Summary:

This study addresses a fundamental question in cognitive neuroscience regarding how the brain transitions from a reactive state of following external instructions to a proactive state of self-directed agency. The authors utilize an ambitious multimodal design by combining the spatial precision of fMRI with the temporal resolution of EEG across two independent datasets from the University of Florida and UC Davis. By applying multivariate pattern analysis, the work demonstrates that while both instructed and willed attention engage the Dorsal Attention Network, willed choices uniquely recruit a frontoparietal decision network including the dACC and anterior insula. Furthermore, the study shows that pre cue alpha oscillations can predict subsequent spontaneous choices. This provides a neural link between pre-existing brain states and intentional action, representing a significant technical effort to characterize the neural scaffolding of internal goal generation.

Strengths:

The primary strengths of this work include the integration of fMRI and EEG which allows the authors to bridge the gap between slow metabolic signals and fast oscillatory brain states. The use of two independent cohorts is a commendable effort to ensure the reproducibility of the willed attention effect, which is often a concern in small sample neuroimaging studies. Additionally, the move beyond univariate activation toward information based mapping demonstrates that the identified networks actually contain specific information about the direction of attention.

Weaknesses:

However, several critical weaknesses must be addressed to support the fundamental claims made in the manuscript. There are significant behavioral differences in performance between the two sites, specifically regarding the UC Davis cohort exhibiting slower reaction times and lower accuracy compared to the UF group. These discrepancies suggest potential differences in subject populations or experimental environments that are not currently accounted for in the neural models. The fMRI analysis lacks temporal precision because the use of beta series regression collapses the complex BOLD response into a single estimate per trial. This loss of temporal information obscures the evolution of the decision process and makes it difficult to distinguish whether the identified patterns represent a truly spontaneous choice or a slow building pre planned strategy.

Furthermore, the EEG decoding approach utilized the entire topography of electrodes rather than a biologically motivated posterior region of interest. Given that alpha mediated spatial attention is traditionally localized to parieto occipital sensors, using the full electrode set risks the inclusion of non neural artifacts such as micro saccades or muscle activity which can contaminate multivariate classifiers. The introduction of the neural efficiency metric also requires further validation as the current ratio is mathematically sensitive to small denominators in the BOLD contrast.

Crucially, the manuscript does not address the physiological implications of recruiting additional frontoparietal networks when behavioral performance remains identical across conditions. The activation of the anterior insula and dACC is frequently associated with increased autonomic arousal and effort. If the willed condition requires more extensive neural scaffolding to reach the same behavioral output as the instructed condition, it raises the question of whether this internal decision process is accompanied by changes in arousal levels. The authors should consider whether the lack of a behavioral tax is due to a compensatory increase in arousal, which could be reflected in the EEG data or pupil diameter if recorded, and potentially also in the amplitude of BOLD activity, which is being masked by the neural efficiency metric. Without an account of how the brain balances this increased computational demand without impacting behavioral performance, the functional significance of the willed attention network remains partially obscured.

Reviewer #2 (Public review):

Summary:

This manuscript combines fMRI and EEG investigations performed at two research sites to examine 'willed' or volitional visuospatial attention, as contrasted with more standard cued (or 'instructed') visuospatial attention. The primary findings are: 1) willed attention (vs. instructed attention) drives additional cortical circuitry across a broad fronto-parietal network; 2) the direction of willed attention, but not instructed attention can be decoded from the pre-cue EEG data and from MVPA analysis of the trial-level fMRI data; and 3) the subjects with high EEG decoding also exhibited high neural efficiency (i.e., high decoding with low BOLD signal change) in the fMRI data. The methods and data analysis are generally sound, and these results appear solid. On the negative side, it is not made clear how the present findings extend our understanding beyond prior published work from one of the senior authors. There are also three significant concerns regarding interpretation of the findings. One has to do with the causal interpretation of the pre-cue alpha EEG signal determining the direction of willed attention. The second concern is the degree to which the present research paradigm adequately examines 'willed attention.' The third is that the MVPA analysis is not sufficiently described, and Permutation testing needs to be done to validate these findings. Otherwise, this manuscript appears methodologically sound, but questions about interpretation may mute the potential impact.

Strengths:

The focus on willed attention attempts to move beyond some of the many limitations of standard laboratory investigations of attention.

The shared paradigm across two modalities and two research sites demonstrates solid reproducibility, even though a few minor differences are observed across sites.

Weaknesses:

(1) There are concerns about this experimental paradigm carrying the banner of Willed Attention, because the application of 'Will' appears quite modest. Yes, extra brain activity is exhibited for this condition vs. its control, but do the cognitive processes isolated adequately stand in for 'Willed Attention?" Willed attention, as operationally defined here, appears to involve a simple decision process prior to the shifting of spatial attention. The cue is internally generated, but after that the rest of the attentional processes appear identical to standard externally cued visuospatial attention experiments. This self-generated cue process likely involves some sort of memory/history of the recently selected cues and then some random-ish selection between A and B. This appears very similar to asking the subject to guess whether a fair coin flip will be heads or tails on each trial. A mental 'coin flip' feels like a very weak version of 'will.' As a potential remedy, it would be helpful to discuss what other phenomena might fall within 'willed attention' and what some future studies might choose to focus on, along with some potential pitfalls (e.g., the reasons why the current study avoided more robust exemplars of will).

(2) The manuscript is lacking a description of the decision processes used during the willed attention paradigm and is lacking evidence as to WHEN subjects made their willed decision. Both of these points are of major concern:

(a) The authors state: "For willed attention, participants were explicitly told to avoid relying on any stereotypical strategies of generating decisions, such as always attending the same/opposite side they attended during the previous trial, as well as to avoid randomizing or equalizing their decisions to choose left or right across trials; prior studies found that decisions to explicitly randomize decisions might invoke additional working memory related processes (Spence & Frith, 1999)." Subjects were instructed NOT to apply a simple heuristic and NOT to randomize or try to equalize their decisions, but exactly HOW the subjects made their decisions is not at all clear. What options does that leave? How does this strategy avoid the working memory-related processes mentioned in the Spence & Frith, 1999 citation? The brain regions that comprise the network of interest (aka Frontoparietal Decision Network) are activated by a very broad range of visual cognitive tasks, including many working memory paradigms. The Anterior Insula and dACC nodes Salience Network often simply reflect task difficulty. Obviously, making a choice is more cognitively demanding than not making a choice. The present experiments do not distinguish functional roles between different regions of the Frontoparietal Decision Network. On the whole, the study does very little to isolate the cognitive processes or neural bases of willed attention beyond calling out the set of 'Usual Suspects' for visual cognition.

(b) The finding that pre-cue EEG signals predicted the postcue decision is intriguing. It could mean that the seemingly irrelevant and transient state of the brain causally and unconsciously biased the subject to one direction or the other. Alternatively, it could mean that the subjects utilized the pre-cue period to make their decision and hold it in case it was needed (i.e., that it was a choice trial). While 2-8 seconds ITI variability makes sense for fMRI decoding, it is a long time for a subject to idly wait, so they might fill that time preparing for the next trial. There appears to have been a substantial amount of individual difference in the pre-cue alpha decoding, which could reflect individual differences in cognitive strategy, specifically in the use of the pre-cue period to make their decision. More efficient decision makers might have pre-decided, which might account for the neural efficiency. The experiments lack any measurement of WHEN participants made their decision. For that reason, I would ask that the authors temper their claims about the significance of the alpha decoding and its possible causality.

(3) Did individual subjects exhibit a choice bias of location for the willed trials? If not, doesn't that raise concerns that subjects were trying to equalize their trials? If they do exhibit location biases, how does that impact the decoding? A simple decoder could learn to always just guess the biased direction for a subject and would perform > 50%. Consider the example in which an individual subject chooses 'Left' 55% of the time. A classifier that simply learns to choose 'Left' on every trial will be correct on 55% of trials. The training data would likely be sufficient to learn the direction of choice bias in each individual subject. So the classifiers could perform significantly above 50% without learning anything beyond the tendency of each subject. That is to say, 50% is not truly chance in this data set. It doesn't appear that Permutation testing has been performed to empirically determine chance for an individual's data. Permutation methods, scrambling the labels 1000 or 10000 times to establish a true baseline would be preferred over simply comparing to 50% and would address concerns about individual subject biases.

(4) The novel contributions of this work beyond the two prior Bengson et al papers from Dr. Mangun's lab appear quite modest. The discussion would be enhanced by specifically stating how the present work advances understanding beyond the prior Bengson studies.

Reviewer #3 (Public review):

Summary:

This manuscript analyzes two independent datasets collected at different sites. Using the same willed-attention paradigm (instructional vs. choice cues) and combining fMRI and EEG analyses, the authors investigate how attentional direction is selected when no external instruction is provided. Their main claims are that the dorsal attention network is engaged by both cue types, whereas the choice cue additionally involves a frontoparietal decision network. Moreover, left-versus-right attentional decisions can be decoded in this decision network only on choice trials, and multichannel pre-stimulus alpha patterns predict the subsequent attentional choice. Finally, individuals with more predictive alpha patterns show greater neural efficiency in the decision network, i.e., higher decoding with lower BOLD activation.

The question is worthwhile and the two-site design is a genuine strength. At the same time, several central inferences rely on decoding analyses for which the statistical testing and cross-validation structure are not described in enough detail to assess robustness. In addition, using a ratio-based neural-efficiency measure make the interpretation more fragile than it needs to be. With a focused revision that tightens inference around MVPA and clarifies a few methodological points, I think the paper could become substantially more convincing.

Strengths:

The work extends previous willed attention studies by attempting to link pre-stimulus alpha pattern predictability to post-cue frontoparietal representations, and by testing reproducibility across two datasets. The conceptual advance beyond previous studies, e.g., Bengson et al. (2015), however, depends on how solid the decoding-based evidence is and whether alternative explanations are convincingly excluded. At present, the strength of support is limited mainly by incomplete reporting and/or controls for MVPA significance testing, as well as potential inflation of decoding estimates if folds are not independent of run structure. Concerns about statistical assessment of decoding accuracy are well documented in the literature (Combrisson & Jerbi, 2015).

Weaknesses:

(1) The manuscript describes the decoding pipeline for both fMRI and EEG MVPA. However, it does not clearly specify how "significantly above chance" is determined for the fMRI ROI decoding, nor how multiple comparisons across ROIs are handled, even though p-values are reported. The same issue applies to the time-resolved EEG analysis across many time points. For each decoding analysis, please specify the inferential test (e.g., permutation test within participant, group-level test on subject accuracies, binomial test, etc.) and report effect sizes with confidence intervals (e.g., Combrisson & Jerbi, 2015). Further, for EEG decoding over time, it would be preferable to control family-wise error, e.g., cluster-based permutation, rather than thresholding pointwise p-values. A standard approach here is the nonparametric cluster framework (e.g., Maris & Oostenveld, 2007).

(2) The cross-validation approach used here is appreciated and appropriate in principle. However, random 10-fold splits across trials can inflate accuracy if training and test folds share run-specific noise, scanner drift, or autocorrelated structure. The manuscript should indicate whether folds were blocked by run or randomized across the entire session. In addition, please report the number of trials per condition after artifact rejection and after removing short ITIs for the long prestimulus epochs (−2500 ms to 0 ms) for each dataset in the section of EEG preprocessing. Similarly, please report how often participants chose left vs. right on choice trials, and whether balanced folds (or an equivalent balancing procedure) were used if needed.

(3) Moreover, ROI definition is not sufficiently specified and independence should be clarified. The ROIs are defined based on peaks from the choice-instructed univariate contrast (Table 2) and then used for MVPA. First, are these ROIs defined as spheres around peaks or using anatomical masks? What radius or voxel count was used? This needs to be explicit. Second, I am concerned about circularity risk. Although choice-vs-instructed selection is not identical to left-vs-right decoding, ROI selection from the same dataset can still bias descriptive estimates and encourages overinterpretation if not carefully justified (Kriegeskorte et al., 2009). At minimum, the authors should explain why their selection criterion is independent of the decoded contrast under the null, and ideally provide a robustness check using either anatomical ROIs or independently defined ROIs, e.g., from prior literature or an atlas.

(4) Using an index of neural efficiency is conceptually interesting. However, if the denominator, computed as the activation difference between choice and instructional conditions, is near zero or noisy, the ratio can become unstable. I would rather see a multivariate model that treats activation and decoding as separate dependent measures, or a latent-variable approach, than a single ratio.

Reviewer #1 (Public review):

Summary:

The authors investigated the relationship between physical activity (PA) and both structural (MRI) and cognitive brain health in the LIFE-Adult Study, with total baseline recruitment of 2576. Hippocampal volume, an MRI-derived BrainAGE marker, and scores from the Trail Making Test were used as outcomes, with the majority of participants measured at baseline and subsets also measured in a follow-up session. The key findings were a lack of direct association between PA and outcomes, but longitudinal evidence for a higher BrainAge at baseline leading to lower physical capacity at follow-up. This supports a reverse-causation hypothesis in contrast to the prevailing understanding of the positive effects of physical activity on brain health.

Strengths:

The Life-Adult study is a rich and carefully acquired dataset, with multiple follow-up time points. The statistical analyses were conducted carefully with appropriate control for confounds and multiple testing. The study design enables an important assessment for reverse causality. The authors are scrupulous in their consideration of a number of factors that could potentially bias their results, performing an age-stratified analysis, and emphasising discrepancies in PA measurements (specifically, age-reporting bias) across the dataset and other limitations.

Weaknesses:

This is an observational study with inconsistent measures of physical activity. Previous studies have used physical activity interventions, and might be more strongly weighted when considering evidence for these effects (specific confounders involved in interventions notwithstanding).

The model identifying potential reverse causality is relatively limited - it seems possible/likely that brainAge could reflect more general health status, which would expand the potential range of factors underlying this observation.

The important quantitative actigraphy subset is small (n=227), as are the longitudinal subsets. Along with the discrepancy of physical activity/capacity at baseline and follow-up, and other complexities of the dataset, it is difficult to make firm conclusions. The authors point out that the actigraphy subset was quite inactive.

Reviewer #2 (Public review):

Summary:

This population-based cohort study found no evidence that physical activity, whether self-reported or objectively measured, positively influenced brain structure (hippocampal volume or BrainAGE) or cognitive function (Trail Making Test scores). Notably, longitudinal analyses suggested the opposite temporal relationship: a higher BrainAGE at baseline predicted higher physical capacity at follow-up, more in line with reverse causation rather than a neuroprotective effect of physical activity.

Strengths:

The study's statistical approach is thorough and well-documented, and the inclusion of two measurements of physical activity (self-report questionnaire and objective accelerometer data) is a strength. The longitudinal aspect also represents a strength.

Weaknesses:

Several aspects of the measurement timing warrant consideration. Physical activity was assessed over 7-day periods, creating a potential mismatch with (commonly less dynamic) brain outcomes examined (hippocampal volume, BrainAGE), which may reflect cumulative exposures over longer timescales. Additionally, the asynchronous measurement protocol (cognitive testing preceding accelerometry, and the MRI occurring weeks after baseline visits) may introduce time lags that attenuate associations. The observed null associations may be influenced by timing misalignment rather than reflecting the absence of consistent effects of physical activity on brain health and cognition.

Other measurement characteristics also warrant consideration when interpreting the null findings. Physical activity was assessed using short-form self-report questionnaires and averaged accelerometer MET/day values, both of which have limited reliability. Additionally, the modest accelerometer subsample size and low/insufficient variation in activity levels observed in this cohort increase the likelihood of missing effects. These factors collectively raise the possibility that true physical activity-brain health associations may have been obscured.

The study's conclusions regarding brain health, structure, and cognitive functioning are broad despite the scope of the selection of outcomes examined. The analyses focus on hippocampal volume, BrainAGE (a global aging metric), and Trail Making Test performance (processing speed and executive function), while omitting other important neuroimaging markers such as cortical thickness, functional connectivity, or white matter microstructure. The null findings presented here cannot exclude positive effects of physical activity on broader constructs of brain health or cognitive functioning.

While the authors appropriately note the use of different physical activity instruments across time points (IPAQ at baseline, VSAQ at follow-up) in the limitations section, the discussion should more explicitly address the interpretive challenges this creates. The observed association between higher baseline brain age gap and lower follow-up physical activity may reflect: (1) a true temporal relationship, (2) an artifact of switching from behavior-focused (IPAQ) to capacity-focused (VSAQ) measurement, or (3) some combination of both. This ambiguity substantially limits causal inference.

Reviewer #1 (Public review):

Summary:

This manuscript by Zhao et. al investigates the canonical hedgehog pathway in testis development of Nile tilapia. They used complementary approaches with genetically modified tilapia and transfected TSL cells (a clonal stem Leydig cell line) previously derived from 3-mo old tilapia. The approach is innovative and provides a means to investigate DHH and each downstream component from the ptch receptors to the gli and sf1 transcription factors. They concluded that Dhh binds Ptch2 to stimulate Gli1 to promote an increase in Sf1 expression leading to the onset of 11-ketotesterone synthesis heralding the differentiation of Leydig cells in the developing male tilapia.'

Strengths of the methods and results:

- The use of Nile tilapia is important as it is an important aquaculture species, it shares the genetic pathway for sex determination of mammalian species, and molecular differentiation pathways are highly conserved<br /> - The approach is rigorous and incorporates a novel TSL, clonal stem Leydig cell model that they developed that is relatively faithful in following endogenous developmental steps and can produce the appropriate steroid.<br /> - Tilapia are relatively amenable to CRISPR/Cas9 targeting and, with their accelerated developmental time frame, provide an excellent model system to interrogate specific signaling pathways.<br /> - The stepwise analysis from dhh-gli-sf1 is thoughtful and well done.

Weaknesses of the methods and results:

- Line 162: need to establish and verify the PKH26-labeled TSL cells were unaffected by the dhh-/- environment. No data to support the claim that they were unaffected.<br /> - The rescued phenotype caused by the addition of ptch2-/- to the dhh-/- model is a compelling. To further define potential ptch1 contributions, it would be helpful to examine the expression level of ptch1 in the context of the ptch2-/- and ptch2-/-;dhh-/- mutant animals. Any compensatory increase in ptch1 in either case, without obvious phenotype changes, would support the dominant role for ptch2.<br /> - Activity of individual gli factors need additional reconciliation. The expression profiles for both alternative gli factors should be quantified in each knockout cell line to establish redundancy and/or compensation.<br /> - Figure 5E: An important control is missing that includes evaluation of HEK293 cells transfected with pcDNA3.1-OnGli1 without the addition of pGL3-sf1.

Achieved Aims:

The authors set out to test the hypothesis that the canonical Dhh signaling pathway for Leydig cell differentiation and steroidogenic activity is mediated via ptch2 and gli1 regulation of sf1. The results are strong, there are additional steps needed to verify that redundancy/compensation is not contributing to the outcomes.

This work is important in better understanding of nuanced commonalities and differences in developmental pathways across species. Specific to Leydig cell differentiation and steroidogenesis, their work with tilapia supports conservation of the canonical Dhh pathway; however, there appear to be some differences in downstream mediators compared to mouse. Specifically, they conclude that ptch2/gli1 stimulates sf1 and steroidogenesis in tilapia where gli1 is dispensable in mouse. Instead, Gli3 has recently been shown to play an important role to stimulate Sf1 and support the hedgehog pathway.

Reviewer #1 (Public review):

Summary:

This paper aims to characterize the relationship between affinity and fitness in the process of affinity maturation. To this end, the authors develop a model of germinal center reaction and a tailored statistical approach, building on recent advances in simulation-based inference. The potential impact of this work is hindered by the poor organization of the manuscript. In crucial sections, the writing style and notations are unclear and difficult to follow.

Strengths:

The model provides a framework for linking affinity measurements and sequence evolution and does so while accounting for the stochasticity inherent to the germinal center reaction. The model's sophistication comes at the cost of numerous parameters and leads to intractable likelihood, which are the primary challenges addressed by the authors. The approach to inference is innovative and relies on training a neural network on extensive simulations of trajectories from the model.

Weaknesses:

The text is challenging to follow. The descriptions of the model and the inference procedure are fragmented and repetitive. In the introduction and the methods section, the same information is often provided multiple times, at different levels of detail. This organization sometimes requires the reader to move back and forth between subsections (there are multiple non-specific references to "above" and "below" in the text).

The choice of some parameter values in simulations appears arbitrary and would benefit from more extensive justification. It remains unclear how the "significant uncertainty" associated with these parameters affects the results of inference. In addition, the performance of the inference scheme on simulated data is difficult to evaluate, as the reported distributions of loss function values are not very informative.

Finally, the discussion of the similarities and differences with an alternative approach to this inference problem, presented in Dewitt et al. (2025), is incomplete.

Reviewer #2 (Public review):

Summary:

This paper presents a new approach for explicitly transforming B-cell receptor affinity into evolutionary fitness in the germinal center. It demonstrates the feasibility of using likelihood-free inference to study this problem and demonstrates how effective birth rates appear to vary with affinity in real-world data.

Strengths:

(1) The authors leverage the unique data they have generated for a separate project to provide novel insights into a fundamental question.

(2) The paper is clearly written, with accessible methods and a straightforward discussion of the limits of this model.

(3) Code and data are publicly available and well-documented.

Weaknesses (minor):

(1) Lines 444-446: I think that "affinity ceiling" and "fitness ceiling" should be considered independent concepts. The former, as the authors ably explain, is a physical limitation. This wouldn't necessarily correspond to a fitness ceiling, though, as Figure 7 shows. Conversely, the model developed here would allow for a fitness ceiling even if the physical limit doesn't exist.

(2) Lines 566-569: I would like to see this caveat fleshed out more and perhaps mentioned earlier in the paper. While relative affinity is far more important, it is not at all clear to me that absolute affinity can be totally ignored in modeling GC behavior.

(3) One other limitation that is worth mentioning, though beyond the scope of the current work to fully address: the evolution of the repertoire is also strongly shaped by competition from circulating antibodies. (Eg: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3600904/, http://www.sciencedirect.com/science/article/pii/S1931312820303978). This is irrelevant for the replay experiment modeled here, but still an important factor in general repertoires.

Reviewer #1 (Public review):

Summary:

This paper aims to characterize the relationship between affinity and fitness in the process of affinity maturation. To this end, the authors develop a model of germinal center reaction and a tailored statistical approach, building on recent advances in simulation-based inference.

The model provides a framework for linking affinity measurements and sequence evolution and does so while accounting for the stochasticity inherent to the germinal center reaction. The model's sophistication comes at the cost of numerous parameters and leads to intractable likelihood, which are the primary challenges addressed by the authors. The approach to inference is innovative and relies on training a neural network on extensive simulations of trajectories from the model.

The revised methods section is easier to follow and better explains the approach. Inference results on simulated data are compelling and the real-data findings are compared with alternative approaches, clarifying the relationship to previous work.

Reviewer #2 (Public review):

Summary:

This paper presents a new approach for explicitly transforming B cell receptor affinity into evolutionary fitness in the germinal center. It demonstrates the feasibility of using likelihood-free inference to study this problem and demonstrates how effective birth rates appear to vary with affinity in real-world data.

Strengths:

• The authors leverage the unique data they have generated for a separate project to provide novel insights to a fundamental question.
• The paper is clearly written, with accessible methods and straightforward discussion of the limits of this model.
• Code and data are publicly available and well-documented.

Weaknesses:

• No substantial weaknesses noted.

Reviewer #1 (Public review):

In this study, Brickwedde et al. leveraged a cross-modal task where visual cues indicated whether upcoming targets required visual or auditory discrimination. Visual and auditory targets were paired with auditory and visual distractors, respectively. The authors found that during the cue-to-target interval, posterior alpha activity increased along with auditory and visual frequency-tagged activity when subjects were anticipating auditory targets. The authors conclude that their results imply that alpha modulation does not solely regulate 'gain control' in early visual areas (also referred to as alpha inhibition hypothesis), but rather orchestrates signal transmission to later stages of the processing stream.

Comments on the first revision:

I thank the authors for their clarifications. The manuscript is much improved now, in my opinion. The new power spectral density plots and revised Figure 1 are much appreciated. However, there is one remaining point that I am unclear about. In the rebuttal, the authors state the following: "To directly address the question of whether the auditory signal was distracting, we conducted a follow-up MEG experiment. In this study, we observed a significant reduction in visual accuracy during the second block when the distractor was present (see Fig. 7B and Suppl. Fig. 1B), providing clear evidence of a distractor cost under conditions where performance was not saturated."

I am very confused by this statement, because both Fig. 7B and Suppl. Fig. 1B show that the visual- (i.e., visual target presented alone) has a lower accuracy and longer reaction time than visual+ (i.e., visual target presented with distractor). In fact, Suppl. Fig. 1B legend states the following: "accuracy: auditory- - auditory+: M = 7.2 %; SD = 7.5; p = .001; t(25) = 4.9; visual- - visual+: M = -7.6%; SD = 10.80; p < .01; t(25) = -3.59; Reaction time: auditory- - auditory +: M = -20.64 ms; SD = 57.6; n.s.: p = .08; t(25) = -1.83; visual- - visual+: M = 60.1 ms ; SD = 58.52; p < .001; t(25) = 5.23)."

These statements appear to directly contradict each other. I appreciate that the difficulty of auditory and visual trials in block 2 of MEG experiments are matched, but this does not address the question of whether the distractor was actually distracting (and thus needed to be inhibited by occipital alpha). Please clarify.

Comments on the latest version:

I am satisfied with the author's response and do not have any additional comments.

Reviewer #1 (Public Review):

Summary:

In their paper, Zhan et al. have used Pf genetic data from simulated data and Ghanaian field samples to elucidate a relationship between multiplicity of infection (MOI) (the number of distinct parasite clones in a single host infection) and force of infection (FOI). Specifically, they use sequencing data from the var genes of Pf along with Bayesian modeling to estimate MOI individual infections and use these values along with methods from queueing theory that rely on various assumptions to estimate FOI. They compare these estimates to known FOIs in a simulated scenario and describe the relationship between these estimated FOI values and another commonly used metric of transmission EIR (entomological inoculation rate).

This approach does fill an important gap in malaria epidemiology, namely estimating the force of infection, which is currently complicated by several factors including superinfection, unknown duration of infection, and highly genetically diverse parasite populations. The authors use a new approach borrowing from other fields of statistics and modeling and make extensive efforts to evaluate their approach under a range of realistic sampling scenarios. However, the write-up would greatly benefit from added clarity both in the description of methods and in the presentation of the results. Without these clarifications, rigorously evaluating whether the author's proposed method of estimating FOI is sound remains difficult. Additionally, there are several limitations that call into question the stated generalizability of this method that should at minimum be further discussed by authors and in some cases require a more thorough evaluation.

Major comments:

(1) Description and evaluation of FOI estimation procedure.

a. The methods section describing the two-moment approximation and accompanying appendix is lacking several important details. Equations on lines 891 and 892 are only a small part of the equations in Choi et al. and do not adequately describe the procedure notably several quantities in those equations are never defined some of them are important to understand the method (e.g. A, S as the main random variables for inter-arrival times and service times, aR and bR which are the known time average quantities, and these also rely on the squared coefficient of variation of the random variable which is also never introduced in the paper). Without going back to the Choi paper to understand these quantities, and to understand the assumptions of this method it was not possible to follow how this works in the paper. At a minimum, all variables used in the equations should be clearly defined.

b. Additionally, the description in the main text of how the queueing procedure can be used to describe malaria infections would benefit from a diagram currently as written it's very difficult to follow.

c. Just observing the box plots of mean and 95% CI on a plot with the FOI estimate (Figures 1, 2, and 10-14) is not sufficient to adequately assess the performance of this estimator. First, it is not clear whether the authors are displaying the bootstrapped 95%CIs or whether they are just showing the distribution of the mean FOI taken over multiple simulations, and then it seems that they are also estimating mean FOI per host on an annual basis. Showing a distribution of those per-host estimates would also be helpful. Second, a more quantitative assessment of the ability of the estimator to recover the truth across simulations (e.g. proportion of simulations where the truth is captured in the 95% CI or something like this) is important in many cases it seems that the estimator is always underestimating the true FOI and may not even contain the true value in the FOI distribution (e.g. Figure 10, Figure 1 under the mid-IRS panel). But it's not possible to conclude one way or the other based on this visualization. This is a major issue since it calls into question whether there is in fact data to support that these methods give good and consistent FOI estimates.

d. Furthermore the authors state in the methods that the choice of mean and variance (and thus second moment) parameters for inter-arrival times are varied widely, however, it's not clear what those ranges are there needs to be a clear table or figure caption showing what combinations of values were tested and which results are produced from them, this is an essential component of the method and it's impossible to fully evaluate its performance without this information. This relates to the issue of selecting the mean and variance values that maximize the likelihood of observing a given distribution of MOI estimates, this is very unclear since no likelihoods have been written down in the methods section of the main text, which likelihood are the authors referring to, is this the probability distribution of the steady state queue length distribution? At other places the authors refer to these quantities as Maximum Likelihood estimators, how do they know they have found the MLE? There are no derivations in the manuscript to support this. The authors should specify the likelihood and include in an appendix an explanation of why their estimation procedure is in fact maximizing this likelihood, preferably with evidence of the shape of the likelihood, and how fine the grid of values they tested is for their mean and variance since this could influence the overall quality of the estimation procedure.

(2) Limitation of FOI estimation procedure.

a. The authors discuss the importance of the duration of infection to this problem. While I agree that empirically estimating this is not possible, there are other options besides assuming that all 1-5-year-olds have the same duration of infection distribution as naïve adults co-infected with syphilis. E.g. it would be useful to test a wide range of assumed infection duration and assess their impact on the estimation procedure. Furthermore, if the authors are going to stick to the described method for duration of infection, the potentially limited generalizability of this method needs to be further highlighted in both the introduction, and the discussion. In particular, for an estimated mean FOI of about 5 per host per year in the pre-IRS season as estimated in Ghana (Figure 3) it seems that this would not translate to 4-year-old being immune naïve, and certainly this would not necessarily generalize well to a school-aged child population or an adult population.

b. The evaluation of the capacity parameter c seems to be quite important and is set at 30, however, the authors only describe trying values of 25 and 30, and claim that this does not impact FOI inference, however it is not clear that this is the case. What happens if the carrying capacity is increased substantially? Alternatively, this would be more convincing if the authors provided a mathematical explanation of why the carrying capacity increase will not influence the FOI inference, but absent that, this should be mentioned and discussed as a limitation.

Reviewer #2 (Public Review):

Summary:

The authors combine a clever use of historical clinical data on infection duration in immunologically naive individuals and queuing theory to infer the force of infection (FOI) from measured multiplicity of infection (MOI) in a sparsely sampled setting. They conduct extensive simulations using agent-based modeling to recapitulate realistic population dynamics and successfully apply their method to recover FOI from measured MOI. They then go on to apply their method to real-world data from Ghana before and after an indoor residual spraying campaign.

Strengths:

(1) The use of historical clinical data is very clever in this context.

(2) The simulations are very sophisticated with respect to trying to capture realistic population dynamics.

(3) The mathematical approach is simple and elegant, and thus easy to understand.

Weaknesses:

(1) The assumptions of the approach are quite strong and should be made more clear. While the historical clinical data is a unique resource, it would be useful to see how misspecification of the duration of infection distribution would impact the estimates.

(2 )Seeing as how the assumption of the duration of infection distribution is drawn from historical data and not informed by the data on hand, it does not substantially expand beyond MOI. The authors could address this by suggesting avenues for more refined estimates of infection duration.

(3) It is unclear in the example how their bootstrap imputation approach is accounting for measurement error due to antimalarial treatment. They supply two approaches. First, there is no effect on measurement, so the measured MOI is unaffected, which is likely false and I think the authors are in agreement. The second approach instead discards the measurement for malaria-treated individuals and imputes their MOI by drawing from the remaining distribution. This is an extremely strong assumption that the distribution of MOI of the treated is the same as the untreated, which seems unlikely simply out of treatment-seeking behavior. By imputing in this way, the authors will also deflate the variability of their estimates.

- For similar reasons, their imputation of microscopy-negative individuals is also questionable, as it also assumes the same distributions of MOI for microscopy-positive and negative individuals.

Reviewer #3 (Public Review):

Summary:

It has been proposed that the FOI is a method of using parasite genetics to determine changes in transmission in areas with high asymptomatic infection. The manuscript attempts to use queuing theory to convert multiplicity of infection estimates (MOI) into estimates of the force of infection (FOI), which they define as the number of genetically distinct blood-stage strains. They look to validate the method by applying it to simulated results from a previously published agent-based model. They then apply these queuing theory methods to previously published and analysed genetic data from Ghana. They then compare their results to previous estimates of FOI.

Strengths:

It would be great to be able to infer FOI from cross-sectional surveys which are easier and cheaper than current FOI estimates which require longitudinal studies. This work proposes a method to convert MOI to FOI for cross-sectional studies. They attempt to validate this process using a previously published agent-based model which helps us understand the complexity of parasite population genetics.

Weaknesses:

(1) I fear that the work could be easily over-interpreted as no true validation was done, as no field estimates of FOI (I think considered true validation) were measured. The authors have developed a method of estimating FOI from MOI which makes a number of biological and structural assumptions. I would not call being able to recreate model results that were generated using a model that makes its own (probably similar) defined set of biological and structural assumptions a validation of what is going on in the field. The authors claim this at times (for example, Line 153 ) and I feel it would be appropriate to differentiate this in the discussion.

(2) Another aspect of the paper is adding greater realism to the previous agent-based model, by including assumptions on missing data and under-sampling. This takes prominence in the figures and results section, but I would imagine is generally not as interesting to the less specialised reader. The apparent lack of impact of drug treatment on MOI is interesting and counterintuitive, though it is not really mentioned in the results or discussion sufficiently to allay my confusion. I would have been interested in understanding the relationship between MOI and FOI as generated by your queuing theory method and the model. It isn't clear to me why these more standard results are not presented, as I would imagine they are outputs of the model (though happy to stand corrected - it isn't entirely clear to me what the model is doing in this manuscript alone).

(3) I would suggest that outside of malaria geneticists, the force of infection is considered to be the entomological inoculation rate, not the number of genetically distinct blood-stage strains. I appreciate that FOI has been used to explain the latter before by others, though the authors could avoid confusion by stating this clearly throughout the manuscript. For example, the abstract says FOI is "the number of new infections acquired by an individual host over a given time interval" which suggests the former, please consider clarifying.

(4) Line 319 says "Nevertheless, overall, our paired EIR (directly measured by the entomological team in Ghana (Tiedje et al., 2022)) and FOI values are reasonably consistent with the data points from previous studies, suggesting the robustness of our proposed methods". I would agree that the results are consistent, given that there is huge variation in Figure 4 despite the transformed scales, but I would not say this suggests a robustness of the method.

(5) The text is a little difficult to follow at times and sometimes requires multiple reads to understand. Greater precision is needed with the language in a few situations and some of the assumptions made in the modelling process are not referenced, making it unclear whether it is a true representation of the biology.

Reviewer #2 (Public review):

Summary:

The authors combine a clever use of historical clinical data on infection duration in immunologically naive individuals and queuing theory to infer the force of infection (FOI) from measured multiplicity of infection (MOI) in a sparsely sampled setting. They conduct extensive simulations using agent based modeling to recapitulate realistic population dynamics and successfully apply their method to recover FOI from measured MOI. They then go on to apply their method to real world data from Ghana before and after an indoor residual spraying campaign.

Strengths:

- The use of historical clinical data is very clever in this context

- The simulations are very sophisticated with respect to trying to capture realistic population dynamics

- The mathematical approach is simple and elegant, and thus easy to understand

Weaknesses:

- The assumptions of the approach are quite strong, and the authors have made clear that applicability is constrained to individuals with immune profiles that are similar to malaria naive patients with neurosyphilis. While the historical clinical data is a unique resource and likely directionally correct, it remains somewhat dubious to use the exact estimated values as inputs to other models without extensive sensitivity analysis.

Comments on revisions:

The authors have adequately responded to all comments.

Reviewer #1 (Public review):

This manuscript adds to the recent, exciting developments in our understanding of the MmpL/S transporters from mycobacteria. This work provides solid support for the trimeric/hexameric arrangement of subunits in the complex, and reveals a possible pathway for substrate translocation.

Overall, I think this manuscript is a solid body of work that adds to several recent studies from this team and others on the structure and mechanism of the MmpL/S transporter family, particularly MmpL4/S4. The combination of AF, disulfide engineering, and experimental structure is good, though it is a bit puzzling that the experimental structure based on disulfide stabilization of the AF prediction does not recapitulate key elements (MmpS periplasmic domain docking to MmpL, and altered CCD configuration).

I have no major concerns about this manuscript.

Reviewer #2 (Public review):

Summary:

The manuscript describes the structure of the Mycobacterium tuberculosis (MmpS4)3-(MmpL4)3 hetero-heximeric transporter complex. The structure was obtained by cryogenic electron microscopy using an engineered construct that cross-links MmpS4 to MmpL4 via a disulfide bond. The position of the disulfide bond was determined using an Alphafold2 model of the hetero-heximer. Although Alphafold2 predicts a symmetric hetero-heximer, the author found that the structure of the coiled-coil domain (CCD) is asymmetric, tilted at about 60° relative to the membrane domains, and only contains two of the three alpha helical hairpins, with the third being disordered.

Strengths:

The strategy of using Alphafold2 models to guide construct design for experimental structure determination is state-of-the-art, and this work provides a great example of its applications and limitations. I.e., the experimental structure does not fully recapitulate the prediction but provides unexpected results.

The comparisons between the authors' structures and the previously published structures of the MmpL4 monomer and MmpL5 trimers strengthen the authors' findings.

Weaknesses:

A more detailed description of the current mechanistic hypothesis would strengthen the manuscript. The authors state that the two periplasmic domains "are expected to undergo rigid body movements that allow substrate transport through these periplasmic domains similar to the conformational changes observed in the E. coli multidrug efflux pump AcrB". A schematic of the proposed transport cycle, as a supplemental figure that shows the current hypothesis regarding transport, would be beneficial for understanding the previous structures and putting the current structure in context. Outside of "the mechanistic basis of how these conformational changes are coupled to protonation of the DY-pairs", what are the major controversies/open questions regarding the mechanism?

The authors provide evidence that the cysteine-depleted S4L4 construct is functional, but do not show that the construct with the introduced disulfide bond #5 (D39C MmpS4 and S434C MmpL4) is also functional. Demonstrating this would allow the authors to better interpret their resulting structures.

The analysis presented in Figure 5 and Supplementary Figure 7 seems to suggest that the authors are proposing that the CCD central cavity acts as a transport pathway for the transported substrate, but I am not sure that this hypothesis is explicitly stated. This makes the reasoning behind the analysis presented unclear. Clarity could be improved by stating that the hypothesis of direct transport of substrate through the CCD central channel is being examined using the structure prediction, and what the implications are for the structure solved with the incompletely formed CCD.

Given that the results emphasize the flexibility of the CCD, the manuscript would be strengthened by 3D variability analysis either in cryoSPARC or using cryoDRGN (or both). This would allow the authors to better quantify the degree of motion in the CCD and how it may correlate to flexibility in other regions. Further 3D flex reconstruction in cryoSPARC may improve the map quality of the CCD.

Reviewer #3 (Public review):

Summary:

This manuscript by Earp et al reports cryoEM structures of the hexameric (MmpS4)₃-(MmpL4)₃ complex from Mycobacterium tuberculosis, which belongs to the RND family of transporters and is known to have a role in the export of siderophores and contribute to drug resistance. The experimental workflow showcased involves the design of disulfide pairs using distance constraints obtained from the AlphaFold predicted structure of the hexameric complex. One such disulfide pair was used to determine the ~3.0 Å structures. The structure reveals density for the previously unresolved coiled-coil domain (CCD), a tilted CCD arrangement, and a cavity within the periplasmic domain, which the authors assert is occupied by detergent. Comparison of this complex with the monomer structure of MmpL4 shows conformational variations interpreted to implicate different domains and conserved residues involved in proton coupling, which might be related to the transport mechanism. While the methodological aspects of the manuscript are solid, enthusiasm for the overall advance/significance is less so, with doubts about the relevance of the tilted CCD structure, considering disulfide trapping and an incomplete validation of the claim that the titled CCD represents a stable intermediate conformation. A clear, updated transport mechanism is largely missing from the manuscript.

Strengths:

Beautiful structures, AF prediction-experimental validation nexus that could be fine-tuned for different systems/difficult to target complexes.

Weaknesses:

Physiological relevance of the tilted CCD conformation. No clear mechanistic model for the transport. While the CCD may indeed be a stable intermediate, the fact that the rest of the trimeric arrangement is unaffected does not fully rule out disulfide trapping as a factor in promoting this. The findings would be strengthened if the same tilted conformation is seen using a different set of disulfides. The significance of the detergent molecule and the new cavity observed could also be better discussed in terms of an updated transport model.

Reviewer #1 (Public review):

Summary:

In "Drift in Individual Behavioral Phenotype as a Strategy for Unpredictable Worlds," Maloney et al. (2026) investigate changes in individual responses over time, referred to as behavioral drift within the lifespan of an animal. Drift, as defined in the paper, complements stable behavioral variation (animal individuality/personality within a lifetime) over shorter timeframes, which the authors associate with an underlying bet-hedging strategy. The third timeframe of behavioral variability that the authors discuss occurs within seasons (across several generations of some insects), termed "adaptive tracking." This division of "adaptive" behavioral variability over different timeframes is intuitively logical and adds valuable depth to the theoretical framework concerning the ecological role of individual behavioral differences in animals.

Strengths:

While the theoretical foundations of the study are compelling, the connection between the experimental data (Fig. 1) and the modeling work (Fig. 2-4) is convincing.

Weaknesses:

In the experimental data (Fig. 1), the authors describe the changes in behavioral preferences over time. While generally plausible, I had identified three significant issues with the experiments that were addressed in the revision:

(1) All of the subsequent theoretical/simulation data is based on changing environments, yet all the experiments are conducted in unchanging environments. While this may suffice to demonstrate the phenomenon of behavioral instability (drift) over time, it does not fully link to the theory-driven work in changing environments. A full experimental investigation of this would be beyond the scope of the current work.

(2) The temporal aspect of behavioral instability has been addressed in Figure 1F.

(3) The temporal dimension leads directly into the third issue: distinguishing between drift and learning (e.g., line 56). This issue has been further discussed in the revised manuscript.

Reviewer #2 (Public review):

Summary:

This is an inspired study that merges the concept of individuality with evolutionary processes to uncover a new strategy that diversifies individual behavior that is also potentially evolutionarily adaptive.

The authors use time-resolved measurement of spontaneous, innate behavior, namely handedness or turn bias in individual, isogenic flies, across several genetic backgrounds.

They find that an individual's behavior changes over time, or drifts. This has been observed before, but what is interesting here is that by looking at multiple genotypes, the authors find the amount of drift is consistent within genotype i.e., genetically regulated, and thus not entirely stochastic. This is not in line with what is known about innate, spontaneous behaviors. Normally, fluctuations in behavior would be ascribed to a response to environmental noise. However, here, the authors go on to find what is the pattern or rule that determines the rate of change of the behavior over time within individuals. Using modeling of behavior and environment in the context of evolutionarily important timeframes such as lifespan or reproductive age, they could show when drift is favored over bet-hedging and that there is an evolutionary purpose to behavioral drift. Namely, drift diversifies behaviors across individuals of the same genotype within the timescale of lifespan, so that the genotype's chance for expressing beneficial behavior is optimally matched with potential variation of environment experienced prior to reproduction. This ultimately increases fitness of the genotype. Because they find that behavioral drift is genetically variable, they argue it can also evolve.

Strengths:

Unlike most studies of individuality, in this study, authors consider the impact of individuality on evolution. This is enabled by the use of multiple natural genetic backgrounds and an appropriately large number of individuals to come to the conclusions presented in the study. I thought it was really creative to study how individual behavior evolves over multiple timescales. And indeed this approach yielded interesting and important insight into individuality. Unlike most studies so far, this one highlights that behavioral individuality is not a static property of an individual, but it dynamically changes. Also, placing these findings in the evolutionary context was beneficial. The conclusion that individual drift and bet-hedging are differently favored over different timescales is, I think, a significant and exciting finding.

Overall, I think this study highlights how little we know about the fundamental, general concepts behind individuality and why behavioral individuality is an important trait. They also show that with simple but elegant behavioral experiments and appropriate modeling, we could uncover fundamental rules underlying the emergence of individual behavior. These rules may not at all be apparent using classical approaches to studying individuality, using individual variation within a single genotype or within a single timeframe.

Weaknesses:

I am unconvinced by the claim that serotonin neuron circuits are regulating behavioral drift, especially because of its bidirectional effect and lack of relative results for other neuromodulators. Without testing other neuromodulators, it will remain unclear if serotonin intervention increases behavioral noise within individuals, or if any other pharmacological or genetic intervention would do the same. Another issue is that the amount of drugs that the individuals ingested was not tracked. Variable amounts can result in variable changes in behavior that are more consistent with the interpretation of environmental plasticity, rather than behavioral drift. With the current evidence presented, individual behavior may change upon serotonin perturbation, but this does not necessarily mean that it changes or regulates drift.

However, I think for the scope of this study, finding out whether serotonin regulates drift or not is less important. I understand that today there is a strong push to find molecular and circuit mechanisms of any behavior, and other peers may have asked for such experiments, perhaps even simply out of habit. Fortunately, the main conclusions derived from behavioral data across multiple genetic backgrounds and the modeling are anyway novel, interesting and in fact more fundamental than showing if it is serotonin that does it or not.

To this point, one thing that was unclear from the methods section is whether genotypes that were tested were raised in replicate vials and how was replication accounted for in the analyses. This is a crucial point - the conclusion that genotypes have different amounts of behavioral drift cannot be drawn without showing that the difference in behavioral drift does not stem from differences in developmental environment.

Comments on the latest version:

The changes to the manuscript sufficiently addressed my few comments. I do not have anything else substantial to add to my review and I am comfortable with my initial assessment.

Reviewer #3 (Public review):

The paper begins by analyzing the drift in individual behavior over time. Specifically, it quantifies the circling direction of freely walking flies in an arena. The main takeaway from this dataset is that while flies exhibit an individual turning bias (when averaged over time), yet their preferences fluctuate over slow timescales.

To understand whether genetic or neuromodulatory mechanisms influence the drift in individual preference, the authors test different fly strains in a Y maze concluding that both genetic background and the neuromodulator serotonin contribute to the degree of drift (although with some contrasting results). The use of a different assay for this different dataset (Y maze istead of wide arena) is justified by previous observation of similar behavioral biases in these assay. Yet the conceptual link between the spectral power analysis used for the first dataset and the autoregressive model used for the second remains unclear.

Finally, the authors use theoretical approaches to show the potential advantage of individual drift for survival in unpredictable, fluctuating environments. They demonstrate that while bet-hedging provides an advantage over timescales matching the generation time (since reproduction is required), it offers less benefit on shorter timescales, where an increased individual drift could be advantageous.

Reviewer #1 (Public review):

This manuscript characterizes the effects of isoflurane on visual processing in layer 2/3 of the mouse primary visual cortex (V1). General anesthesia, including isoflurane, has been reported to modulate various neural processes, such as size tuning, direction selectivity, and spatial selectivity in V1. Using two-photon calcium imaging, the authors monitored neural responses to visual stimuli under isoflurane anaesthesia and found that spatial frequency preferences are also affected across cell types, with the magnitude and direction of these effects varying between cell types.

The authors performed careful and rigorous comparisons of neuronal responses between the two conditions using well-chosen nonparametric statistics. At the same time, because two-photon calcium imaging can be combined with cell-type-specific labeling, the authors labelled inhibitory neurons with tdTomato, allowing them to distinguish GCaMP activities in excitatory and specific inhibitory cell classes. We also appreciated that the manuscript provides not only summary statistics but also example GCaMP traces (Figure 1), which makes it easy for readers to understand the quality of the raw data.

We believe that the manuscript could be improved by emphasizing the following three points.

(1) The analyses are limited to the neurons that responded to visual stimuli in both the anesthetized and awake states. According to Table S1, the proportion of visually responsive neurons that met such criteria is only 27.4% for the excitatory neurons. This raises the potential concerns that the reported effects of isoflurane may not fully reflect population-level changes in visual coding. We suggest that the authors repeat the same analyses, including average tuning curves and decoding analyses, for all recorded neurons in each condition.

(2) The manuscript would benefit from tuning curves of spatial frequency preference for individual neurons, as this would help readers assess whether the reported statistics are appropriate (Figures 2A-D). In addition, more in-depth single-neuron analyses would help distinguish between the two proposed hypotheses in Figure 5 that may not be evident from average responses alone. This is because, with the current analysis, it is not clear how the shape of the tuning curves will affect the estimation of spatial frequency preference. To address this potential concern and strengthen the interpretation of the results, we suggest:<br /> a) repeating the analysis at the level of individual neuronal responses, instead of average responses, and<br /> b) using simulated data to examine how changes in tuning-curve width could affect estimated spatial frequency preference.

For example, using the neuronal responses in the awake condition, one could broaden the tuning curves and recompute the preferred spatial frequency, then compare the resulting distribution with that observed under anesthesia.

(3) We believe the manuscript's overall framing is a little broader than what is directly supported by the data. In particular:

(a) the statement "reduced sensory perception during anesthesia is linked to a degradation in spatial resolution at the cellular level" in the Abstract is an unclear and unsupported claim. We suggest removing this sentence and more directly summarizing the findings.

(b) given the discrepancy between the effects of urethane and isoflurane as laid out in the discussion, the current title "Anesthesia Lowers Spatial Frequency Preference in the Primary Visual Cortex" appears overstated and should be revised to explicitly reflect the specific anesthetic tested: "Isoflurane Anesthesia Lowers Spatial Frequency Preference in the Primary Visual Cortex".

Reviewer #2 (Public review):

Summary:

The main objective of the study was to link the changes in brain state due to anesthesia to consequences on visual neural processing, particularly effects on spatial frequency tuning. This is accomplished by 2-photon imaging of excitatory and inhibitory neurons (separating PV- and SST-positive subtypes) in mouse visual cortex during full-field visual stimulation with gratings, and tracking neuronal tuning for spatial frequency before, during, and after isoflurane anesthesia. The main finding is that anesthesia induces lower spatial frequency preferences in excitatory neurons, and this leads to poorer population representations (decoding) of higher spatial frequency responses during anesthesia. A second main finding is that anesthesia impacts inhibitory neuron subtypes in distinct ways, with the most pronounced effects of anesthesia on somatostatin inhibitory neurons.

Strengths:

(1) A main strength is that the study is that it is straightforward, and reassuringly, the results confirm multiple previous studies showing anesthesia's effects on the amplitude of cortical responses: larger and less selective responses in excitatory neurons (versus awake responses); strongly reduced responses in somatostatin inhibitory neurons (versus awake responses) (Fig. 5I-L), with less differences across anesthetized and awake states on response amplitude of PV neurons.

(2) These confirmations of prior observations (on the amplitude of responses) establish good ground for the new results on spatial frequency tuning. For excitatory neurons, spatial frequency selectivity shifts to higher values in awake versus anesthetized conditions; this is because anesthesia induces larger responses to lower spatial frequencies. In somatostatin neurons, instead, wakefulness reduces the lower spatial frequency responses present in anesthesia, and dramatically increases the overall amplitude of responses and medium and higher spatial frequencies. This is consistent with prior work showing that in awake states, somatostatin neurons exert broad inhibition in V1; this study extends that finding to the tuning of spatial frequencies.

Weaknesses:

(1) A first weakness of the study is the lack of examination of changes to single neuron receptive field sizes and/or surround suppression across conditions, and how these may relate to the effects on spatial frequency tuning with full field gratings. There is a well-known relationship between the size of the receptive field and the resulting selectivity for spatial frequencies (i.e., large receptive fields prefer lower spatial frequency stimuli). Likewise, there are many studies showing how surround suppression / spatial integration is impacted by anesthesia (and arousal). A more detailed examination of all these related quantities on an individual neuron basis would provide a greater understanding of the factors underlying the effects on spatial frequency tuning. One could imagine that receptive field changes, and/or changes in surround suppression, influence the selectivity to full-field gratings.

(2) A second weakness is the lack of examination/insight into the temporal dynamics of the effects. The experimental paradigm records activity across control, anesthesia, and recovery epochs in a single duration (~40 mins) session. The epochs are simply binned together ("Awake", "Anes.", "Recover"). It is not clear how the start of the anesthesia bin is defined, nor is it clear how the recovery period is defined. It is also not clear what the changes are to motor tone, brain state, etc., that are also strong influences on visual responses in mouse V1. Presumably, these onset/offset effects are similar enough across mice and sessions that they affect all the bins in the same way, but greater examination of the temporal effects in excitatory, PV, and SOM neurons could shed light on interactions driving the changes. Is there some temporal dependence of anesthesia on selectivity changes across the cell types? For example, at the onset of anesthesia, are SOM neurons losing broadband frequency responses before the excitatory neurons gain low frequency responses? Do PV neurons also show effects after the changes in SOM neurons (suggesting strong SOM -> PV inhibition)? Such analysis might shed light on the timing/causality of the effects among these 3 neuron types.

(3) A third weakness concerns the interpretation of the low and high arousal conditions during awake states (Figure 6). It is not clear how movement (or lack of movement) impacts the high arousal epochs, nor is it clear how the low arousal condition compares to the brain state during anesthesia. For example, deep versus light anesthesia can lead to synchronized or asynchronous states, respectively, and low arousal in wakefulness can show strong low-frequency oscillations of activity, which could promote a lower excitability state than light anesthesia. Without some more detail about commonly measured brain state or body/face motion metrics, it is difficult to know what brain states are represented by the bins and how to interpret the comparisons.

Overall, the study uses adequate methods and experimental design to demonstrate solid support for the (somewhat narrow) central finding that anesthesia lowers the spatial resolution of mouse V1 responses.

Since this is a very well-examined topic, the findings here are not totally surprising, but confirmatory and slightly extend prior findings (a good thing). As such, the study will likely have most relevance to specialists in the mouse visual system, but if the study could address some of the remaining questions discussed above, this would potentially broaden the implications of the study to general insights about the operation of cortical circuitry.

Reviewer #3 (Public review):

Summary:

This manuscript is focused on studying the spatial frequency selectivity of individual neurons in the mouse primary visual cortex (V1) in the anesthetized and awake brain states using 2-photon calcium imaging. Although previous studies have demonstrated that anesthesia decreases both size tuning and spatial selectivity in V1 neurons, the strength of this study is its focus on characterization of the same neurons in awake and anesthetized states in combination with transgenic mouse lines selectively labeling pan-inhibitory neurons and also more specific neuronal subtypes, including parvalbumin-positive (PV+) or somatostatin-positive (SOM+) interneurons. A combination of these methodologies allows for a more in-depth mechanistic study of the properties of different types of neurons. The main findings suggest that in excitatory neurons, anesthesia leads to a shift in preferred SF and broadening of SF tuning, with no changes in orientation and direction selectivity. Downward shift in preferred SF was more pronounced in both SOM+ and PV+ interneurons.

Strengths:

(1) 2-photon calcium imaging with single-cell resolution.

(2) Characterization of excitatory and two types of inhibitory neurons.

Weaknesses:

(1) VIP interneurons are critical to the neural circuit, and their characterization would be critical to the mechanistic understanding of this process, but is missing.

(2) Unfortunately, the manuscript does not lead to an additional insight into the nature of this anesthesia-induced shift in SF preference.

(3) Furthermore, it also doesn't help understand how SF preference is encoded in V1.

(4) Finally, some critical histological controls are missing.

Reviewer #1 (Public review):

Summary:

The manuscript entitled "Evaluation of Antibiotic and Peptide Vaccine Strategies for Mirror Bacterial Infections" addresses a topic that is well established in the literature. The authors investigate the activity of enantiomeric (D-form) antibiotics against bacteria and the immunogenicity of D-form peptides, proposing that D-enantiomers are ineffective both as antibacterial agents and as vaccine candidates. While the subject matter is relevant, the concepts explored are already well known, and the manuscript offers limited novelty.

The authors demonstrate that D-enantiomeric antibiotics lack antibacterial activity compared to their naturally occurring L-forms and that D-form peptides fail to elicit detectable immune responses. These observations are consistent with existing knowledge regarding molecular chirality in biological systems. However, the manuscript relies on a limited experimental dataset while extrapolating the findings broadly, which weakens the strength of the conclusions.

Strengths:

The manuscript introduces the topic of Mirror Bacterial Infections, likely to occur if no regulations or restrictions are placed immediately.

The manuscript addresses a relevant topic and has potential value, particularly in framing discussions around chirality and pathogen interactions. With a more cautious interpretation of the results, the manuscript could better justify its conceptual framework and strengthen its contribution to the field.

Weaknesses:

(1) Several sections of the manuscript are overly descriptive and would benefit from deeper comparative analysis and critical synthesis. In multiple instances, the discussion relies on hypothetical scenarios supported primarily by selective citations rather than robust experimental evidence. The introduction of the term "mirror microbiology" or "mirror bacteria" appears largely conceptual and is used to unify what are essentially two separate lines of investigation, enantioselective antibiotic activity and peptide chirality in immune recognition, without sufficient mechanistic integration.

(2) To the best of this reviewer's understanding, the manuscript does not present substantial novelty. The pronounced differences in biological activity between L- and D-forms of small molecules and peptides are well documented, including their implications for antimicrobial efficacy and immune recognition. While the manuscript is written in clear and accessible language suitable for both specialists and interdisciplinary readers, novelty remains limited.

The manuscript reiterates well-established principles of stereochemistry and biological recognition. Given the extensive existing literature demonstrating that enantiomeric antibiotics are typically inactive due to stereospecific target interactions, the failure of D-form antibiotics is expected and does not constitute a novel finding.

(3) Critical experimental details are lacking, particularly regarding the peptide design. It is unclear whether the peptides were synthesized entirely in the D-configuration or whether only select amino acids were substituted. This distinction is essential for interpreting immunogenicity results and for comparison with prior studies.

(4) The authors conclude that D-form peptides are poorly recognized by the immune system. However, the data presented indicate that neither the L- nor the D-form peptides tested elicited a measurable immune response. Without demonstrating immunogenicity of the corresponding L-form peptides, the conclusion that immune non-recognition is specific to the D-form is not sufficiently supported.

Reviewer #2 (Public review):

This paper by Kleinman et al. tackles an increasingly discussed biosecurity scenario, namely the possibility that "mirror bacteria" could evade key elements of host immunity and therefore demand bespoke medical countermeasures. The authors experimentally probe two such countermeasure concepts: (1) whether existing chiral antibiotics might still work against mirror bacteria (this is tested indirectly by measuring the activity of antibiotic enantiomers against natural-chirality bacteria), and (2) whether D-peptide antigens can be made immunogenic. Briefly, the authors show that enantiomers of four approved antibiotics have little to no activity in MIC assays, argue this implies the parent drugs would likely fail against mirror bacteria, report limited single-dose tolerability data for the enantiomers in mice, and show that selected bacterially derived D-peptides can elicit strong binding antibody titers when conjugated to a carrier protein and given with adjuvant.

Overall, the study is quite interesting but constrained by the fact that D-peptide immunogens and related ideas have been explored for decades, by prior literature showing that D-enantiomeric peptides can themselves be strongly antimicrobial vs conventional bacteria, and by a number of conceptual and experimental limitations outlined below.

(1) A blanket statement indicating that flipping chirality makes antibiotics ineffective cannot be true across all classes. Indeed, there is extensive precedent for "mirror" (D-amino-acid) peptides that retain, or even improve, antimicrobial activity against natural bacteria.

(2) The paper's key claim ("parent antibiotics won't work on mirror bacteria") is based on the observation that the enantiomers of chloramphenicol/linezolid/tedizolid/aztreonam largely lose activity against natural bacteria. This is a reasonable proxy experiment given the absence of mirror organisms, but it remains an inference and should be described as such.

(3) The chiral purity needs to be documented more rigorously. The methods mention structural confirmation by NMR and >95% purity by LC-MS/HPLC for enantiomeric compounds, but this is not the same as demonstrating high enantiomeric excess or excluding low-level contamination by the active parent enantiomer.

(4) The residual activity of ent-aztreonam is quite interesting. The authors report slight activity for ent-aztreonam (MIC of 32-128 µg/mL in a subset), still far weaker than aztreonam but nonzero.

(5) For antibiotics, MIC is a starting point, but further experiments are needed. To justify countermeasure relevance, it would help to include at least one additional pharmacodynamic readout (time-kill kinetics, post-antibiotic effect, inoculum effect, or activity in the presence of human serum).

(6) The acute toxicity study is limited (single-dose, short follow-up, small n, one sex/strain, and no histopathology).

(7) The Discussion leans on human equivalent dosing logic to reassure feasibility. Given the lack of PK, bioavailability, metabolism, and repeat-dose data, these comparisons risk overreach.

(8) The readout is ELISA endpoint binding (IgG; and IgA in BALF for one antigen), which is fine for an initial immunogenicity screen. But the manuscript then drifts toward "vaccine strategy" claims without showing any antibody functionality (opsonophagocytosis, complement deposition, neutralization, blocking adhesion, and so on) or even binding to a more native-like antigen format (e.g., D-peptide displayed on particles; D-protein fragments; or any surrogate that goes beyond plate-bound peptide).

(9) The methods report peptide conjugates containing ~10-200 EU/mL endotoxin. That is not trivial and could materially amplify immunogenicity, and should be discussed.

(10) The authors should report how many technical/biological replicates were performed for MIC determinations and for ELISAs.

Reviewer #3 (Public review):

Summary:

There is a threat of mirror life bacteria, which could possibly evade immunity and cause problems for human/animal hosts. This paper evaluates enantiomeric antibiotics and vaccines as a means to understand how this could be combatted in the future.

Strengths:

It is valuable to collect such information, as it is not always clear how an antibiotic in its enantiomeric form would interact with a bacterium in terms of its MIC or towards toxicity. The paper is scientifically sound with regard to assays and statistical methods.

Weaknesses:

The beginning of the paper could be described as hyperbolic. For a paper that demonstrates that mirror-image molecules have (expected) lower MICs and toxicity, some of the claims in the beginning that they are going to cause a pandemic of evading the immune system seem to be a bit overstated. If they are mirror images, how are these bacteria going to generate virulence factors or mediate pathogenesis mechanisms? It seems like the lack of adaptation would go both ways - supported by the empirical data gathered in this manuscript. There is also the issue of only relatively simple and accessible mirror-image antibiotics being available. This is a limitation that - to their credit - the authors do discuss in the discussion section.

Reviewer #1 (Public review):

Summary:

The authors provide in vivo and in vitro evidence for an interaction between AIRE and AID. This has implications for the dynamics of the germinal center response and autoimmunity related to the APSI disease.

The manuscript describes an unexpected function of AIRE, which is more well known for its function to regulate negative selection of T cells in the thymus. Here, the gene has also been shown to be expressed by B cells (Immunity 2015: 26070482). They describe that AIRE interacts with AID, and in its absence, B cells acquire more hypermutations and also produce auto-antibodies against IL-17. These autoantibodies have been described previously.

Strengths:

The study is interesting and provides some additional information about how AIRE regulates immune cell function. Several biochemical and in vivo experiments show the interaction and the function of AIREs in the regulation of AID activity in the GC response.

Weaknesses:

Some of the hypothetical consequences of this regulation are not investigated. This includes responses to model antigens and dynamics of the germinal center related to kinetics.

Major Comments:

(1) AID regulates both switch and somatic hypermutation. Switch is easier to achieve, so which of these processes does AIRE influence the most? Also, the switch is thought to occur before the B cell enters the GC. Looking at the histology, is AIRE also expressed at the early proliferative stage that has been described by Ann Haberman?

(2) In experiments determining anti-CD40-dependent upregulation of AIRE, naïve resting B cells were used from mice. A proportion of the B-cells got activated. Are these MZB or FOB cells as MZBs are more easily activated?

(3) In the BM chimeric experiments in Figure 3. Do the AIRE+ and AIRE - populations distribute equally among B cell subpopulations?

(4) Furthermore, in the NP-KLH experiments, one would expect that B cells with increased affinity would leave the GC earlier and become plasma cells. Thus, the kinetics of the AIRE+ vs AIRE- B cells within the GC would be different? Also, would they maybe take over at some point, as the increased affinity would favor help from Tfh cells that are known to be limited?

(5) Given the previous studies on AIRE's function in regulating transcription (PMID: 34518235), how does this interaction fit into this picture?

(6) In the uracil experiments, the readout for AID to induce double-stranded breaks could be tested.

(7) The candida experiments are a nice connection to the situation in patients. However, why is it mostly auto-antibodies against IL-17? How about other immune responses, as well as T cell-independent type I and II responses?

Reviewer #2 (Public review):

Summary:

In this study, Zhou et al investigated the expression and function of AIRE in B cells in peripheral lymphoid tissues. First, they found the expression of AIRE protein in mature B cells in the follicles in human tonsils and spleens from healthy donors. Flow cytometry analyses using human samples as well as Aire-reporter mice demonstrated AIRE expression in germinal center B cells. The expression of Aire in B cells was induced by CD40 signals. Then, to investigate the impact of AIRE deficiency on B-cell function, the authors used a method of transplanting bone marrow cells from Aire-KO and WT mice into B-cell-deficient mice, comparing B-cell development and function reconstituted in the recipient mice. Their results showed that Aire-deficient B cells strongly responded to immunization with antigens, exhibiting enhanced class switching and somatic hypermutation of antibodies compared with WT B cells. The same phenomena were observed in CRISPRed B cell lines lacking Aire. The authors successfully utilized the Aire-deficient B cell line to demonstrate that Aire suppresses antibody class switching and somatic hypermutation via its interaction with AID. Finally, using B cell transfer into B cell-deficient mice demonstrated that mice harboring Aire-deficient B cells produced high levels of autoantibodies against Th17 cytokines and exhibited reduced resistance to Candida infection. This mirrors characteristic symptoms in AIRE-deficient patients. The findings of this study not only reveal an unexpected function of AIRE in B cells but also have the potential to contribute to understanding the pathogenesis of APECED and to offering a new direction for developing therapies.

Strengths:

The strength of this study lies in demonstrating the expression of the function of AIRE in B cells in both mice and humans. It also revealed the direct interaction between AIRE and AID, along with its binding mode (requiring CARD and NLS domains of AIRE), and showed that this interaction is crucial for AIRE function in B cells. It is also significant that the study demonstrated how B-cell-intrinsic dysfunction of AIRE leads to autoantibody production against cytokines.

Weaknesses:

As for loss-of-function analysis of Aire in B cells, in addition to the B cell transfer from Aire-KO mice performed in this study, generating B cell-specific Aire-deficient mice using Aire-flox mice (Dobes et al, Eur J Immunol 2018) would further reinforce the conclusions of this study. Furthermore, the relationship with Aire function in thymic B cells reported by previous studies remains unclear, posing an unresolved challenge. This study also failed to address whether Aire deficiency affects gene expression in GC B cells, in particular, whether it induces the expression of various self-antigens as reported in thymic B cells or mTECs.

Reviewer #1 (Public review):

Summary:

Knowing that small pupil-size variations accompany brightness variations (even when these are illusory), the authors asked whether pupil constrictions would accompany the synesthetic perception of a brighter color (compared with a darker one), induced by the presentation of a black-white character. This grapheme-colour synesthesia is only experienced by a few participants, sixteen of whom were enrolled in this study. The results reliably showed that a relative pupil constriction would "betray" the perception of a brighter color in these participants, while no such effect would be observed in control participants who were asked to report a color in association with each grapheme, even though they did not perceive any.

Strengths:

The main strength of the study lies in its combination of psychophysics (brightness ratings) and pupillometry, which allowed for showing clear-cut results.

Weaknesses:

Some relatively minor weaknesses concern the ancillary analyses, which tackle secondary questions and are not entirely convincing.

(1) The linear mixed model approach is a powerful way to identify important variables, but it does not clarify whether the key factors are between-subject or between-trial variations. Some variables are inherently defined at a subject level (e.g., PA scores), others are not. I would strongly recommend an alternative visualisation of the results to examine inter-individual variability.

(2) It is not clear why taking the first derivative of pupil size in Figure 5 would isolate the effect of arousal, eliminating those of luminance and contrast changes (in fact, one could argue for the opposite, since arousal effects are generally constant for extended periods of time while contrast effects are typically more local and transient).

(3) It is a pity that responses to physical brightness modulations were only measured in the synesthete group, not in controls, as this would have allowed for ruling out differences in pupil reactivity across the two populations.

(4) Another concern is with the visualisation of the pupil traces in Figure 3 (main results); these were heavily pre-processed (per-participant demeaned), losing any feature besides the effect of interest and generating the unrealistic expectation that perception of dark/bright colors generate a net dilation/constriction of the pupil - whereas perception-related modulations of pupil size are always relative and generally small compared to the numerous other effects registered in pupil size. It would be far better to see the actual profiles, preserving the unfolding of dilations and constrictions over time, especially since these are further analysed in Figures 4 and 5.

Impact:

Despite these weaknesses, and especially if they are adequately addressed in the review, this work is likely to improve our understanding of synesthesia, providing a new tool to quantify the subjective sensations; an interesting potential extension would be using pupillometry for tracking changes over time of the synesthetic experiences, opening up the possibility to evaluate the importance of learning for this peculiar experience.

Reviewer #2 (Public review):

Synesthesia is a neurological condition where stimulation of one sensory channel leads to involuntary, automatic, and consistent experience of another, unrelated percept. For example, Sir Francis Galton (1880, Nature) famously described the robust tendency of some individuals (synesthetes) to associate numerals with a distinct color. Ever since, synesthesia has continued to attract a broad interest in the cognitive neurosciences in light of its implications for the study of domains such as perception, consciousness, and brain connectivity, among others.

Strauch, Leenaars, and Rouw measured pupil size in a group of 16 grapheme-color synesthetes and two matched control groups. The participants were presented with gray digits - that is, visual stimuli having identical physical properties in terms of brightness. Each participant subsequently rated the corresponding evoked color and brightness: unlike controls, synesthetes did so in a very consistent and reliable fashion. Accordingly, this was also shown in their pupils: despite the same objective luminance, digits associated with brighter percepts caused their pupils to constrict, and digits associated with darker percepts caused their pupils to dilate more than controls. These results highlight how crossmodal correspondences are deeply rooted in synesthetes, and put forward pupillometry as a particularly appealing biomarker for some phenomenological experience (at least those grounded in "brightness").

Further strengths of the technique are its temporal resolution and its responsiveness to several constructs. Across several tasks, the authors show, for example, that responses to synesthetic light are somewhat slower than responses to real light (i.e., they are likely mediated), but at the same time faster than responses to mental imagery. The role of mental imagery can also be reasonably dismissed when considering the second feature of pupil size: its responsiveness to mental effort and cognitive load. The pupils tend to dilate with demanding, challenging tasks, and this was the case when control participants were asked to report the color of a digit for which they did not consistently experience a synesthetic association. The same task was, instead, seemingly effortless for synesthetes, again speaking in favor of the automaticity of number-color correspondences in their case.

Overall, the findings by Strauch, Leenaars, and Rouw are highly significant for the field and likely to be impactful. The strength of their evidence, when accounting for the relatively small sample size and the inherent variability of both phenomenology (color perception and subjective reporting) and physiology (pupil size), is adequate and sufficiently convincing.

Reviewer #3 (Public review):

Summary:

In the present study, the authors examined pupillary responses to uncolored stimuli (number graphemes) among number-color synesthetes and non-synesthetes. After seeing a digit, the synesthetes and active control participants were asked to indicate which color they perceived using three dimensions of hue, saturation, and lightness. The lightness values were the primary independent variable for follow-up analyses. To see how the pupil responded to psychologically "bright" and "dark" digits, the authors split the reported lightness values at the median and plotted them. The synesthetes showed a pupillary constriction to digits they perceived as bright and dilation to digits they perceived as dark. Active control participants did not show that effect. In a subsequent block, only the synesthetes were shown the colors they reported perceiving as colored discs. Their pupillary responses were similar. The authors also found that the differences in pupillary responses between light and dark perceptions (with digits) were only slightly delayed in their onset to the perception of a colored disc, and therefore, the color perception accompanying a digit is unlikely to be effortful or a retrieved association, but occurs rather automatically.

Strengths:

The authors employed a well-controlled and designed quasi-experiment comparing color-grapheme synesthetes to non-synesthetes and showed convincingly that the color perceptions accompanying graphemes alter the physical perception of brightness. They also made a reasoned attempt to rule out the possibility that color associations are occurring effortfully via retrieved associations.

Weaknesses:

There are some areas in which the implications of these findings could be elaborated upon. I had the following questions:

(1) Are the pupillary responses among synesthetes, which objectively do not seem to match the degree of physical stimulation entering the retina, in any way maladaptive for eye functioning? I understand the constriction/dilation of the pupil to not only benefit visual acuity but also to protect the retina from damage. Are synesthetes at any risk of retinal damage due to over-dilation of the pupil to brighter stimuli? Or are these effects of a magnitude that is too small to matter? As reported in arbitrary units, it was hard to know how large these effects were in terms of measurable changes in dilation (e.g., millimeters).

(2) Likewise, is the automatic synesthetic merging of two percepts something that could be learned such that natural synesthetes and "artificial" synesthetes would look similar? For example, if a group of non-synesthetic participants were to learn a color-grapheme association to automaticity, would you expect their pupillary responses to the graphemes look similar to the synesthetes'? If so (or if not), what would this tell us anything about the phenomenology of synesthesia?

(3) Do the synesthetic perceptions of digit graphemes merge in a sensible way? For example, if a synesthete sees a particular color with the digit 1, and a different color with the digit 9, what do they perceive when they see 19? or 1-9, or 1 9? Is there color blending, or an altogether different color perception?

Reviewer #1 (Public review):

Summary:

In this work, Huang et al. revealed the complex regulatory functions and transcription network of 172 unknown transcriptional factors (TFs) in Pseudomonas aeruginosa PAO1. They have built a global TF-DNA binding landscape and elucidated binding preferences and functional roles of these TFs. More specifically, the authors established a hierarchical regulatory network and identified ternary regulatory motifs, and co-association modules. Since P. aeruginosa is a well known pathogen, the authors thus identified key TFs associated with virulence pathways (e.g., quorum sensing [QS], motility, biofilm formation), which could be potential drug targets for future development. The authors also explored the TF conservation and functional evolution through pan-genome and phylogenetic analyses. For the easy searching by other researchers, the authors developed a publicly accessible database (PATF_Net) integrating ChIP-seq and HT-SELEX data.

Strengths:

(1) The authors performed ChIP-seq analysis of 172 TFs (nearly half of the 373 predicted TFs in P. aeruginosa) and identified 81,009 significant binding peaks, representing one of the largest TF-DNA interaction studies in the field. Also, The integration of HT-SELEX, pan-genome, and phylogenetic analyses provided multi-dimensional insights into TF conservation and function.

(2) The authors provided informative analytical Framework for presenting the TFs, where a hierarchical network model based on the "hierarchy index (h)" classified TFs into top, middle, and bottom levels. They identified 13 ternary regulatory motifs and co-association clusters, which deepened our understanding of complex regulatory interactions.

(3) The PATF_Net database provides TF-target network visualization and data-sharing capabilities, offering practical utility for researchers especially for the P. aeruginosa field.

Weaknesses:

(1) There is very limited experimental validation for this study. Although 24 virulence-related master regulators (e.g., PA0815 regulating motility, biofilm, and QS) were identified, functional validation (e.g., gene knockout or phenotypic assays) is lacking, leaving some conclusions reliant on bioinformatic predictions. Another approach for validation is checking the mutations of these TFs from clinical strains of P. aeruginosa, where chronically adapted isolates often gain mutations in virulence regulators.

(2) ChIP-seq in bacteria may suffer from low-abundance TF signals and off-target effects. The functional implications of non-promoter binding peaks (e.g., coding regions) were not discussed.

(3) PATF_Net currently supports basic queries but lacks advanced tools (e.g., dynamic network modeling or cross-species comparisons). User experience and accessibility remain under-evaluated. But this could be improved in the future.

Achievement of Aims and Support for Conclusions

(1) The authors successfully mapped global P. aeruginosa TF binding sites, constructed hierarchical networks and co-association modules, and identified virulence-related TFs, fulfilling the primary objectives. The database and pan-genome analysis provide foundational resources for future studies.

(2) The hierarchical model aligns with known virulence mechanisms (e.g., LasR and ExsA at the bottom level directly regulating virulence genes). Co-association findings (e.g., PA2417 and PA2718 co-regulating pqsH) resonate with prior studies, though experimental confirmation of synergy is needed.

Impact on the Field and Utility of Data/Methods

(1) This study fills critical gaps in TF functional annotation in P. aeruginosa, offering new insights into pathogenicity mechanisms (e.g., antibiotic resistance, host adaptation). The hierarchical and co-association frameworks are transferable to other pathogens, advancing comparative studies of bacterial regulatory networks.

(2) PATF_Net enables rapid exploration of TF-target interactions, accelerating candidate regulator discovery.

Comments on revisions:

The authors have done a good job of revising their manuscript. The manuscript is now more concise and logical for readers.

Reviewer #3 (Public review):

Summary:

The authors utilized ChIP-seq on strains containing tagged transcription factor (TF)-overexpression plasmids to identify binding sites for 172 transcription factors in P. aeruginosa. High-quality binding site data provides a rich resource for understanding regulation in this critical pathogen. These TFs were selected to fill gaps in prior studies measuring TF binding sites in P. aeruginosa. The authors further perform a structured analysis of the resulting transcriptional regulatory network, focusing on regulators of virulence and metabolism, in addition to performing a pangenomic analysis of the TFs. The resulting dataset has been made available through an online database. While the implemented approach to determining functional TF binding sites has limitations, the resulting dataset still has substantial value to P. aeruginosa research.

Strengths:

The generated TF binding site database fills an important gap in regulatory data in the key pathogen P. aeruginosa. Key analyses of this dataset presented include an analysis of TF interactions and regulators of virulence and metabolism, which should provide important context for future studies into these processes. Experimental validation has been included in the revised version. The online database containing this data is well organized and easy to access. As a data resource, this work should be of significant value to the infectious disease community.

Weaknesses:

Drawbacks of the study, which have been mitigated in a revised version, include 1) challenges interpreting binding site data obtained from TF overexpression due to unknown activity state of the TFs on the measured conditions (discussed by the authors), and 2) remaining challenges in the practical utilization of the TRN topological analysis.

Reviewer #1 (Public review):

This study presents a useful finding about development of task representations in mouse medial prefrontal cortex using 1-photon calcium recordings in an olfactory-guided spatial memory task. A key strength of the study is the use of longitudinal recordings allowing identification of task-related activity that emerges after learning. The study also reports existence of neuronal sequences during learning and their replay at reward locations. The evidence provided is solid, providing quantification of functional classes of cells over the course of learning using the longitudinal calcium recordings in prefrontal cortex, and quantification of prefrontal sequences.

(1) The authors continue to state that task phase selective cells (non-splitter) cells can be considered as "cross-condition generalization" and interpret them as "potential building blocks of schemas". However, cross-condition generalization requires demonstration of cross-condition generalization performance (CCGP) of neural decoders across task conditions, which is not shown here.

(2) The authors note that correlations on short time scales are not similar between sampling and reward phase, acknowledging that these two represent different behavioral states in a cued-memory task, and that the manuscript should more clearly distinguish replay with "pure sequences". However, while the last line in the abstract states that "sub-second neural sequences in the mPFC are more likely involved in behavioral outcomes rather than planning future actions", references are made throughout the manuscript to preplay/replay sequences, including results primarily for non-cued spatial memory tasks, in which there is no cued sampling phase. For example, lines 259-263 state "During odor sampling phase, no such significant replay was observed..." and "... sequence clusters showed small but significant bias to preplay in the sampling phase". If the authors want to distinguish between replay and "pure" sequences, then the terminology "replay" and "preplay" should not be used here.

Further, large parts of the Discussion are devoted to comparison to hippocampal ripple-associated replay. Lines 355-356 in Discussion state that "the suggestion that mPFC sequences may also support planning [Tang et al., 2021] could not be confirmed by our work as sequences in the odor sampling phase were absent". It should be clarified that this is a comparison between what the authors term "pure sequences" in the sampling phase of an odor-cued task, and internally generated sequences during hippocampal ripples in a non-cued spatial memory task, so this is not a like-for-like comparison.

Reviewer #1 (Public review):

Summary:

Jocher, Janssen et al examine the robustness of comparative functional genomics studies in primates that make use of induced pluripotent stem cell-derived cells. Comparative studies in primates, especially amongst the great apes, are generally hindered by the very limited availability of samples, and iPSCs, which can be maintained in the laboratory indefinitely and defined into other cell types, have emerged as promising model systems because they allow the generation of data from tissues and cells that would otherwise would be unobservable.

Undirected differentation of iPSCs into many cell types at once, using a method known as embryoid body differentiation, requires researchers to manually assign all cell types in the dataset so they can be correctly analysed. Typically, this is done using marker genes associated with a specific cell type. These are defined a priori, and have historically tended to be characterised in mice and human and then employed to annotate other species. Jocher, Janssen et al ask if the marker genes and features used to define a given cell type in one species are suitable for use in a second species, and then quantify the degree of usefulness of these markers. They find that genes that are informative and cell type specific in a given species are less valuable for cell type identification in other species, and that this value, or transferability, drops off as the evolutionary distance between species increases.

This paper will help guide future comparative studies of gene expression in primates (and more broadly) as well as add to the growing literature on the broader challenges of selecting powerful and reliable marker genes for use in single cell transcriptomics.

Strengths:

Marker gene selection and cell type annotation is challenging problem in scRNA studies, and successful classification of cells often requires manual expert input. This can be hard to reproduce across studies, as despite general agreement on the identity of many cell types, different methods for identifying marker genes will return different sets of genes. The rise of comparative functional genomics complicates this even further, as a robust marker gene in one species need not always be as useful in a different taxon. The finding that so many marker genes have poor transferability is striking, and by interrogating the assumption of transferability in a thorough and systematic fashion, this paper reminds us of the importance of systematically validating analytical choices. The focus on identifying how transferability varies across different types of marker genes (especially when comparing TFs to lncRNAs), and on exploring different methods to identify marker genes, also suggests additional criteria by which future researchers could select robust marker genes in their own data.

The paper is built on a substantial amount of clearly reported and thoroughly considered data, including EBs and cells from four different primate species - humans, orangutans, and two macaque species. The authors go to great lengths to ensure the EBs are as comparable as possible across species, and take similar care with their computational analyses, always erring on the side of drawing conservative conclusions that are robustly supported by their data over more tenuously supported ones that could be impacted by data processing artefacts such as differences in mappability etc. For example, I like the approach of using liftoff to robustly identify genes in non-human species that can be mapped to and compared across species confidently, rather than relying on the likely incomplete annotation of the non-human primate genomes. The authors also provide an interactive data visualisation website that allows users to explore the dataset in depth, examine expression patterns of their own favourite marker genes and perform the same kinds of analyses on their own data if desired, facilitating consistency between comparative primate studies.

Weaknesses and recommendations:

(1) Embryoid body generation is known to be highly variable from one replicate to the next for both technical and biological reasons, and the authors do their best to account for this, both by their testing of different ways of generating EBs, and by including multiple technical replicates/clones per species. However, there is still some variability that could be worth exploring in more depth. For example, the orangutan seems to have differentiated preferentially towards cardiac mesoderm whereas the other species seemed to prefer ectoderm fates, as shown in Figure 2C. Likewise, Supplementary Figure 2C suggests significant unbalance in the contributions across replicates within a species, which is not surprising given the nature of EBs, while Supplementary Figure 6 suggests that despite including three different clones from a single rhesus macaque, most of the data came from a single clone. The manuscript would be strengthened by a more thorough exploration of the intra-species patterns of variability, especially for the taxa with multiple biological replicates, and how they impact the number of cell types detected across taxa etc.

The same holds for the temporal aspect of the data, which is not really discussed in depth despite being a strength of the design. Instead, days 8 and 16 are analysed jointly, without much attention being paid to the possible differences between them. Are EBs at day 16 more variable between species than at day 8? Is day 8 too soon to do these kinds of analyses? Are markers for earlier developmental progenitors better/more transferable than those for more derived cell types?

(2) Closely tied to the point above, by necessity the authors collapse their data into seven fairly coarse cell types, and then examine the performance of canonical marker genes (as well as those discovered de novo) across the species. But some of the clusters they use are somewhat broad, and so it is worth asking whether the lack of specificity exhibited by some marker genes and driving their conclusions is driven by inter-species heterogeneity within a given cluster.

Comments on revisions:

I think the authors have addressed my previous comments to my satisfaction, and I thank them for the changes they have made, it's good to see that the manuscript is just as sound as it seemed the first time around.

Reviewer #2 (Public review):

Summary:

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

Strengths:

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

Revision comment:

The revised manuscript is improved in clarity and methodological detail. An important addition is the retrograde labeling data showing a degree of anatomical segregation between CN->CL and CN->VAL pathways that strengthens their reported different functional roles. I still think that potential effects on motor execution when cerebellar nuclei are silenced during task performance may complicate interpretations specifically related to learning. However, the evidence supporting a role of the cerebellar nuclei in off-line consolidation is convincing.

Overall, the study outlines a multifaceted role of the cerebellum in motor learning, consolidation, and execution. The demonstration that cerebellar projections to distinct forebrain structures contribute to these processes is significant.

Reviewer #3 (Public review):

Summary:

Varani et al present important findings regarding the role of distinct cerebellothalamic connections in motor learning and performance. Their key findings are that: 1) cerebellothalamic connections are important for learning motor skills, 2) cerebellar efferents specifically to the central lateral (CL) thalamus are important for short-term learning, 3) cerebellar efferents specifically to the ventral anterior lateral (VAL) complex are important for offline consolidation of learned skills, and 4) that once a skill is acquired, cerebellothalamic connections become important for online task performance. The authors went to great lengths to separate effects on motor performance from learning, for the most part successfully. While one could argue about some of the specifics, there is little doubt that the CN-CL and CN-VAL pathways play distinct roles in motor learning and performance. An important next step will be to dissect the downstream mechanisms by which these cerebellothalamic pathways mediate motor learning and adaptation.

Strengths:

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

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

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

(4) The retrograde tracing experiments (Supplementary Figure 5) demonstrate convincingly that the CN-VAL and CN-CL projections are almost entirely segregated,

Weaknesses:

(1) Despite the care the authors took to demonstrate that their chemogenetic approach does not impair online performance, there is (as they acknowledge in the Discussion) impaired rotarod performance at fixed higher speeds in Supplementary Figure 4f for CN-VAL projections, suggesting that there could be subtle changes in motor performance below the level of detection of their assays. There is also a trend in the same direction that did not pass significance for CN-CL at higher speeds, suggesting that part of the effects could be related to subtle deficits in performance.

Reviewer #1 (Public review):

Summary:

Fecal virome transfer (FVT) has the potential to take advantage of microbiome associated phages to treat diseases such as NEC. However, FVT is also associated with toxicity due to the presence of eukaryotic viruses in the mixture, which are difficult to filter out. The authors use a chemostat propagation system to reduce the presence of eukaryotic viruses (these become lost over time during culture). They show in pig models of NEC that chemostat propagation reduce the incidence of diarrhea induced by FVTs.

Strengths:

The authors report an innovative yet simple approach that has the potential to be useful for future applications. Most of the experiments are easy to follow and are performed well.

Weaknesses:

The biggest weakness is that the authors show that their technique addresses safety, but they are unable to demonstrate that they retain efficacy in their NEC model. This could be due to technical issues or perhaps the efficacy of FVT reported in the literature is not robust.

During the revision, the authors have acknowledged these limitations and added clarifications where necessary.

Reviewer #2 (Public review):

The authors hypothesized that chemostat propagated viromes could modulate the GM and reduce NEC lesions while avoiding potential side effects, such as the earlier onset of diarrhea. This is interesting.

Major revision

(1) As authors said, the aim of the research is 'We hypothesized that chemostat propagated viromes could modulate the GM and reduce NEC lesions while avoiding potentialside effects, such as earlier onset of diarrhea'.

(a) For the efficacy, in Fig 5, there are no significance in stomach pathology and enterocolitis between groups, even between the control group and the experimental groups, is it because of the low incidence of NEC? This may affect the statistical power of the conclusions. And how can you draw the conclusion that chemostat can reduce NEC lesions?

(b) Lack of gross view pictures of animal tissues or any other pathological pictures is not convincing.

(c) For the safety, such as body weight development, FVT had no statistical significance with control, CVT and CVT-MO, so how can you draw the conclusion that chemostat can avoiding potentialside effects?

(d) The evidence to prove the decrease of eukaryotic viruses are not enough and quantitative.

(2) Fig 3F,

(a) How can a medium have 'the baseline viral content' ?

(b) Statistical significance of relative abundance of specific eukaryotic viral contigs between different times is unkown.

(c) Some of listed eukaryotic viruses, their hosts are not pigs, piglets or even human, so what's the meaning if these eukaryotic viruses decreased?

(3) In this study, pH 6.5 was selected as the pH value for chemostat cultivation, but considering the different adaptability of different bacteria to pH, it is recommended to further explore the effect of pH on bacteria and virus groups. In particular, it was optimized to maintain the growth of beneficial bacteria such as Lactobacillaceae and Bacteroides in order to improve the effect of chemostat cultivation.

(4) In some charts, the annotation of error lines, statistical significance markers (even 'ns' should be marked), etc., should be more standardized and clearer. And in your results section, the combination of pictures is messy, thus maybe you should do some recombination.

Comments on revisions:

(1) At the design level, the study posited "reduction of necrotizing enterocolitis (NEC)" as the primary hypothesis and endpoint. Yet neither of the two in-vivo experiments demonstrated any NEC-protective signal; Experiment 2 even showed a trend toward more severe gastric lesions. Although delayed onset of diarrhea can be listed as a secondary endpoint, its clinical significance is limited. The work remains a safety proof-of-concept and falls short of efficacy validation, yielding insufficient scientific value for publication.

(2) The manuscript postulates a link between the loss of Lactobacillaceae phages and the absence of NEC protection, but no reverse verification (e.g., re-introducing these phages or optimizing culture to retain them) was performed within the study.

(3) Culturing intestinal microbiota ex vivo is inherently challenging, owing to oxygen sensitivity, pH drift, nutrient depletion, and other factors. This study not only failed to demonstrate stable congruence between the cultured community and the original fecal inoculum, but also documented a marked loss of Lactobacillaceae and a 75 % drop in viral diversity. In the absence of any NEC-protective efficacy, the authors likewise provide no functional validation of phage viability (lysis assays, MOI determination, etc.). Consequently, the data are inadequate to support expectations of therapeutic benefit in vivo.

Reviewer #3 (Public review):

This study investigated the in vitro amplification of donor fecal virus using chemostat culturing technology, aiming to reduce eukaryotic virus load while preserving bacteriophage community diversity, thereby optimizing the safety and efficacy of FVT. The research employed a preterm pig model to evaluate the effects of chemostat-propagated viromes (CVT) in preventing necrotizing enterocolitis (NEC) and mitigating adverse effects such as diarrhea.

Strengths:

(1) Enhanced Safety Profile:<br /> Chemostat cultivation effectively reduced eukaryotic virus load, thereby minimizing the potential infection risks associated with virome transplantation and offering a safer virome preparation method for clinical applications.

(2) Process Reproducibility:<br /> The chemostat system achieved stable amplification of bacteriophage communities (Bray-Curtis similarity >70%), mitigating the impact of donor fecal variability on therapeutic efficacy.

Comments on revision:

The authors have satisfactorily addressed all comments and concerns raised during the review process. The revised manuscript is clear, complete, and meets the standards of the journal.

Reviewer #1 (Public review):

Summary:

The paper investigates the interplay between fluid flow and biofilm development using Pseudomonas aeruginosa PAO1 in microfluidic channels. By combining experimental observations with mathematical modeling, the study identifies the significant impact of nutrient limitation and hydrodynamic forces on biofilm growth and detachment. The authors demonstrate that nutrient limitation drives the longitudinal distribution of biomass, while flow-induced detachment influences the maximum clogging and temporal dynamics. The study highlights that pressure buildup plays a critical role in biofilm detachment, leading to cyclic episodes of sloughing and regrowth. A stochastic model is used to describe the detachment process, capturing the apparent randomness of sloughing events. The findings offer insights into biofilm behavior during clogging and fouling, potentially relevant to infections, environmental processes, and engineering applications.

Strengths:

This paper demonstrates a strong integration of experimental work and mathematical modeling, providing a comprehensive understanding of biofilm dynamics in a straight microfluidic channel. The simplicity of the microchannel geometry allows for accurate modeling, and the findings have the potential to be applied to more complex geometries. The detailed analysis of nutrient limitation and its impact on biofilm growth offers valuable insights into the conditions that drive biofilm formation. The model effectively describes biofilm development across different stages, capturing both initial growth and cyclic detachment processes. While cyclic pressure buildup has been studied previously, the incorporation of a stochastic model to describe detachment events is a novel and significant contribution, capturing the complexity and randomness of biofilm behavior. Finally, the investigation of pressure buildup and its role in cyclic detachment and regrowth enhances our understanding of the mechanical forces at play, making the findings applicable to a wide range of technological and clinical contexts.

Weaknesses:

The study achieves its primary objective of combining experiments and modeling to elucidate the coupling between flow, biofilm growth, and detachment in a confined microfluidic channel. In the revised manuscript, the authors have clarified several methodological choices and underlying assumptions. The points below are best viewed not as weaknesses, but as aspects that define the scope of the approach.

• Biofilm porosity and permeability. The authors now discuss biofilm porosity and provide a clear rationale for neglecting permeability effects in their system, arguing that flow around dense biofilm structures dominates over flow through the matrix. While this assumption appears reasonable for the conditions explored, permeability effects are not explicitly modeled and could become relevant in less compact or more heterogeneous biofilms.

• Characterization of the EPS matrix. The role of the extracellular matrix is convincingly addressed using polysaccharide‑deficient mutants, which provides a strong and causal link between EPS composition and mechanical stability. At the same time, the absence of complementary biochemical or imaging‑based characterization means that spatial or temporal variations in EPS distribution are not directly resolved, limiting the level of structural details.

• Three‑dimensional interpretation of biofilm development. The authors clarify that three‑dimensional information is primarily obtained from pressure‑based measurements, with two‑dimensional imaging serving as a validation tool. This approach is coherent and supported by scaling arguments and reproducibility across experiments.

Reviewer #1 (Public review):

Summary:

Matsen et al. describe an approach for training an antibody language model that explicitly tries to remove effects of "neutral mutation" from the language model training task, e.g. learning the codon table, which they claim results in biased functional predictions. They do so by modeling empirical sequence-derived likelihoods through a combination of a "mutation" model and a "selection" model; the mutation model is a non-neural Thrifty model previously developed by the authors, and the selection model is a small Transformer that is trained via gradient descent. The sequence likelihoods themselves are obtained from analyzing parent-child relationships in natural SHM datasets. The authors validate their method on several standard benchmark datasets and demonstrate its favorable computational cost. They discuss how deep learning models explicitly designed to capture selection and not mutation, trained on parent-child pairs, could potentially apply to other domains such as viral evolution or protein evolution at large.

Overall, we think the idea behind this manuscript is really clever and shows promising empirical results. Two aspects of the study are conceptually interesting: the first is factorizing the training likelihood objective to learn properties that are not explained by simple neutral mutation rules, and the second is training not on self-supervised sequence statistics but on the differences between sequences along an antibody evolutionary trajectory. If this approach generalizes to other domains of life, it could offer a new paradigm for training sequence-to-fitness models that is less biased by phylogeny or other aspects of the underlying mutation process.

Future versions of the work can consider extending the ideas to additional datasets, species, definitions of fitness, or even different proteins entirely.

Comments on revisions:

We thank the authors for addressing our points and have no remaining questions.

Reviewer #2 (Public review):

Summary:

Endowing protein language models with an ability to predict the function of antibodies would open a world of translational possibilities. However, antibody language models have yet to achieve the breakthrough success, which large language models have achieved for the understanding and generation of natural language. This paper elegantly demonstrates how training objectives imported from natural language applications lead antibody language models astray on function prediction tasks. Training models to predict masked amino acids teaches models to exploit biases of nucleotide-level mutational processes, rather than protein biophysics. Taking the underlying biology of antibody diversification and selection seriously allows disentangling these processes, through what the authors call deep amino acid selection models. These models extend previous work by the authors (Matsen MBE 2025) by providing predictions not only for the selection strength at individual sites, but also for individual amino acids substitutions. This represents a practically important advance.

Strengths:

The paper is based on a deep conceptual insight, the existence of multitude of biological processes that affect antibody maturation trajectories. The figures and writing a very clear, which should help make the broader field aware of this important but sometimes overlooked insight. The paper adds to a growing literature proposing biology-informed tweaks for training protein language models, and should thus be of interest to a wide readership interested in the application of machine learning to protein sequence understanding and design.

Weaknesses:

Proponents of the state-of-the-art protein language models might counter the claims of the paper by appealing to the ability of fine-tuning to deconvolve selection and mutation-related signatures in their high-dimensional representation spaces. Leaving the exercise of assessing this claim entirely to future work somewhat diminishes the heft of the (otherwise good!) argument. In the context of predicting antibody binding affinity, the modeling strategy only allows prediction of mutations that improve affinity on average but not those which improve binding to specific epitopes.

Comments on revisions:

We thank the authors for clarifying the description of the methods and for adding additional discussion of important directions for future work.

Reviewer #3 (Public review):

Summary:

This work proposes DASM, a new transformer-based approach to learning the distribution of antibody sequences which outperforms current foundational models at the task of predicting mutation propensities under selected phenotypes, such as protein expression levels and target binding affinity. The key ingredient is the disentanglement, by construction, of selection-induced mutational effects and biases intrinsic to the somatic hypermutation process (which are embedded in a pre-trained model).

Strengths:

The approach is benchmarked on a variety of available datasets and for two different phenotypes (expression and binding affinity). The biologically informed logic for model construction implemented is compelling and the advantage, in terms of mutational effects prediction as well as computational efficiency, is clearly demonstrated via comparisons to state-of-the-art models.

Weaknesses:

While all the main points are well addressed and supported, it could have been interesting to strengthen the claim of gain in interpretability by investigating it explicitly in relation to the functional effects studied in this paper.

Comments on revisions:

I thank the authors for clarifying a few points I had flagged up and I appreciate much better that the content of the companion paper was precisely covering model selection and structural interpretability.

Regarding my first point (references for language models for antibodies), I feel that the parenthetical citation format shouldn't be a problem (but the editors might advise here). Antiberta2 is this paper: https://www.biorxiv.org/content/10.1101/2023.12.12.569610v1.full.pdf (yet, I understand if the authors want to focus on models purely sequence-based). A couple of additional references could be: https://academic.oup.com/bioinformatics/article/40/11/btae659/7888884; https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1012646; https://www.pnas.org/doi/10.1073/pnas.2418918121; https://arxiv.org/abs/2506.13006.

A very minor comment: could one add some p-value (it could be a supplementary table) for the Pearson correlation coefficients? The comparison between methods is rather clear, but for some correlations it's a bit unclear whether they should be considered significant. It would be important to understand the extent to which in different datasets one might expect functional prediction power based on an evolutionary objective function alone.

Reviewer #1 (Public review):

Summary:

The paper by ILBAY et al describes a screen in C. elegans for loss-of-function of factors that are presumed to constitutively downregulate heat shock or stress genes regulated by HSF-1. The hypothesis posits an active mechanism of downregulation of these genes under non-stressed conditions. The screen robustly identified ZNF-236, a multi zinc finger containing protein, whose loss upregulates heat-shock and stress-induced prion-like protein genes, but which does not appear to act in cis at the relevant promoters. The authors speculate that ZNF-236 acts indirectly on chromatin or chromatin domains to repress hs genes under non-stressed conditions.

Strengths:

The screen is clever, well-controlled and quite straightforward. I am convinced that ZNF-236 has something to do with keeping heat shock and other stress transcripts low. The mapping of potential binding sites of ZNF-236 is negative, despite the development of a new method to monitor binding sites. I am not sure whether this assay has a detection/sensitivity threshold limit, as it is not widely used. Up to this point, the data are solid, and the logic is easy to follow.

Weaknesses:

While the primary observations are well-documented, the mode of action of ZNF-236 is inadequately explored. Multi Zn finger proteins often bind RNA (TFIII3A is a classic example), and the following paper addresses multivalent functions of Zn finger proteins in RNA stability and processing: Mol Cell 2024 Oct 3;84(19):3826-3842.e8. doi: 10.1016/j.molcel.2024.08.010.). I see no evidence that would point to a role for ZNF-236 in nuclear organization, yet this is the authors' favorite hypothesis. In my opinion, this proposed mechanism is poorly justified, and certainly should not be posited without first testing whether ZNF-236 acts post-transcriptionally, directly down-regulating the relevant mRNAs in some way. It could regulate RNA stability, splicing, export or translation of the relevant RNAs rather than their transcription rates. This can be tested by monitoring whether ZNF-236 alters run-on transcription rates or not. If nascent RNA synthesis rates are not altered, but rather co- and/or post-transcriptional events, and if ZNF-236 is shown to bind RNA (which is likely), the paper could still postulate that the protein plays a role in downregulating stress and heat shock proteins. However, they could rule out that it acts on the promoter by altering RNA Pol II engagement. Another option that should be tested is that ZNF-236 acts by nucleating an H3K9me domain that might shift the affected genes to the nuclear envelope, sequestering them in a zone of low-level transcription. That is also easily tested by tracking the position of an affected gene in the presence and absence of SNF-236. This latter mechanism is also right in line with known modes of action for Zn finger proteins (in mammals, acting through KAP1 and SETDB1). A role for nucleating H3K9me could be easily tested in worms by screening MET-2 or SET-25 knockouts for heat shock or stress mRNA levels. These data sets are already published.

Without testing these two obvious pathways of action (through RNA or through H3K9me deposition), this paper is too preliminary.

Appraisal:

The authors achieved their initial aim with the screen, and the paper is of interest to the field. However, they do not adequately address the likely modes of action. Indeed, I think their results fail to support the conclusion or speculation that ZNF-236 acts on long-range chromatin organization. No solid evidence is presented to support this claim.

Impact:

If the paper were to address and/or rule out likely modes of action, the paper would be of major value to the field of heat shock and stress mRNA control.

Reviewer #2 (Public review):

Summary:

This manuscript reports the identification of ZNF-236 as a key regulator that maintains quiescence of heat shock inducible genes in C. elegans. Using a forward genetic screen for constitutive activation of an endogenous hsp-16.41 reporter, the authors show that loss of znf-236 leads to widespread, HSF-1-dependent expression of inducible heat shock proteins (iHSPs) and a subset of prion-like stress-responsive genes, in the absence of proteotoxic stress. Transcriptomic analysis reveals that znf-236 mutants partially overlap with the canonical heat shock response, selectively activating highly inducible iHSPs rather than the full HSR program. iHSP transgenes integrated throughout the genome generally become de-repressed in znf-236 mutants, whereas the same constructs on extrachromosomal arrays or inserted into the rDNA locus re insensitive to znf-236 loss. Using a newly developed method, Transcription Factor Deaminase Sequencing (TFD-seq), the authors show that ZNF-236 binds sparsely across the genome and does not associate with iHSP promoters, supporting an indirect mode of regulation. Physiologically, znf-236 mutants exhibit increased thermotolerance and maintain iHSP expression during aging.

Strengths:

This is a carefully executed and internally consistent study that identifies a new regulator of stress-induced gene quiescence in C. elegans. The genetics are clean and the phenotypes are robust.

Weaknesses:

The manuscript is largely descriptive. It would be substantially strengthened by deeper mechanistic insight into what ZNF-236 does beyond being required for default silencing.

Reviewer #3 (Public review):

Summary:

The researchers performed a genetic screen to identify a protein, ZNF-236, which belongs to the zinc finger family, and is required for repression of heat shock inducible genes. The researchers applied a new method to map the binding sites of ZNF-236, and based on the data, suggested that the protein does not repress genes by directly binding to their regulatory regions targeted by HSF1. Insertion of a reporter in multiple genomic regions indicates that repression is not needed in repetitive genomic contexts. Together, this work identifies ZNF-236, a protein that is important to repress heat-shock-responsive genes in the absence of heat shock.

Strengths:

A hit from a productive genetic screen was validated, and followed up by a series of well-designed experiments to characterize how the repression occurs. The evidence that the identified protein is required for the repression of heat shock response genes is strong.

Weaknesses:

The researchers propose and discuss one model of repression based on protein binding data, which depends on a new technique and data that are not fully characterized.

Major Comments:

(1) The phrase "results from a shift in genome organization" in the abstract lacks strong evidence. This interpretation heavily relies on the protein binding technique, using ELT-2 as a positive and an imperfect negative control. If we assume that the binding is a red herring, the interpretation would require some other indirect regulation mechanism. Is it possible that ZNF-236 binds to the RNA of a protein that is required to limit HSF-1 and potentially other transcription factors' activation function? In the extrachromosomal array/rDNA context, perhaps other repressive mechanisms are redundant, and thus active repression by ZNF-236 is not required. This possibility is mentioned in one sentence in the discussion, but most of the other interpretations rely on the ZNF-236 binding data to be correct. Given that there is other evidence for a transcriptional role for ZNF-236, and no negative control (e.g. deletion of the zinc fingers, or a control akin to those done for ChIP-seq (like a null mutant or knockdown), a stronger foundation is needed for the presented model for genome organization.

(2) Continuing along the same line, the study assumes that ZNF-236 function is transcriptional. Is it possible to tag a protein and look at localization? If it is in the nucleus, it could be additional evidence that this is true.

(3) I suggest that the authors analyze the genomic data further. A MEME analysis for ZNF-236 can be done to test if the motif occurrences are enriched at the binding sites. Binding site locations in the genome with respect to genes (exon, intron, promoter, enhancer?) can be analyzed and compared to existing data, such as ATAC-seq. The authors also propose that this protein could be similar to CTCF. There are numerous high-quality and high-resolution Hi-C data in C. elegans larvae, and so the authors can readily compare their binding peak locations to the insulation scores to test their hypothesis.

(4) The researchers suggest that ZNF-236 is important for some genomic context. Based on the transcriptomic data, can they find a clue for what that context may be? Are the ZNF-236 repressed genes enriched for not expressed genes in regions surrounded by highly expressed genes?

Reviewer #1 (Public review):

In this revised manuscript, Qin and colleagues aim to delineate a neural mechanism that is engaged specifically in the sated flies to suppress the intake of sugar solution (the "brake" mechanism for sugar consumption). They identified a three-step neuropeptidergic system that downregulates the sensitivity of sweet-sensing gustatory sensory neurons in sated flies. First, neurons that release a neuropeptide Hugin (which is an insect homolog of vertebrate Neuromedin U (NMU)) are in active state when the concentration of glucose is high. This activation depends on the cell-autonomous function of Hugin-releasing neurons that sense hemolymph glucose levels directly. Next, the Hugin neuropeptides activate Allatostatin A (AstA)-releasing neurons via one of Hugin receptors, PK2-R1. Finally, the released AstA neuropeptide suppresses sugar response in sugar-sensing Gr5a-expressing gustatory sensory neurons through AstA-R1 receptor. Suppression of sugar response in Gr5a-expressing neurons reduces fly's sugar intake motivation. They also found that NMU-expressing neurons in the ventromedial hypothalamus (VMH) of mice (which project to the rostal nucleus of the solitary tract (rNST)) are also activated by high concentrations of glucose independent of synaptic transmission, and that injection of NMU reduces the glucose-induced activity in the downstream of NMU-expressing neurons in rNST. These data suggest that the function of Hugin neuropeptide in the fly is analogous to the function of NMU in the mouse.

The shift of the narrative, which focuses specifically on the hugin-AstA axis as the "brake" on the satiety signal and feeding behavior, clarified the central message of the presented work. The authors have provided multiple lines of compelling evidence generated through rigorous experiments. The parallel study in mice adds a unique comparative perspective that makes the paper interesting to a wide range of readers.

While I deeply appreciate the authors' efforts to substantially restructure the manuscript, I have a few suggestions for further improvements. First, there remains room for discussion whether the "brake" function of the hugin-AstA axis is truly satiety state-dependent. The fact that neural activation (Fig. Supp. 8), peptide injection (Fig. 3A, 4A), receptor knockdown (Fig. 3C,G, 4E), and receptor mutants (Fig. Supp. 10, 12) all robustly modulate PER irrespective of the feeding status suggests that the hugin-AstA axis influences feeding behaviors both in sated and hungry flies. Additionally, their new data (Fig. Supp. 13B, C) now shows that synaptic transmission from hugin-releasing neurons is necessary for completely suppressing feeding even in sated flies. If the hugin-AstA axis engages specifically in sated (high glucose) state, disruption of this neuromodulatory system is expected to have relatively little effect in starved flies (in which the "brake" is already disengaged).

In this context, it is intriguing that the knockdown of PK2-R2 hugin receptor modestly but consistently decreases proboscis extension reflex specifically in starved flies (Fig. 3D, H). The manuscript does not discuss this interesting phenotype at all. Given the heterogeneity of hugin-releasing neurons (Fig. Supp. 7), there remains a possibility that a subset of hugin-releasing neurons and/or downstream neurons can provide a complementary (or even opposing) effect on the feeding behavior.

Given these intriguing yet unresolved issues, it is important to acknowledge that whether this system is "selectively engaged in fed states to dampen sweet sensation (in Discussion)" requires further functional investigations. Consistent effects of manipulation of the hugin-AstA system across multiple experimental approaches underscores the importance of this molecular circuitry axis for controlling feeding behaviors. Moderation of conclusions to accommodate alternative interpretation of data will be beneficial for field to determine the precise mechanism that controls feeding behaviors in future studies.

Reviewer #2 (Public review):

Summary:

The question of how caloric and taste information interact and consolidate remains both active and highly relevant to human health and cognition. The authors of this work sought to understand how nutrient sensing of glucose modulates sweet sensation. They found that glucose intake activates hugin signaling to AstA neurons to suppress feeding, which contributes to our mechanistic understanding of nutrient sensation. They did this by leveraging the genetic tools of Drosophila to carry out nuanced experimental manipulations, and confirmed the conservation of their main mechanism in a mammalian model. This work builds on previous studies examining sugar taste and caloric sensing, enhancing the resolution of our understanding.

Strengths:

Fully discovering neural circuits that connect body state with perception remains central to understanding homeostasis and behavior. This study expands our understanding of sugar sensing, providing mechanistic evidence for a hugin/AstA circuit that is responsive to sugar intake and suppresses feeding. In addition to effectively leveraging the genetic tools of Drosophila, this study further extends their findings into a mammalian model with the discovery that NMU neural signaling is also responsive to sugar intake.

Weaknesses:

The effect of Glut1 knockdown on PER in hugin neurons is modest in both fed and starved flies, suggesting that glucose intake through Glut1 may only be part of the mechanism. Additionally, many of the manipulations testing the "brake" circuitry throughout the study show similar effects in both fed and starved flies. This suggests that the focus of the discussion and Supplemental Figure 16 on a satiety-specific "brake" mechanism may not be fully supported by the data.

Reviewer #1 (Public review):

Summary:

The paper uses rigorous methods to determine phase dynamics from human cortical stereotactic EEGs. It finds that the power of the phase is higher at the lowest spatial phase. The application to data illustrates the solidity of the method and their potential for discovery.

Comments on revisions:

The authors have provided responses to the previous recommendations. The paper does not seem to contain further significant improvements. I am thus not inclined to change my judgement.

Reviewer #3 (Public review):

Summary:

The authors propose a method for estimating the spatial power spectrum of cortical activity from irregularly sampled data and apply it to iEEG data from human patients during a delayed free recall task. The main findings are that the spatial spectra of cortical activity peak at low spatial frequencies and decrease with increasing spatial frequency. This is observed over a broad range of temporal frequencies (2-100 Hz).

Strengths:

A strength of the study is the type of data that is used. As pointed out by the authors, spatial spectra of cortical activity are difficult to estimate from non-invasive measurements (EEG and MEG) and from commonly used intracranial measurements (i.e. electrocorticography or Utah arrays) due to their limited spatial extent. In contrast, iEEG measurements are easier to interpret than EEG/MEG measurements and typically have larger spatial coverage than Utah arrays. However, iEEG is irregularly sampled within the three-dimensional brain volume and this poses a methodological problem that the proposed method aims to address.

Weaknesses:

Although the proposed method is evaluated in several indirect ways, a direct evaluation is lacking. This would entail simulating cortical current source density (CSD) with known spatial spectrum and using a realistic iEEG volume-conductor model to generate iEEG signals.

Comments on revisions:

I would like to clarify two points:

(1) In their response, the authors frame the role of simulations primarily as a means of assessing the effects of volume conduction. However, the purpose of evaluating a proposed estimation method through simulations extends beyond this specific issue. More generally, simulations are essential for establishing that the proposed method-particularly given the multiple non-trivial transformations applied to the observed data-produces accurate and reliable estimates under controlled conditions.

(2) The authors seem to interpret my use of the term current source density as referring to the current source density (CSD) method, which is an approach to mitigating volume conduction by inverting Poisson's equation. This was not my intention: current source density refers to the physical quantity (i.e., the spatial density of current sources) underlying macroscopic brain activity, and is independent of any specific estimation or inversion technique.

Reviewer #1 (Public review):

The authors present an approach that uses the transformer architecture to model epistasis in deep mutational scanning datasets. This is an original and very interesting idea. Applying the approach to 10 datasets they quantify the contribution of higher order epistasis, showing it varies quite extensively.

Comments on revisions:

The authors have addressed my concerns.

Reviewer #2 (Public review):

Summary:

This paper presents a novel transformer-based neural network model, termed the epistatic transformer, designed to isolate and quantify higher-order epistasis in protein sequence-function relationships. By modifying the multi-head attention architecture, the authors claim they can precisely control the order of specific epistatic interactions captured by the model. The approach is applied to both simulated data and ten diverse experimental deep mutational scanning (DMS) datasets, including full-length proteins. The authors argue that higher-order epistasis, although often modest in global contribution, plays critical roles in extrapolation and capturing distant genotypic effects, especially in multi-peak fitness landscapes.

Strengths:

(1) The study tackles a long-standing question in molecular evolution and protein engineering: "how significant are epistatic interactions beyond pairwise effects?" The question is relevant given the growing availability of large-scale DMS datasets and increasing reliance on machine learning in protein design.

(2) The manuscript includes both simulation and real-data experiments, as well as extrapolation tasks (e.g., predicting distant genotypes, cross-ortholog transfer). These well-rounded evaluations demonstrate robustness and applicability.

(3) The code is made available for reproducibility.

Weaknesses:

(1) The paper mainly compares its transformer models to additive models and occasionally to linear pairwise interaction models. However, other strong baselines exist. For example, the authors should compare baseline methods such as "DANGO: Predicting higher-order genetic interactions". There are many works related to pairwise interaction detection, such as: "Detecting statistical interactions from neural network weights", "shapiq: Shapley interactions for machine learning", and "Error-controlled non-additive interaction discovery in machine learning models".

(2) While the transformer architecture is cleverly adapted, the claim that it allows for "explicit control" and "interpretability" over interaction order may be overstated. Although the 2^M scaling with MHA layers is shown empirically, the actual biological interactions captured by the attention mechanism remain opaque. A deeper analysis of learned attention maps or embedding similarities (e.g., visualizations, site-specific interaction clusters) could substantiate claims about interpretability.

(3) The distinction between nonspecific (global) and specific epistasis is central to the modeling framework, yet it remains conceptually underdeveloped. While a sigmoid function is used to model global effects, it's unclear to what extent this functional form suffices. The authors should justify this choice more rigorously or at least acknowledge its limitations and potential implications.

(4) The manuscript refers to "pairwise", "3-4-way", and ">4-way" interactions without always clearly defining the boundaries of these groupings or how exactly the order is inferred from transformer layer depth. This can be confusing to readers unfamiliar with the architecture or with statistical definitions of interaction order. The authors should clarify terminology consistently. Including a visual mapping or table linking a number of layers to the maximum modeled interaction order could be helpful.

Comments for the revision:

I want to thank the authors for their efforts in revising the manuscript. Most of the concerns raised in the initial review have been adequately addressed.

However, one important issue remains. I previously asked the authors to benchmark their method against stronger baselines. The authors declined, arguing that these alternatives are "not directly applicable to the types of analyses." I am not persuaded by this rationale. In my view, these baseline methods target essentially the same underlying problem, and at least some, if not all, should be included in a comparative evaluation (or the manuscript should provide a clearer, more technically grounded explanation of why such comparisons are not feasible or not meaningful).

Reviewer #3 (Public review):

Summary:

Sethi and Zou present a new neural network to study the importance of epistatic interactions in pairs and groups of amino acids to the function of proteins. Their new model is validated on a small simulated data set, and then applied to 10 empirical data sets. Results show that epistatic interactions in groups of amino acids can be important to predict the phenotype of a protein, especially for sequences that are not very similar to the training data.

Strengths:

The manuscript relies on a novel neural network architecture that makes it easy to study specifically the contribution of interactions between 2, 3, 4 or more amino acids. The novel network architecture achieves such a level of interpretability without noticeable performance penalty. The study of 10 different protein families shows that there is variation among protein families in the importance of these interactions, and that higher order interactions are particularly important to predict the phenotypes of distant proteins.

Weaknesses:

The Github repository provides a README file to run a standard pipeline, but a user will need to go through the code to actually know what that pipeline is doing.

Reviewer #1 (Public review):

The manuscript by Luciano et al is a collection of experiments about the yeast histone 3 lysine 4 methyltransferase, Set1, starting with 10 yeast two-hybrid screens (Y2H). Y2H screens were briefly popular 20+ years ago, but the persistently unfavourable false-to-true positive ratios limited their utility, and the conclusion emerged that Y2H is an unreliable approach for gathering protein-protein interaction data. Y2H outcomes are candidate interaction lists at best, strongly contaminated by false positives. Here, the authors employed a company (Hybridomics) to perform the Y2H screens.

The primary data is not presented, and the outcomes are summarized using the Hybridomics in-house quality scoring system in Figure 1A. It is not possible to evaluate these data, and the manuscript presents cartoon summaries that the reader must accept as valuable.

(1) Based on the extensive knowledge about Set1C/COMPASS acquired from genetics and biochemistry by many labs (including the Geli lab), the results presented here from the 10 Y2H screens are notably patchy. Of the 7 subunits of this complex, only one (Spp1) was identified using Set1 as bait. Conversely, as baits, Swd2, Spp1, Shg1, captured Set1, and the Bre2-Sdc1 interaction was reciprocally identified. These interactions were scored at the highest confidence level, which lends some confidence to the screens. However, the missing interactions, even at the third confidence level, indicate that any Y2H conclusions using these data must be qualified with caution. The authors do not appear to be cautious in their lengthy evaluations of these candidate interactions, which are illustrated with cartoons in Figures 2 and 3, with some support from the literature but almost without additional evidence. Snf2 is a particularly interesting candidate, which the authors support with pull-down experiments after mixing the two proteins in vitro (Figure 4). After Y2H, this is the least convincing evidence for a protein-protein interaction, and no further, more reliable evidence is supplied.

(2) Figure 5 continues the cartoon summary of extrapolations from the Y2H screens, again without supporting evidence, except that the authors state, "We have refined the interaction region between Set1, Prp8 and Prp22, showing that Prp8 and Prp22 interact strongly with Set1-F4 (n-SET). Prp22 interacts in addition with Set1-F1 (Figure S2)." However, Figure S2 does not show this evidence and is incoherent.

The figure legends for Figure S2B and C (copied here in bold) do not correspond to the figure.

B - Expression of the F1-F5 fragments in yeast cells. Fusion proteins were detected with an anti-GAL4 monoclonal antibody. TOTO yeast cells (Hybrigenics) were transformed with the different pB66-Set1-F1 to F5 plasmids and subsequently with either P6, pP6-Snf2 762-968, pP6-Prp8 37-250, or pP6-Prp22 379-763 that were identified in the Y2H screens. Transformed cells were incubated 3 days at 30{degree sign}C on SD-LEU-TRP and then restreaked on SD-LEU-TRP-HIS with 3AT. Cell growth was monitored after 2 days at 30{degree sign}C.

C - Solid and dotted arrows indicate that transformed TOTO cells transformed with pB66-Set1-F1 to F5 and the indicated prey (Snf2, Prp8, and Prp22) are growing in the presence of 20 mM and 5 mM of AT, respectively.

Figure S2D is two almost featureless dark grey panels accompanied by the figure legend D) Control experiment showing that TOTO cells transformed with p6 and pB66-Set1-F4 are not gowing (sic) in the presence of 5 mM or 20 mM AT.

Line 343. Interestingly, the two-hybrid screens reveal that Set1 1-754 interacted with Gag capsid-like proteins of Ty1 (Figure S5), raising the possibility that Set1 binding to Ty1 mRNA is linked to the interaction of Set1 1-754 with Gag.

This is another example of the primary mistake repeatedly made by the authors -Y2H interactions are candidate results and not conclusive evidence. To further illustrate this point, the authors highlight the candidate interaction between Nis1 and 3 Set1C subunits.

(3) After multiple speculations based on the Y2H candidates, the authors changed to focus on sumoylation of Set1, which has previously reported to be sumoylated. Evidence identifying two sumoylation sites in Set1, in the N-SET and SET domains, is valuable and adds important progress to the role of sumoylation in the regulation of H3K4 methyltransferase, relevant for all eukaryotes. This illuminating part of the manuscript is only tenuously connected to the preceding Y2H screens and concomitant speculations.

(4) The manuscript then describes a red herring exercise involving Set1 methylation of Nrm1. In an already speculative and difficult manuscript, it is exasperating to read a paragraph about a failed idea. Apart from panel E, Figure 7 is a distraction, and I believe it should not be shared.

(5) However, despite the failure with Nrm1, Line 443 - The H3K4-like domain in Nrm1 raised our attention to other yeast proteins that carry such sequences. This line of thinking is even less connected to the Y2H screens than the sumoylation work.

However, the authors present a reasonable evaluation of the yeast proteome screened for six amino acids similar to the known H3K4 motif ARTKQT (Figure 7e).

(6) However, this evaluation goes nowhere and has no connection with the next section of the manuscript, which is entirely speculation about the regulation of metabolism and stress responses based on the Y2H results and selected evidence from the literature.

(7) The manuscript then describes more failed experiments regarding lysine methylation of Snf2 by Set1C, which unexpectedly reports arginine methylation rather than lysine. The manuscript does not currently meet the standard expected for this type of paper - the composition is somewhat incoherent and there are no previous reports of arginine methylation by SET domain proteins.

The manuscript presents a very experienced grasp of the literature and a sophisticated appreciation of the forefront issues, but a surprising failure to eliminate uninformative failures and peripheral distractions. The overinterpretation of Y2H results is a dominating failure. There are some valuable parts within this manuscript, and hopefully, the authors can reformat to eliminate the defects and appropriately qualify the candidate data.

Reviewer #2 (Public review):

Summary:

This paper starts with a large-scale yeast two-hybrid (Y2H) screen using Set1 (full-length and smaller parts) and other Set1C/COMPASS subunits as bait. There are hundreds of possible interactions identified, but only a small number are given any follow-up. While it's useful to document all the possible interactions, the unfocused and preliminary nature of the results makes the paper feel scattered and incomplete.

Strengths:

The Y2H screen was very comprehensive, producing lots of interesting possible leads for further experiments.

Weaknesses:

The results are useful but incomplete because only a small subset of the Y2H interactions is further examined. Even in the case of those that were further tested, the validating experiments are only partial or inconclusive.

Reviewer #3 (Public review):

The SET1C/COMPASS complex is the histone H3K4 methyltransferase in Saccharomyces cerevisiae, where it plays pivotal roles in transcriptional regulation, DNA repair, and chromatin dynamics. While its canonical function in histone methylation is well-established, its full interactome remains poorly defined. Moreover, whether SET1C methylates non-histone substrates has been an open question.

In this study, Luciano et al. employ systematic yeast two-hybrid (Y2H) screening to uncover novel interactors and functions of SET1C. Their findings reveal potential functional connections to RNA biogenesis, chromatin remodeling, and non-histone methylation.

The authors performed multiple Y2H screens using Set1 (full-length, N-terminal, and C-terminal fragments) and each of its seven subunits as baits. They identified high-confidence interactors that link SET1C to diverse cellular processes, including chromatin regulation (e.g., the SWI/SNF complex via Snf2), DNA replication (e.g., Mcm2, Orc6), RNA biogenesis (e.g., spliceosome components Prp8 and Prp22; polyadenylation factors Pta1 and Ref2), tRNA processing (e.g., Trm1, Trm732), and nuclear import/export (e.g., importins Kap104 and Kap123). Some of these interactions were further validated by immunoprecipitation or in vitro assays.

Given the interaction of Set1 with Slx5 and Wss1 - proteins involved in SUMO-dependent processes - the authors investigated and convincingly demonstrated that Set1 is sumoylated. This modification may influence the function and regulation of the SET1C complex.

Finally, the authors provide evidence that SET1C methylates proteins beyond histone H3K4, notably Nrm1, a transcriptional corepressor, and Snf2, the catalytic subunit of the SWI/SNF chromatin remodeling complex. Although Nrm1 contains a domain resembling the H3K4-methylated sequence (H3K4-like domain), this region does not appear to be required for its methylation. The search for other proteins containing similar domains as potential methylation candidates (p.12, first paragraph) seems less justified, given the lack of evidence supporting the requirement for the H3K4-like domain in methylation.

This study offers valuable insights into the interactome of SET1C, suggesting potential links between the complex and a wide range of cellular processes. However, the functional implications of the Y2H interactions remain to be explored further. Additionally, the study provides intriguing information on the possible regulation of Set1 by sumoylation. The discovery of Nrm1 and Snf2 as methylation substrates could significantly expand the known targets and functions of SET1C.

The results are supported by high-quality data.

Reviewer #1 (Public review):

Summary:

This work stratifies depression subgroups based on white matter integrity (Fractional Anisotropy, FA) and evaluates the relationship between white matter (WM) alterations in these subgroups and clinical symptoms. Furthermore, the authors tested these subgroup findings in an independent cohort. This paper provides WM-based depression subtypes that are linked to the clinical symptom profile (anxiety, cognitive, hopelessness, sleep, and psychomotor retardation) and presents the prediction of treatment outcome using these subtypes.

Strengths:

Applying a novel NMF (Non-negative Matrix Factorization) biclustering approach to stratify depression subtypes using white matter integrity. Following the recent functional MRI-based depression subtype stratification, this work provides a structural signature for depression heterogeneity. These subtypes were also tested in an independent cohort, with findings regarding clinical symptom profiles.

Weaknesses:

Although this novel method successfully subgroups depression patients, it is difficult to understand the spatial patterns of WM alteration and which structural connections, such as DMN, SN, ECN, and Limbic, because the findings are distributed across multiple WM bundles in each subgroup. Furthermore, these subtypes fail to predict optimal treatment selection within each group, since all subgroups benefit from different treatments.

Reviewer #2 (Public review):

Summary:

The authors measure the directional consistency of water diffusion in white matter (functional anisotropy: FA) to stratify depression subtypes across young adults. These findings are significant in that they highlight white matter as an underappreciated aspect of neural heterogeneity in major depressive disorder. While the evidence for meaningful, lower-dimensional structure in depression heterogeneity within their Nanjing cohorts is strong, claims that their subtypes are characterized by specific clinical symptom profiles and reflect neuroplasticity reserve are not supported by the same strength of evidence.

Strengths:

Circumscribing analyses to a simple white matter measure, across a sparse skeleton, with explicit sparsity-promoting algorithms yielded heterogeneity subdivisions that are much more interpretable than most depression heterogeneity clustering papers. Replication of their 3-cluster solution in an external dataset bolsters confidence in the existence of these 3 clusters, although generalizability to more diverse populations remains untested. The authors also tested a wide variety of treatment outcomes, which is difficult data to aggregate but ultimately critical for validating the utility of depression subtypes.

Weaknesses:

sCCA and SVR results were less interpretable. In part, this is due to core features of these methods (broad distribution of weights, instability across iterations). However, these inherent components of sCCA and SVR opacity were exacerbated by the opacity surrounding several analytic choices made by the authors and intermediate results associated with them. Without more transparency, it's unclear how these results extend the neuroclinical differentiation established (or not established) by their original NMF analyses.

To be more specific, a central claim of the paper is that their biotypes are "pathophysiologically distinct" and demonstrate "symptom-specific neurobiological substrates". However, only 3/18 pairwise symptom differences generalize across both datasets (Figures 1 and 2), implying that these biotypes have more symptom overlap than distinction. Brain-based distinctions are real and replicable, but because their NMF approach specifically optimizes for separating clusters on the basis of brain features, this is more of a methodological validation than a scientific finding. While several brain-symptom relationships reported later using sCCA and SVR are interesting, it is not currently possible to evaluate the robustness of these relationships and whether or not these relationships are nested within NMF-derived clusters or exist regardless of subtype.

To be clear, the heterogeneity problem in depression is extremely difficult to solve and beyond the scope of this manuscript. Despite the scale of this problem, the authors do report tangible progress in this aim, largely through finding an interpretable set of white matter features distinguishing patient clusters. These findings may lead researchers to meaningfully incorporate white matter features into heterogeneity analyses more in the future. However, many of the claims made are not fully supported, particularly surrounding clinical specificity and neuroplasticity reserve.

Reviewer #1 (Public review):

Summary:

The paper reports an analysis of whole-genome sequence data from 40 Faroese. The authors investigate aspects of demographic history and natural selection in this population. The key findings are that Faroese (as expected) have a small population size and are broadly of Northwest European ancestry. Accordingly, selection signatures are largely shared with other Northwest European populations although the authors identify signals that may be specific to the Faroes. Finally they identify a few predicted deleterious coding variants that may be enriched in the Faroes.

Strengths:

The data are appropriately quality controlled and appear to be high quality. Some aspects of Faroese population history are characterized - in particular, the relatively (compared to other European populations) high proportion of long runs of homozygosity, which may be relevant for disease mapping of recessive variants. The selection analysis is presented reasonably, although as the authors point out, many aspects, for example differences in iHS, can reflect differences in demographic history or population-specific drift and thus can't reliably be interpreted in terms of differences in the strength of selection.

Weaknesses:

The main limitations of the paper are as follows:

(1) The data are not available. I appreciate that (even de-identified) genotype data cannot be shared, however, that does substantially reduce the value of the paper. I appreciate the authors sharing summary statistics for the selection scan.

(2) The insight into the population history of the Faroes is limited, relative to what is already known (i.e. they were settled around 1200 years ago, by people with a mixture of Scandinavian and British ancestry, have a small effective population size, and any admixture since then comes from substantially similar populations). It's obvious, for example that the Faroese population has a smaller bottleneck than, say, GBR.

More sophisticated analyses (for example, ARG-based methods, or IBD or rare variant sharing) would be able to reveal more detailed and fine-scale information about the history of the populations that is not already known. PCA, ADMIXTURE and HaplotNet analysis are broad summaries, but the interesting questions here would be more specific to the Faroes, for example, What are the proportions of Scandinavian vs Celtic ancestry? What is the date and extent of sex bias (as suggested by the uniparental data) in this admixture? I think that it a bit of a missed opportunity not to address these questions.

(3) I don't really understand the rationale for looking at HLA-B allele frequencies. The authors write that "Observational evidence from the FarGen project recruitment data suggest that ankylosing spondylitis (AS) may be at a higher prevalence in the Faroe Islands". But nothing beyond that. So there's no evidence (certainly no published evidence) that AS is more prevalent, and hence nothing to explain with the HLA allele frequencies? This section seems preliminary.

Reviewer #2 (Public review):

In this paper, Hamid et al present 40 genomes from the Faroe Islands. They use these data (a pilot study for an anticipated larger-scale sequencing effort) to discuss the population genetic diversity and history of the sample, and the Faroes population. I think this is an overall solid paper; it is overall well-polished and well-written. It is somewhat descriptive (as might be expected for an explorative pilot study), but does make good use of the data.

The data processing and annotation follows a state-of-the-art protocol, and at least I could not find any evidence in the results that would pinpoint towards bioinformatic issues having substantially biased some of the results, and at least preliminary results lead to the identification of some candidate disease alleles, showing that small, isolated cohorts can be an efficient way to find populations with locally common, but globally rare disease alleles.

I also enjoyed the population structure analysis in the context of ancient samples, which gives some context to the genetic ancestry of Faroese, although it would have been nice if that could have been quantified, and it is unfortunate that the sampling scheme effectively precludes within-Faroes analyses.

Comments on the revision:

I appreciate the authors' detailed and thoughtful response to my review. They have addressed all my concerns to my satisfaction and I have no additional comments.

Reviewer #2 (Public review):

The major strengths of the manuscript are in the Plasmodium falciparum genetic and phenotyping approaches. PfMSP2 knockouts are made in two different strains, which is important as it is know that invasion pathways can vary between strains, but is a level of comprehensiveness that is not always delivered in P. falciparum genetic studies. The knockout strains are characterised very thoroughly using multiple different assays and the authors should be commended for publishing a good deal of negative data, where no phenotype was detected. This is not always done but is very helpful for the field and reduces the potential for experimental redundancy, i.e. others repeating work that has already been performed but never published. The quality of the writing, referencing and figures is also generally strong.

There are certainly some areas of the manuscript that would benefit from deeper exploration, such as electron microscopy/other imaging approaches to explore whether deletion of PfMSP2 has a visible impact on merozoite surface structure, further replicates of the video microscopy assays to see whether trends in the data could reach significance (although these are very time-consuming and technically difficult assays), and follow up of some of the genes where expression is changed by PfMSP2 knockout (as the authors point out, there are no candidates that have a very obvious link to invasion suggesting that they may be compensating for PfMSP2 function, although several are expressed in schizont stages). However, there is already a substantial amount of data in the manuscript, and more detailed follow-up is reasonable to leave to future work. Overall, with the modifications made through the review process, including the addition of new controls for key experiments, the claims and conclusions are justified by the data, and the manuscript generates important new information about a highly studied Plasmodium falciparum merozoite surface protein.

Reviewer #3 (Public review):

Henshall et al. study invasion of human erythrocytes by Plasmodium falciparum merozoites and report knockout of PfMSP2, a critical merozoite surface protein with unknown function. They describe conservation of MSP2 in P. falciparum and key avian malaria parasites, unabated growth of two knockout lines (∆MSP2) produced in divergent 3D7 and Dd2 strains, no differences in expression of key invasion-associated genes, no effect on invasion kinetics (with or without protease treatment of erythrocytes), nonsignificant effects of knockout on parasite growth inhibition by antibodies directed against key invasion-associated antigens, and do find a significant effect on potentiating AMA1 invasion inhibitory antibodies. The studies are interesting and have potential for directing vaccine design targeting erythrocyte invasion, a critical step in bloodstream expansion of malaria parasites.

Major points:

(1) Much of the manuscript describes negative results and this reviewer found it arduous to get through many negative or nonsignificant results before finally getting to the significant effect on AMA1 inhibitory antibodies, not presented until Figure 6! Computational studies in Fig. 1 could be a supplementary figure. Figs. 2 and 3. demonstrate knockout in 3D7 and Dd2, respectively and could be assembled into a single figure. (Notably Fig. 2A and 3A are almost identical with use of some different primers.) Fig. 2E, 2F, 3D-H, all of Fig. 4, most of Fig. 5 are all negative or insignificant results that could also be moved to supplementary data. As MSP4, MSP5, and SUB1 are presumably included in the whole genome RNA-seq experiments shown in Fig. 4C, it makes sense to remove Fig. 4A data from the paper fully. These consolidating changes would help highlight the key finding of improved binding and block of AMA1's role in invasion.

(2) The potentiating effects on anti-AMA1 antibodies are shown with rabbit sera and purified antibodies, mouse monoclonal antibodies, and smaller i-bodies inspired by shark antibody-like receptors but not with human monoclonal antibodies (hmAbs). As naturally acquired hmAbs targeting AMA1 have been identified and characterized (PMIDs: 39632799, 40020675), would it not be important to test these antibodies in the ∆MSP2, especially as the authors emphasize the importance of their model in designing better human malaria vaccines?

(3) Fig. 7 presents quantitative fluorescence microscopy to measure anti-AMA1 binding and support a model where MSP2 serves to sterically hinder antibody access to AMA1 on individual merozoites. I understand that the negative WD33 control is useful to contrast to the positive WD34 antibody (both bind AMA1 but only WD34 exhibits parasite growth inhibitory effects), but it seems that use of smaller i-bodies rather than conventional larger mouse or ideally human monoclonal antibodies may compromise demonstration of steric hindrance by MSP2 because smaller i-bodies may be less hinder.

(4) Some explanation for why WD33 fails to inhibit growth despite targeting the same antigen as WD34 is needed. Are the epitopes known? Does one bind further from the RON2 binding pocket?

Reviewer #1 (Public review):

Summary:

In this manuscript, the authors describe the generation of a Drosophila model of RVCL-S by disrupting the fly TREX1 ortholog cg3165 and by expressing human TREX1 transgenes (WT and the RVCL-S-associated V235Gfs variant). They evaluate organismal phenotypes using OCT-based cardiac imaging, climbing assays, and lifespan analysis. The authors show that loss of cg3165 compromises heart performance and locomotion, and that expression of human TREX1 partially rescues these phenotypes. They further report modest differences between WT and mutant hTREX1 under overexpression conditions. The study aims to establish Drosophila as an in vivo model for RVCL-S biology and future therapeutic testing.

Strengths:

(1) The manuscript addresses an understudied monogenic vascular disease where animal models are scarce.

(2) The use of OCT imaging to quantify fly cardiac performance is technically strong and may be useful for broader applications.

(3) The authors generated both cg3165 null mutants and humanized transgenes at a defined genomic landing site.

(4) The study provided initial in vivo evidence that human TREX1 truncation variants can induce functional impairments in flies.

Weaknesses:

(1) Limited mechanistic insight.

RVCL-S pathogenesis is strongly linked to mislocalization of truncated TREX1, DNA damage accumulation, and endothelial/podocyte cellular senescence. The current manuscript does not examine any cellular, molecular, or mechanistic readouts - e.g. DNA damage markers, TREX1 subcellular localization in fly tissues, oxidative stress, apoptosis, or senescence-related pathways. As a result, the model remains largely phenotypic and descriptive.

To strengthen the impact, the authors should provide at least one mechanistic assay demonstrating that the humanized TREX1 variants induce expected molecular consequences in vivo.

(2) The distinction between WT and RVCL-S TREX1 variants is modest.

In the cg3165 rescue experiments, the authors do not observe differences between hTREX1 and the V235Gfs variant (e.g., Figure 3A-B). Phenotypic differences only emerge under ubiquitous overexpression, raising two issues:

(i) It is unclear whether these differences reflect disease-relevant biology or artifacts of strong Act5C-driven expression.

(ii) The authors conclude that the model captures RVCL-S pathogenicity, yet the data do not robustly separate WT from mutant TREX1 under physiological expression levels.

The authors should clarify these limitations and consider additional data or explanations to support the claim that the model distinguishes WT vs RVCL-S variants.

(3) Heart phenotypes are presented as vascular defects without sufficient justification.

RVCL-S is a small-vessel vasculopathy, but the Drosophila heart is a contractile tube without an endothelial lining. The authors refer to "vascular integrity restoration," but the Drosophila heart lacks vasculature.

The manuscript would benefit from careful wording and from a discussion of how the fly heart phenotypes relate to RVCL-S microvascular pathology.

(4) General absence of tissue-level or cellular imaging.

No images of fly hearts, brains, eyes, or other tissues are shown. TREX1 nuclear mislocalization is a hallmark of RVCL-S, yet no localization studies are included in this manuscript.

Adding one or two imaging experiments demonstrating TREX1 localization or tissue pathology would greatly enhance confidence in the model.

Reviewer #2 (Public review):

Summary:

The authors used the Drosophila heart tube to model Retinal vasculopathy with the goal of building a model that could be used to identify druggable targets and for testing chemical compounds that might target the disease. They generated flies expressing human TREX1 as well as a line expressing the V235G mutation that causes a C-terminal truncation that has been linked to the disease. In humans, this mutation is dominant. Heart tube function was monitored using OCM; the most robust change upon overexpression of wild-type or mutant TREX1was heart tube restriction, and this effect was similar for both forms of TREX1. Lifespan and climbing assays did show differential effects between wt and mutant forms when they were strongly and ubiquitously expressed by an actin-Gal4 driver. Unfortunately, these types of assays are less useful as drug screening tools. Their conclusion that the primary effect of TREX is on neuronal function is inferential and not directly supported by the data.

Strengths:

The authors do not show that CG3165 is normally expressed in the heart. Further fly heart tube function was similarly restricted in response to expression of either wild-type or mutant TREX1. The fact that expression of any form of human TREX1 had deleterious effects on heart function suggests that TREX1 serves different roles in flies compared to humans. Thus, in the case of this gene, it may not be a useful model to use to identify targets or use it as a drug screening tool.

The significant effects on lifespan and climbing that did show differential effects required ubiquitous overexpression using an actin-gal4 driver that does not allow the identification of tissue-specific effects. Thus, their assertion that the results suggested a strong positive correlation between Drosophila neuromotor regulation and transgenic hTREX1 presence and a negative impact from hTREX1 V235G" is not supported by these data. Also worrisome was the inability to identify the mutant TREX1 protein by Western blot despite the enhanced expression levels suggested by qPCR analysis. Mutant TREX1 cannot exert a dominant effect on cell function if it isn't present.

There are also some technical problems. The lifespan assays lack important controls, and the climbing assays do not appear to have been performed correctly. It is unclear what the WT genetic background is in Figure 1-3, so it is unclear if the appropriate controls have been used. Finally, the lack of information on the specific statistical analyses used for each graph makes it difficult to judge the significance of the data. Overall, the current findings establish the Retinal vasculopathy disease model platform, but with only incremental new data and without any mechanistic insights.

Reviewer #1 (Public review):

Summary:

The NF-kB signaling pathway plays a critical role in the development and survival of conventional alpha beta T cells. Gamma delta T cells are evolutionarily conserved T cells that occupy a unique niche in the host immune system and that develop and function in a manner distinct from conventional alpha beta T cells. Specifically, unlike the case for conventional alpha beta T cells, a large portion of gamma delta T cells acquire functionality during thymic development, after which they emigrate from the thymus and populate a variety of mucosal tissues. Exactly how gamma delta T cells are functionally programmed remains unclear. In this manuscript, Islam et al., use a wide variety of mouse genetic models to examine the influence of the NF-kB signaling pathway on gamma delta T cell development and survival. They find that the inhibitor of kappa B kinase complex (IKK) is critical to the development of gamma delta T1 subsets, but not adaptive/naïve gamma delta T cells. In contrast, IKK-dependent NF-kB activation is required for their long-term survival. They find that caspase 8-deficiency renders gamma delta T cells sensitive to RIPK1-mediated necroptosis and they conclude that IKK repression of RIPK1 is required for the long-term survival of gamma delta T1 and adaptive/naïve gamma delta T cells subsets. These data will be invaluable in comparing and contrasting the signaling pathways critical for the development/survival of both alpha beta and gamma delta T cells.

Comments on revisions:

The word adaptive is misspelt throughout most figures.

Reviewer #2 (Public review):

This study presents a comprehensive genetic dissection of the role of IKK signaling in the development and maintenance of lymphoid gd T cells. By employing a variety of conditional and mutant mouse models, the authors demonstrate that IKK-dependent NF-κB activation is essential for the generation of type 1 gd T cells, while adaptive gd T cells require this pathway primarily for long-term survival. The use of multiple complementary genetic strategies, including IKK deletion and modulation of RIPK1 and CASPASE8 activity, provides robust mechanistic insight into subset-specific regulation of gd T cell homeostasis. Overall, the study provides mechanistic insight for IKK-dependent regulation of gd T cell development and peripheral maintenance.

Comments on revisions:

Thank you for your comments and clarifications.

Reviewer #1 (Public review):

The central pair apparatus of motile cilia consists of two singlet microtubules, termed C1 and C2, each of which is associated with a set of projections, referred to as the C1 and C2 projections. Each projection comprises multiple distinct structural domains, designated a, b, c, and so on. Biochemical studies combined with genetic analyses in Chlamydomonas identified three proteins as the major components of the C2a projection, and subsequent cryo-EM studies confirmed these findings.

In this paper, the authors aim to study the homologues of these three proteins-CCDC108/CFAP65, CFAP70, and MYCBPAP/CFAP147-using knockout mouse models. Biochemical and cell biological analyses demonstrate that, as in Chlamydomonas, these proteins are components of the C2 projection and form a complex that depends on the presence of each other. In addition, the authors use affinity purification to identify two previously uncharacterized proteins and show that they are central pair apparatus proteins that associate with the aforementioned complex. Knockout mice lacking any of the three core proteins exhibit phenotypes consistent with primary ciliary dyskinesia (PCD).

Overall, the manuscript is clearly written, and the data are convincing and support the authors' conclusions. However, given the previous findings in Chlamydomonas, this work provides limited conceptual advances to the field. Nonetheless, it represents a useful and well-documented resource for understanding the conserved organization of the central pair apparatus in motile cilia. It will be of interest to cell and developmental biologists, biochemists, and clinicians studying and treating human ciliopathies.

Reviewer #2 (Public review):

Summary:

This manuscript investigates the protein composition and functional role of the C2a projection of the central apparatus (CA) in vertebrate motile cilia. Using three knockout mouse models (Ccdc108, Mycbpap, and Cfap70), the authors demonstrate that these genes - homologs of Chlamydomonas FAP65, FAP147, and FAP70 - are required for normal motile cilia function in ependymal and tracheal multiciliated cells. Specifically, the authors show that:

(1) Knockout mice for each gene exhibit primary ciliary dyskinesia phenotypes (hydrocephalus and sinusitis), accompanied by abnormal ciliary motion and reduced ciliary beat frequency.

(2) CCDC108, MYCBPAP, and CFAP70 physically interact and localize to the axonemal central lumen, consistent with the C2a projection.

(3) Loss of any one of these proteins destabilizes the others and disrupts CA integrity in a tissue-specific manner.

(4) ARMC3 and MYCBP are C2a-associated proteins.

Strengths:

(1) Clarity: the results are presented in a coherent sequence that facilitates understanding of both the rationale and conclusions.

(2) Genetic rigor: three independent knockout mouse lines that exhibit consistent motile cilia phenotypes provide in vivo support for the proposed role of these proteins.

(3) Integration of structural and functional analyses: combination of ultrastructural (TEM) and immunofluorescence data with CBF measurements provides convincing correlation between structural defects and impaired ciliary function.

(4) Mutual dependency model: reciprocal destabilization of CCDC108, MYCBPAP, and CFAP70 supports their interdependence in the C2a assembly.

(5) Expansion of the vertebrate C2a proteome: the identification of ARMC3 and MYCBP as C2a-associated proteins provides a foundation for future mechanistic studies.

Weaknesses:

(1) Mechanistic depth: the data show a convincing correlation between C2a and ciliary function, but the cell type-specificity of CCDC108, MYCBPAP, and CFAP70 knockout effects is underdeveloped. This is an interesting observation that raises mechanistic/structural questions not addressed in the study, such as what is the role of C2a in CP nucleation, maintenance, or mechanical stabilization? Is C2a composition different in different cell types?

(2) Cell model choice: co-immunoprecipitation was performed using mouse testis lysates. While this is a reasonable source of CA proteins from flagellated cells, the functional analyses in this study focus on ependymal and tracheal multiciliated cells. It would therefore be helpful for the authors to clarify the extent to which these interactions are expected to be conserved across ciliated cell types, and to discuss potential tissue-specific differences in CA assembly.

(3) Statistical analysis: the manuscript states "Statistical significance was defined as P < 0.5", which is likely a typo, but should be P < 0.05. In general, the statistical methods require more clarification. In several figures (e.g., 2B, 2D, 5J, 5K), multiple knockout genotypes are compared with WT, yet unpaired t-tests are reported. When more than two groups are analyzed, multiple pairwise t-tests inflate Type I error unless appropriately corrected; a one-way ANOVA with post hoc comparisons (e.g., Dunnett's test for WT-referenced comparisons) would be more appropriate. Furthermore, the analysis of ciliary movement modes (Figure 2D) involves categorical data, for which a t-test is not statistically appropriate. These comparisons could instead be evaluated using chi-square or Fisher's exact tests. Addressing these issues is important to ensure accurate statistical inference.

(4) Methods section: does not sufficiently describe how image-based quantifications were performed. For example, the criteria used to define cilia number, basal body number, and rotational beating are not specified, nor is how CBF measurements were analyzed. The authors should also provide details regarding analysis software and imaging parameters used (and whether they were kept constant across genotypes).

Reviewer #1 (Public review):

Summary:

This manuscript addresses an important question in cardiac biology: whether distinct cardiomyocyte (CM) subpopulations play specialized roles during heart development and regeneration. Using single-cell RNA sequencing and newly generated genetic tools, the authors identify phlda2 as a specific marker of primordial cardiomyocytes in the adult zebrafish heart. They further show that these primordial CMs function are essential for myocardial morphogenesis and coronary vascularization but are dispensable for myocardial regeneration or revascularization after injury. These findings indicate that heart regeneration doesn't simply recapitulate developmental processes.

Strengths:

A major strength of the study is the generation of a phlda2 BAC reporter, which provides a specific and reliable marker for primordial cardiomyocytes. The lack of genetic tools has previously limited functional analysis of this CM population. By using phlda2 regulatory elements to generate reporter and NTR-based ablation lines, the authors can visualize and selectively manipulate primordial CMs in vivo. This enables a direct functional interrogation rather than relying on lineage tracing or correlative evidence. Through genetic ablation, the authors convincingly demonstrate that primordial CMs are essential for myocardial morphogenesis and coronary vascular organization during development but are not necessary for heart regeneration.

Weaknesses:

(1) The manuscript would benefit from clarifying whether the primordial cardiomyocytes ablation affects epicardial cell behaviors during heart development, given that the well-established role of the epicardium in supporting coronary vessel growth, it is possible that the vascular phenotypes observed after primordial CM ablation may be affected, at least in part, by altered epicardial cells.

(2) Because primordial cardiomyocytes form a dense, single-cell-thick layer covering the ventricular surface, it would be informative to determine whether their loss alters the spatial distribution or inward migration of coronary endothelial cells or epicardial cells.

(3) The manuscript carefully examines the relationship between primordial CMs and gata4⁺ cardiomyocytes during regeneration. However, their relationship during heart development should be more fully addressed.

(4) As loss of cardiomyocytes is known to induce gata4:GFP activation during regeneration, it would be important to determine whether ablation of primordial cardiomyocytes alone triggers gata4:GFP expression in neighboring cardiomyocytes. This analysis would further support the conclusion that primordial cardiomyocytes are not required for regenerative responses.

Reviewer #2 (Public review):

Summary:

In the manuscript "Primordial Cardiomyocytes orchestrate myocardial morphogenesis and vascularization but are dispensable for regeneration", Sun et al. identify a novel marker of primordial cardiomyocytes and use it to visualize and ablate the population during development and regeneration. The role of the primordial layer has not been investigated because the tools to manipulate this population have not existed. The manuscript is straightforward, easy to understand, and addresses an important question that has not been explored.

While the manuscript provides important insights into the role of primordial CMs, backed by a convincing methodology, the authors should clarify their requirements for heart development and maturation. Specifically, is the primordial layer required for the fish to survive? Do primordial CMs regenerate when ablated during development, and do the defects observed (in trabecular and compact CMs and coronary vessels) resolve after 10 days post-treatment when they were detected?

Strengths:

The major strengths are the identification of a marker that enables manipulation of primordial cardiomyocytes and the tools generated by the team.

Weaknesses:

The major weakness is not considering the longer-term consequences of primordial layer ablation during development, as it is unclear whether the animals succumb to the acute cardiac defects observed or fully recover.

Reviewer #3 (Public review):

Summary:

The authors performed single-cell RNA sequencing of adult zebrafish hearts and identified markers for distinct cardiomyocyte subpopulations. One marker, phlda2, marks primordial cardiomyocytes. They generated transgenic reporter lines to characterize phlda2 expression patterns and a phlda2-NTR ablation line to determine the functional requirement of primordial cardiomyocytes during heart regeneration. They found that phlda2+ primordial cardiomyocytes are essential for myocardial morphogenesis and coronary vessel development. Interestingly, when phlda2+ primordial cardiomyocytes are ablated during heart regeneration, gata4+ cortical cardiomyocytes, coronary vessel revascularization, and scar tissue formation are not affected.

Strengths:

The authors identified a new primordial cardiomyocyte marker, phlda2. They further demonstrated that primordial cardiomyocytes are important for heart morphogenesis but dispensable for heart regeneration. Their findings reveal a potential difference between heart development and regeneration programs.

Weakness:

Despite the interesting findings, the authors did not provide supplemental data for their scRNAseq to demonstrate the data quality and support their conclusions, and some results are not well described.

Reviewer #1 (Public review):

In the manuscript entitled "Flexible and high-throughput simultaneous profiling of gene expression and chromatin accessibility in single cells," Soltys and colleagues present easySHARE-seq, a method described as an improvement upon SHARE-seq for the simultaneous measurement of RNA transcripts and chromatin accessibility.

The authors demonstrate the utility of easySHARE-seq by profiling approximately 20,000 nuclei from the murine liver, successfully annotating cell types and linking cis-regulatory elements to target genes. The authors claim that easySHARE-seq supports longer read lengths potentially enabling better variant discovery or allele-specific signal assessment, though they do not provide direct evidence to support these specific claims.

A key strength of the protocol is enhanced sequencing efficiency, achieved by shortening the Index 1 read from 99 to 17 nucleotides. This reduction does not come at a significant cost to barcode diversity, retaining approximately 3.5 million combinations. Additionally, the approach allows for the sequencing of a sub-library to assess quality prior to final barcoding and sequencing which seems quite clever.

While the increase in RNA transcript recovery is substantial, it appears to come at a cost: there is a notable decrease in ATAC fragments per cell compared to the original SHARE-seq (and other platforms). Likely as a result, the dimensionality reduction (UMAP) shows good resolution for RNA profiles but relatively poor resolution for accessibility profiles. Furthermore, the presented data suggests potential ambient RNA contamination; specifically, the detection of Albumin in HSCs and B cells is likely an artifact of the protocol rather than a biological signal.

Overall, the study is well-presented and represents a promising advance. However, there are significant shortcomings that should be addressed, particularly regarding "leaky" transcript recovery and reduced ATAC performance.

Recommendations:

(1) To provide a comprehensive view of the current field, the authors should include Scale Biosciences (Scale Bio) in their discussion of available commercial platforms.

(2) A head-to-head comparison with the 10x Genomics Multiome platform would be of significant interest to the single-cell genomics community and would better contextualize the performance of easySHARE-seq.

(3) Optimizing ATAC Performance: I strongly suggest exploring methods to improve ATAC sensitivity. As the authors note, the improvement in RNA recovery may result from fewer processing steps and stronger fixation. It would be valuable to test if decreasing fixation back to 2% (as in the original SHARE-seq) recovers ATAC data quality, and to determine if the fixation level or the number of steps is the key variable in preserving transcripts.

(4) The authors allude to the possibility of scaling this assay using a barcoded poly(T). Explicit inclusion or demonstration of this capability would dramatically increase interest in this protocol. Perhaps ATAC could be scaled using a barcoded Tn5?

(5) The number of HSCs and B cells expressing Albumin is problematic and suggests significant ambient RNA issues that need to be addressed or computationally corrected.

Reviewer #2 (Public review):

Aims:

The authors sought to optimize SHARE-seq, a multimodal single-cell method, to improve the simultaneous profiling of gene expression and chromatin accessibility. Their goal was to enhance barcode design for better sequencing efficiency and cost savings, while improving overall data quality. They then applied their optimized method, easySHARE-seq, to study liver sinusoidal endothelial cells (LSECs) to demonstrate its utility in examining gene regulation and spatial zonation.

Strengths:

The improved barcode design is an advance, increasing the proportion of sequencing reads dedicated to biological information rather than barcode identification. This modification offers practical benefits in terms of sequencing costs and read length, potentially reducing alignment errors. The method also demonstrates improved RNA detection compared to the original SHARE-seq protocol. The biological applications showcase how simultaneous measurement of both modalities enables analyses that would be practically impossible with single-modality approaches, particularly in examining how chromatin states change along developmental or spatial trajectories.

Weaknesses:

There is a notable reduction in chromatin accessibility detection compared to the original SHARE-seq method, likely limiting the broad use of the method. While the authors are transparent about this tradeoff, additional discussion would be helpful regarding how this affects data interpretation. Comparisons showing consistency between easySHARE-seq and SHARE-seq chromatin accessibility patterns at the single-cell level would strengthen confidence in the method.

Overall:

The authors achieve their aim of creating an optimized protocol with improved barcode design and enhanced RNA detection. The method represents a useful advance for specific experimental contexts where the tradeoffs are appropriate.

Reviewer #1 (Public review):

Summary:

They use cultures of insulinoma MIN6 cells that form spheroids in a micro-patterned PEG-hydrogel to measure Ca2+ oscillations in multiple cells simultaneously.

Strengths:

They demonstrate that insulinoma spheroids are formed in multi-well plates and that Ca2+ imaging can be performed on them.

Weaknesses:

The type of equipment and multi-wells used for the experiments are very specialized to be used as a common tool. Insulinoma cells are tumoral cell lines that divide, unlike primary beta cells. Pancreatic islets are very different from this preparation, as they are highly heterogeneous, whereas these cells all respond equally. It would be good to see the same technique applied to primary cells.

MIN6 cells do not respond to glucose and other secretagogues in the same way as primary cells, and they cycle, depending on the phase of the cycle to which they are exposed.

The authors should report the number of cells per spheroid and the number of cells that are alive and dead.

I would like to examine the effects of calcium channel blockers on calcium transients, and the use of pregnenolone is already described in the literature, but remains less well known.

MIN6 cells secrete much insulin, because detecting the hormone in ELISAs requires too many primary cells. The authors should discuss the model in greater detail and compare it with primary beta cells. Also, they take 3 mM glucose as the basal concentration, which is low.

Reviewer #2 (Public review):

Summary:

The study by Robben et al., show 3D beta-cell spheroid platform, a valuable tool allowing high-throughput monitoring of cytoplasmic Ca concentrations and insulin secretion, with Ca signals comparable to those recorded in primary islets. The authors demonstrate a solid method to culturing MIN6 cells in a 3D culture system, recording Ca signals in a high-throughput format and characterizing these Ca signals using pharmacological tools, including TRPM3 channel and K-ATP channel modulators. This highlights the utility of the 3D beta-cell spheroid for screening new ion channel modulators in beta-cells of the pancreas.

Strengths:

- The study shows that the MIN-6-based 3D beta-cell model is better to study Ca-signaling and insulin secretion compared to 2D culture of single MIN-6 cells.

- The method allows imaging of Ca signaling in many spheroids in parallel followed by collecting medium to measure insulin release and correlate both effects.

- The authors demonstrate that this system is suitable for screening new pharmacological modulators and used as an agonist of the ATP-sensitive potassium channel (diazoxide) and the agonist and antagonist of the TRPM3 channel.

Weaknesses:

- The study is based on only one cell line, the MIN6 insulinoma cells, which may not fully mimic the pancreatic beta-cells within the islet.

- The authors show only spheroids cultured overnight. A long-term culture is missing to assess beta-cell viability long term function.

- The authors tested their platform using only two compounds. Testing a larger compound library is necessary to make a clear conclusion about the suitability of the platform for high-throughput screening.

Reviewer #3 (Public review):

Summary:

The primary objective of this study is to develop high-throughput screening assays utilizing homogeneous 3D cell cultures that more accurately replicate the intricate architecture and cellular communication found in tissues. The authors have chosen pancreatic islet β-cells as a model system to evaluate agents that modulate insulin release, which is particularly relevant given the increasing prevalence of diabetes mellitus-a significant global health concern. Moreover, the incorporation of human-based 3D spheroids, organoids, or organ-on-chip technologies into drug discovery protocols is essential for enhancing clinical translation, as candidate compounds identified using animal models have often demonstrated limited success in clinical settings.

Strengths:

This study was thoughtfully planned and skillfully carried out. The use of micropatterned hydrogels to observe 19 spheroids at once is an ingenious aspect, which has been effectively validated with Ca microfluorography. Overall, I found this investigation to be exceptionally well-executed and free from notable flaws, as the results clearly back up the conclusions. Additionally, the developed method achieved the proposed aims, providing a high-throughput format with 3D cultures. I believe this study deserves publication.

Weaknesses:

For an HTS assay, authors should incorporate the Z-factor.

Reviewer #1 (Public review):

Summary:

In this report, Dr Jie Sun and colleagues employed high-resolution single-cell technologies (transcriptomic + cytometry) to build a temporal map of lung responses after IAV infection in young and old mice. They performed detailed analyses of several innate and adaptive immune compartments and described how age influences each of them. The data are robustly generated, and the analyses provide interesting observations that could be associated with disease severity in aged mice. Mechanistically, the authors provide evidence that IFNa/g signaling after viral clearance could mediate some long-term respiratory outcomes, possibly by acting on MoIMs.

Strengths:

(1) Comprehensive temporal profiling of lung responses.

(2) Combination of scRNA_seq and flow cytometry.

(3) Mechanistic part assessing the role of IFNa/g signaling.

Weaknesses:

(1) Descriptive nature of the study.

(2) Lack of quantification of lung lesions.

(3) Lung functional measurements were only assessed in aged mice (with or without treatment).

(4) No assessment of global and virus-specific humoral responses, which could be related to changes in B cells.

(5) Recently described "pro-repair" Ly6G+ macrophages after IAV infection (PMID: 39093958) are not considered here, and the gating strategy encompasses them in the neutrophil gate.

(6) The authors suggest that IMs in the aged lung may serve as a major contributor to the pathogenesis of long-term sequelae observed in aged hosts, but do not assess this possibility experimentally.

Addressing the weaknesses identified above would substantially strengthen the conclusions of the manuscript.

Reviewer #2 (Public review):

Summary:

In this paper, the authors leverage single-cell approaches to delve deeper into the host responses and immune cells involved in immunopathogenesis of influenza virus infection in aged mice. The dynamics of gene expression and immune cell frequencies were also tracked at multiple time-points to examine the acute and chronic changes in young and aged mice after influenza virus infection. Their analyses demonstrated that the immune cell frequencies and gene signatures differed in young and aged mice, especially macrophages, T cells and B cells. Furthermore, interferon pathways were found to be differentially regulated in the young and aged mice, and blocking the interferon pathway with monoclonal antibodies led to improvement in lung respiratory functions and reduced inflammation.

Strengths:

A strength of this study is that multiple time points are considered for analyses, allowing assessment of temporal changes in gene expression and immune cell frequencies after virus infection during the acute and chronic phases of the disease. The data presented could also serve as a potential resource for other researchers interested in understanding the host responses to the influenza virus, especially in aged mice. Another interesting finding was that blocking interferon signalling can reduce the chronic severe symptoms caused by the influenza virus in aged mice.

Weaknesses:

The manuscript could greatly benefit from more rigorous approaches, particularly in the statistical analyses and data visualisation. Moreover, the scientific rationale and logic for several parts of the manuscript can be improved. Finally, the authors did not adequately dissect whether the contribution of host responses was from virus infection or from bystander effects. Specifically, my major comments are as follows:

(1) While it is interesting to compare the difference in host responses between aged and young mice, the authors should also more deeply characterise the differences in phenotypic and infection kinetics between aged and young mice, so that the readers can better appreciate the effects of virus infection and host immune tolerance to viral infection. For instance, what are the differences in virus infection kinetics between the aged and young mice? Are the levels of infection different? Are the virus dynamics and kinetics different between aged and young mice? Besides lung tissue damage, are there also tissue damage or inflammatory responses beyond lung tissues that differ between aged and young mice?

(2) Figure 1B: Could the authors quantify the extent of tissue damage in aged and young adults? It is challenging to interpret the extent of tissue damage, especially across the different time points.

(3) Figure 1D: The authors claim that the senescence signatures are higher in aged mice, justifying that the pathway analyses are consistent with ageing signatures. However, it is also important to note that the senescence signatures were insignificant in aged mice after day 14. Is this expected?

(4) Figure 1E: The stacked bar charts are difficult to read. It is unclear if the cell type frequencies or proportions are significantly changed, especially as the authors are showing these changes with averaged values. Moreover, the authors should keep the colours of the bar charts consistent with the UMAP.

(5) Figure 1F-M: The charts show increased frequencies of innate and adaptive immune cells in aged mice. How about the young mice? Which type of cells are increased to allow these mice to be more tolerant to infection?

(6) Figure 2D and Figure S2C: Besides showing the dynamics of the different clusters, the authors should also display the statistics for individual mice. If the analyses have to be pooled for the single-cell analysis, the authors should declare the challenges and show the statistical comparisons for the flow cytometry.

(7) Figure 3E: The authors should not claim differences in somatic hypermutation based on gene expression. This will require BCR sequencing and evidence for clonal expansion to confirm that there are differences in somatic hypermutation. Moreover, the authors did not measure the quality and quantity of antibody responses between aged and young mice. The claims for the antibody responses are thus extrapolated, and the B cell identities cannot be identified without any functional or phenotypic readouts.

(8) Figure 4H. Why did the authors not perform the experiments for aged mice with a higher virus dose? Also, the spider plots do not display the variability between individual mice, making it challenging to interpret whether the changes were statistically different between the conditions.

(9) Figure 5A. Is the interferon pathway the only pathway that was significantly enriched in the aged mice? Is it the top pathway? The authors should also show the other pathways that were significantly enriched in aged mice. Did the authors also analyse whether the differences in interferon pathways were caused by infected cells or by bystander cells?

(10) Figure 5B: Based on the pathway analyses, the peak responses for interferon are at day 9 post-infection. However, the interferon treatment is performed on day 14, where differences were less apparent. Why did the authors choose to do the interferon treatment at day 14 instead?

(11) Figure 6: How about interferon-mediated cell-cell interactions? The authors should consider using established libraries such as Cell Chat to determine if there are any cell-cell communications that lead to differences in interferon responses and signaling.

(12) Throughout the whole manuscript, the authors kept emphasising that the aged mice displayed uncoordinated immune responses, yet, based on the pathway analyses and phenotypic characterisation, it appears that only interferon was mainly dysregulated. I would thus like to recommend that the authors adjust the tone of the manuscript to tailor it to the results obtained from their investigations.

Reviewer #1 (Public review):

Summary:

In the ecological interactions between wild plants and specialized herbivorous insects, structural innovation-based diversification of secondary metabolites often occurs. In this study, Agrawal et al. utilized two milkweed species (Asclepias curassavica and Asclepias incarnata) and the specialist Monarch butterfly (Danaus plexippus) as a model system to investigate the effects of two N,S-cardenolides-formed through structural diversification and innovation in A. curassavica-on the growth, feeding, and chemical sequestration of D. plexippus, compared to other conventional cardenolides. Additionally, the study examined how cardenolide diversification resulting from the formation of N,S-cardenolides influences the growth and sequestration of D. plexippus. On this basis, the research elucidates the ecophysiological impact of toxin diversity in wild plants on the detoxification and transport mechanisms of highly adapted herbivores.

Strengths:

The study is characterized by the use of milkweed plants and the specialist Monarch butterfly, which represent a well-established model in chemical ecology research. On one hand, these two organisms have undergone extensive co-evolutionary interactions; on the other hand, the butterfly has developed a remarkable capacity for toxin sequestration. The authors, building upon their substantial prior research in this field and earlier observations of structural evolutionary innovation in cardenolides in A. curassavica, proposed two novel ecological hypotheses. While experimentally validating these hypotheses, they introduced the intriguing concept of a "non-additive diversity effect" of trace plant secondary metabolites when mixed-contrasting with traditional synergistic perspectives-in their impact on herbivores.

Weaknesses:

The manuscript has two main weaknesses. First, as a study reliant on the control of compound concentrations, the authors did not provide sufficient or persuasive justification for their selection of the natural proportions (and concentrations) of cardenolides. The ratios of these compounds likely vary significantly across different environmental conditions, developmental stages, pre- and post-herbivory, and different plant tissues. The ecological relevance of the "natural proportions" emphasized by the authors remains questionable. Furthermore, the same compound may even exert different effects on herbivorous insects at different concentrations. The authors should address this issue in detail within the Introduction, Methods, or Discussion sections.

Second, the study was conducted using leaf discs in an in vitro setting, which may not accurately reflect the responses of Monarch butterflies on living plants. This limitation undermines the foundation for the novel ecological theory proposed by the authors. If the observed phenomena could be validated using specifically engineered plant lines-such as those created through gene editing, knockdown, or overexpression of key enzymes involved in the synthesis of specific N,S-cardenolides-the findings would be substantially more compelling.

Reviewer #2 (Public review):

I have reviewed both the original and revised version of this manuscript and while no additional experiments were added, I find the interpretations and discussion of the limitations of the study have improved. This is appreciated.

My original concern regarding the mixture treatments largely remains. Figure 4 nicely shows that the mixtures are more potent than the average of all single compounds. However, Fig S3 shows that the effects of mixtures are not significantly different from effects of at least one, single N,S compound (voruscharin or uscharin) across all measured growth/sequestration responses. Essentially, the effects of single N,S compounds is similar to mixtures (which also contain N,S compounds).

While the current results are certainly interesting as presented, in my view the main takeaway of the manuscript would be more compelling if it could be demonstrated that it isn't simply the presence of N,S compounds in the mixtures driving the observations. For example, does a mixture of all compounds except voruscharin or uscharin still have stronger growth/sequestration effects compared to single non-N,S compounds?

Reviewer #1 (Public review):

Summary:

Sheidaei and colleagues report a novel and potentially important role for an early mitotic actomyosin-based mechanism, PANEM contraction, in promoting timely congression of chromosomes located at the nuclear periphery, particularly those in polar positions. The manuscript will interest researchers studying cell division, cytoskeletal dynamics, and motor proteins. Although some data overlap with the group's prior work, the authors extend those findings by optimizing key perturbations and performing more detailed analyses of chromosome movements, which together provide a clearer mechanistic explanation. The study also builds naturally on recent ideas from other groups about how chromosome positioning influences both early and later mitotic movements.

In its current form, however, the manuscript suffers from major organizational problems, an overcrowded and confusing Results section and figures, and a lack of essential experimental controls and contextual discussion. These deficiencies make it difficult to evaluate the data and the authors' conclusions. A substantial structural revision is required to improve clarity and persuasiveness. In addition, several key control experiments and more conceptual context are needed to establish the specificity and relevance of PANEM relative to other microtubule- and actin-based mitotic mechanisms. Testing PANEM in additional cell lines or contexts would also strengthen the claim. I therefore recommend addressing the structural, conceptual, and experimental issues detailed below.

Major Comments:

(1) Structural overhaul and figure reorganization<br /> The Results section is overly dense, lacks clear structure, and includes descriptive content that belongs in the Methods. Many figure panels should be moved to Supplementary Materials. A substantial reorganization is required to transform the manuscript into a focused, "Reports"-type article.<br /> - Move methodological and descriptive details (e.g., especially from the second Results subheading and Figure 2) to the Methods or Supplementary Materials.<br /> - Remove repetitive statements that simply restate that later phenotypes arise as consequences of delayed Phase 1 (applicable to subheadings 3 onward).<br /> - Figure 4I: This panel is currently unclear and should be drastically simplified.<br /> I recommend to reorganize figures as follows:<br /> - Figure I: Keep as single figure but simplify. Figure 1D and 1E could be combined, move unnormalized SCV to supplementary materials. Same goes for 1F.<br /> - New Figure 2: Combine current Figures 2A, 3A, 3C, 3D, 4C, 4F, and 4H to illustrate how PANEM contraction facilitates initial interactions of peripheral chromosomes with spindle microtubules which increases speed of congression initiation.<br /> - New Figure 3: Combine current Figures 5A, 5C, 5D, 5F, 6B, 6C, and lower panels of 4H to show how PANEM contraction repositions polar chromosomes and reduces chromosome volume in early mitosis to enable rapid initiation of congression.<br /> - New Figure 4: Combine Figures 7A, 7B, 7D, 7E, 7F, expanded Supplementary Figure S7, and new data to demonstrate that PANEM actively pushes peripheral chromosomes inward which is important for efficient chromosome congression in diverse cellular contexts.

(2) Specificity and redundancy of actin perturbation<br /> To establish the specificity and relevance of PANEM, the authors should include or discuss appropriate controls:<br /> - Apply global actin inhibitors (e.g., cytochalasin D, latrunculin A) to disrupt the entire actin cytoskeleton. These perturbations strongly affect mitotic rounding and cytokinesis but only modestly influence early chromosome movements, as reported previously (Lancaster et al., 2013; Dewey et al., 2017; Koprivec et al., 2025). The minimal effect of global inhibition must be addressed when proposing a localized actomyosin mechanism. Comment if the apparent differences in this approach and one that the authors were using arises due to different cell types.<br /> - Clarify why spindle-associated actin, especially near centrosomes, as reported in prior studies using human cultured cells (Kita et al., 2019; Plessner et al., 2019; Aquino-Perez et al., 2024), was not observed in this study. The Myosin-10 and actin were also observed close to centrosomes during mitosis in X. laevis mitotic spindles (Woolner et al., 2008). Possible explanations include differences in fixation, probe selection, imaging methods, or cell type. Note that some actin probes (e.g., phalloidin) poorly penetrate internal actin, and certain antibodies require harsh extraction protocols. Comment on possibility that interference with a pool of Myo10 at the centrosomes is important for effects on congression.

(3) Expansion of PANEM functional analysis<br /> To strengthen the conclusions and broaden the study beyond the group's previous work, PANEM function should be tested in additional contexts (some may be considered optional but important for broader impact):<br /> - Test PANEM function in at least one additional cell line that displays PANEM to rule out cell-line-specific effects.<br /> - Examine higher-ploidy or binucleated cells to determine whether multiple PANEM contractions are coordinated and if PANEM contraction contributes more in cells of higher ploidies or specific nuclear morphologies.<br /> - Investigate dependency on nuclear shape or lamina stiffness; test whether PANEM force transmission requires a rigid nuclear remnant.<br /> - Analyze PANEM's contribution under mild microtubule perturbations that are known to induce congression problems (e.g., low-dose nocodazole).<br /> - Evaluate PANEM contraction role in unsynchronized U2OS cells, where centrosome separation can occur before NEBD in a subset of cells (Koprivec et al., 2025), and in other cell types with variable spindle elongation timing.<br /> - Quantify not only the percentage of affected cells after azBB but also the number of chromosomes per cell with congression defects in the current and future experiments.

(4) Conceptual integration in Introduction and Discussion<br /> The manuscript should better situate its findings within the context of early mitotic chromosome movements:<br /> - Clearly state in the Introduction and elaborate in the Discussion that initiation of congression is coupled to biorientation (Vukušić & Tolić, 2025). This provides essential context for how PANEM-mediated nuclear volume reduction supports efficient congression of polar chromosomes.<br /> - Explain that PANEM is most critical for polar chromosomes because their peripheral positions are unfavorable for rapid biorientation (Barišić et al., 2014; Vukušić & Tolić, 2025).<br /> - Discuss how cell lines lacking PANEM (e.g., HeLa and others) nonetheless achieve efficient congression, and what alternative mechanisms compensate in the absence of PANEM. For example, it is well established that cells congress chromosomes after monastrol or nocodazole washout, which essentially bypasses the contribution of PANEM contraction.

Significance:

Advance:<br /> This study's main strength is its novel and potentially important demonstration that contraction of PANEM, a peripheral actomyosin network that operates contracts early mitosis, contributes to the timely initiation of chromosome congression, especially for polar chromosomes. While PANEM itself was previously described by this group, this manuscript provides new mechanistic evidence, improved perturbations, and detailed chromosome tracking. To my knowledge, no prior studies have mechanistically connected this contraction to polar chromosome congression in this level of detail. The work complements dominant microtubule-centric models of chromosome congression and introduces actomyosin-based forces as a cooperating system during very early mitosis. However, the impact of the study is currently limited by major organizational issues, insufficient controls, and incomplete contextualization within existing literature.

Audience:<br /> Primary audience of this study will be researchers working in cell division, mitosis, cytoskeleton dynamics, and motor proteins. The findings may interest also the wider cell biology community, particularly those studying chromosome segregation fidelity, spindle mechanics, and cytoskeletal crosstalk. If validated and clarified, the concept of PANEM could be integrated into textbooks and models of chromosome congression and could inform studies on mitotic errors and cancer cell mechanics.

Expertise:<br /> My expertise lies in kinetochore-microtubule interactions, spindle mechanics, chromosome congression, and mitotic signaling pathways.

Reviewer #2 (Public review):

Summary:

In this manuscript, Sheidaei et al. reported on their study of chromosome congression during the early stages of mitotic spindle assembly. Building on their previous study (ref. #15, Booth et al., eLife, 2019), they focused on the exact role of the actin-myosin-based contraction of the nuclear envelope. First, they addressed a technical issue from their previous study, finding a way to specifically impair the actomyosin contraction of the nuclear membrane without affecting the contraction of the plasma membrane. This allowed them to study the former more specifically. They then tracked individual kinetochores to reveal which were affected by nuclear membrane contraction and at what stage of displacement towards the metaphase plate. The investigation is rigorous, with all the necessary controls performed. The images are of high quality. The analyses are accurate and supported by convincing quantifications. In summary, they found that peripheral chromosomes, which are close to the nuclear membrane, are more influenced by nuclear membrane contraction than internal chromosomes. They discovered that nuclear membrane contraction primarily contributes to the initial displacement of peripheral chromosomes by moving them towards the microtubules. The microtubules then become the sole contributors to their motion towards the pole and subsequently the midplane. This step is particularly critical for the outermost chromosomes, which are located behind the spindle pole and are most likely to be mis-segregated.

Significance:

While the conclusions are somewhat intuitive and could be considered incremental with regard to previous works, they are solid and improve our understanding of mitotic fidelity. The authors had already reported the overall role of nuclear membrane contraction in reducing chromosome mis-segregation in their previous study, as mentioned fairly and transparently in the text. However, the reason for this is now described in more detail with solid quantification. Overall, this is good-quality work which does not drastically change our understanding of chromosome congression but contributes to improving it. Personally, I am surprised by the impact of such a small contraction (of around one micron) on the proper capture of chromosomes and wonder whether the signalling associated with the contraction has a local impact on microtubule dynamics. However, investigating this point is clearly beyond the scope of this study.

Reviewer #3 (Public review):

Summary:

Sheidaei et al., report how chromosomes are brought to positions that facilitate kinetochore-microtubule interactions during mitosis. The study focusses on an important early step of the highly orchestrated chromosome segregation process. Studying kinetochore capture during early prophase is extremely difficult due to kinetochore crowding but the team has taken up the challenge by classifying the types of kinetochore movements, carefully marking kinetochore positions in early mitosis and linking these to map their fate/next-positions over time. The work is an excellent addition to the field as most of the literature has thus far focussed on tracking kinetochore in slightly later stages of mitosis. The authors show that the PANEM facilitates chromosome positioning towards the interior of the newly forming spindle, which in turn facilitates chromosome congression - in the absence of PANEM chromosomes end up in unfavourable locations, and they fail to form proper kinetochore-microtubule interactions. The work highlights the perinuclear actomyosin network in early mitosis (PANEM) as a key spatial and temporal element of chromosome congression which precedes the segregation process.

Major Comments:

(1) The complexity of tracking has been managed by classifying kinetochore movements into 4 categories, considering motions towards or away from the spindle mid-plane. While this is a very creative solution in most cases, there may be some difficult phases that involve movement in both directions or no dominant direction (e.g. Phase3-like). It is unclear if all kinetochores go through phase1, 2, 3 and 4 in a sequential or a few deviate from this pattern. A comment on this would be helpful. Also, it may be interesting to compare those that deviate from the sequence and ask how they recover in the presence and absence of azBB.

(2) Would peripheral kinetochore close to poles behave differently compared to peripheral kinetochore close to the midplane (figure S4) ?In figure 3D, are they separated? If not, would it look different?

(3) Uncongressed polar chromosomes (e.g., CENPE inhibited cells) are known to promote tumbling of the spindle. In figure 5B with polar chromosomes, it will be helpful to indicate how the authors decouple spindle pole movements from individual kinetochore movements.

(4) The work has high quality manual tracking of objects in early mitosis- if this would be made available to the field, it can help build AI models for tracking. The authors could consider depositing the tracking data and increasing the impact of their work.

Significance:

The current work builds upon their previous work, in which the authors demonstrated that an actomyosin network forms on the cytoplasmic side of the nuclear envelope during prophase. This work explains how the network facilitates chromosome capture and congression by tracking motions of individual kinetochores during early mitosis. The findings can be broadly useful for cell division and the cytoskeletal fields.

Reviewer #1 (Public review):

Summary:

In this manuscript, Shen et al. have improved upon the mitotic clone analysis tool MAGIC that their lab previously developed. MAGIC uses CRISPR/Cas9-mediated double-stranded breaks to induce mitotic recombination. The authors have replaced the sgRNA scaffold with a more effective scaffold to increase clone frequency. They also introduced modifications to positive and negative clonal markers to improve signal-to-noise and mark the cytoplasm of the cells instead of the nuclei. The changes result in increase in clonal frequencies and marker brightness. The authors also generated the MAGIC transgenics to target all chromosome arms and tested the clone induction efficacy.

Strengths:

MAGIC is a mitotic clone generation tool that works without prior recombination to special chromosomes (e.g., FRT). It can also generate mutant clones for genes for which the existing FRT lines could not be used (e.g., the genes that are between the FRT transgene and the centromere).

This manuscript does a thorough job in describing the method and provides compelling data that support improvement over the existing method.

Reviewer #2 (Public review):

Summary:

In this study, the authors present the latest improvement of their previously published methods, pMAGIC and nMAGIC, which can be used to engineer mosaic gene expression in wild-type animals and in a tissue-specific manner. They address the main limitation of MAGIC, the lack of gRNA-marker transgenes, which has hampered the broader adoption of MAGIC in the fly community. To do so, they create an entire toolkit of gRNA markers for every Drosophila chromosome and test them across a range of different tissues and in the context of making Drosophila species hybrid mosaic animals. The study provides a significant and broadly useful improvement compared to earlier versions, as it broadens the use-cases for transgenic manipulation with MAGIC to virtually any subfield of Drosophila cell biology.

Strengths:

Major improvements to MAGIC were made in terms of clone induction efficiency and usability across the Drosophila model system, including wild-type genotypes and the use in non-melanogaster species.

Notably, mosaic mutants can now be created for genes residing on the 4th chromosome, which is exciting and possibly long-awaited by 4th chromosome gene enthusiasts.

Selection of the standard set of gRNA markers was done thoughtfully, using non-repetitive conserved and unique sequences.

The authors demonstrate that MAGIC can be used easily in the context of interspecific hybrids. I believe this is a great advancement for the Drosophila community, especially for evolutionary biologists, because this may allow for easy access to mechanistic, tissue-specific insight into the process of a range of hybrid incompatibilities, an important speciation process that is normally difficult to study at the level of molecular and cell biology.

In the same way, because it is not limited to usage in any particular genetic background, genome-wide MAGIC can be potentially used in wild-type genotypes relatively easily. This is exciting, especially because natural genetic diversity is rarely investigated more mechanistically and at the scale/resolution of cells or specific tissues. Now, one can ask how a particular naturally occurring allele influences cell physiology compared to another (control) while keeping the global physiological context of the particular genetic background largely intact.

Reviewer #3 (Public review):

Summary:

In the manuscript by Shen, Yeung, and colleagues, the authors generate an improved and expanded Mosaic analysis by gRNA-induced crossing-over (MAGIC) toolkit for use in making mosaic clones in Drosophila. This is a clever method by which mitotic clones can be induced in dividing cells by using CRISPR/Cas9 to generate double-strand breaks at specific locations that induce crossing over at those locations. This is conceptually similar to previous mosaic methods in flies that utilized FRT sites that had been inserted near centromeres along with heat-shock inducible FLPase. The advantage of the MAGIC system is that it can be used along with chromosomes lacking FRT sites already introduced, such as those found in many deficiency collections or in EMS mutant lines. It may also be simpler to implement than FRT-based mosaic systems. There are two flavors of the MAGIC system: nMAGIC and pMAGIC. In nMAGIC, the main constituents are a transgene insertion that contains gRNAs that target DNA near the centromere, along with a fluorescent marker. In pMAGIC, the main constituents are a transgenic insertion that contains gRNAs that target DNA near the centromere, along with ubiquitous expression of GAL80. As such, nMAGIC can be used to generate clones that are not labelled, whereas pMAGIC (along with a GAL4 line and UAS-marker) can be used much like MARCM to positively label a clone of cells. This manuscript introduces MAGIC transgenic reagents that allow all 4 chromosomes to be targeted. They demonstrate its use in a variety of tissues, including with mutants not compatible with current FLP/FRT methods, and also show it works well in tissues that prove challenging for FLP/FRT mosaic analyses (such as motor neurons). They further demonstrate that it can be used to generate mosaic clones in non-melanogaster hybrid tissues. Overall, this work represents a valuable improvement to the MAGIC method that should promote even more widespread adoption of this powerful genetic technique.

Strengths:

(1) Improves the design of the gRNA-marker by updating the gRNA backbone and also the markers used. GAL80 now includes a DE region that reduces the perdurance of the protein and thus better labeling of pMAGIC clones. The data presented to demonstrate these improvements is rigorous and of high quality.

(2) Introduces a toolkit that now covers all chromosome arms in Drosophila. In addition, the efficiency of 3 target different sites is characterized for each chromosome arm (e.g., 3 different gRNA-Marker combinations), which demonstrate differences in efficiency. This could be useful to titrate how many clones an experimenter might want (e.g., lower efficiency combinations might prove advantageous).

(3) The manuscript is well written and easy to follow. The authors achieved their aims of creating and demonstrating MAGIC reagents suitable for mosaic analysis of any Drosophila chromosome arm.

(4) The MAGIC method is a valuable addition to the Drosophila genetics toolkit, and the new reagents described in this manuscript should allow it to become more widely adopted.

Comments on revised version:

The authors have done a great job addressing reviewer concerns with the addition of updated figures, new experiments, and changes to the manuscript. I am supportive of this version and agree with the updated assessment.

Reviewer #1 (Public review):

Summary:

Synaptotagmin (Syt) 1 and Syt7 specifically promote (are critical for) MAIT cell activation in response to M.tb-infected bronchial epithelial cell line BEAS-2B (Fig. 1) and monocyte-like cell line THP-1 (Fig. 3), but not at the M.smeg-infected conditions. Esyt2 shows a similar effect. This work also displayed co-localization of Syt1 and Syt7 with Rab7a and Lamp1, but not with Rab5a (Fig. 5). Loss of Syt1 and Syt7 resulted in a larger area of MR1 vesicles (Fig. 6f) and an increased number of MR1 vesicles in close proximity to an Auxotrophic Mtb-containing vacuoles during infection (Fig. 7ab). Moreover, flow organellometry to separate phagosomes from other subcellular fractions and identify enrichment of auxotrophic Mtb-containing vacuoles in fractions 42-50, which were enriched with Lamp1+ vacuoles or phagosomes (Fig.7e-f).

Strengths:

This work convincingly associated Syt1 and Syt7 with late endocytic compartments and Mtb+ vacuoles. Gene editing of Syt1 and Syt7 loci of bronchial epithelial and monocyte-like cells supported Syt1 and Syt7 facilitated maintaining a normal level of antigen presentation for MAIT cell activation in Mtb infection. Imaging analyses provided solid evidence to support that Syt1 and Syt7 mutants enhanced the size of MR1-resided vesicles, the overlaps of MR1 with M.tb fluorescent signal, and the MR1 proximity with Mtb-infected vacuoles, suggesting that Syt1 and Syt7 proteins help antigen presentation for MAIT activation in Mtb infection.

Weaknesses:

Current data could be improved to support the conclusion that "This study identifies a pathway in which Syt1 and Syt7 facilitate the translocation of MR1 from Mtb-containing vacuoles, potentially to the cell surface for antigen presentation". Likewise, the current data are more supportive of a different conclusion.

Comments on revisions:

Authors have been very responsive to the review comments, except for keeping a very strong conclusion. Suggest rewriting the conclusions "identifies a specialized pathway", "facilitate the translocation", "from Mtb-containing vacuoles", and "potentially to the cell surface" to be more reflective of the data.

Reviewer #3 (Public review):

Summary:

In the submitted manuscript the authors investigate the role of Synaptotagmins (Syt1) and (Syt7) in MR1 presentation of Mtb antigens. By using Syt1 and Syt7 knock down the authors determine that these molecules are required to effectively control Mtb infection.

Strengths:

In the first series of experiments, the authors determined that knocking down Syt1 and Sy7 in antigen-presenting cells decreases IFN-γ production following cellular infection with Mtb. These experiments are well performed and controlled.

Comments on revisions:

The revised manuscript offers further support to the role of Synaptogamins 1 and 7 in MR1 trafficking during MT infection

Reviewer #1 (Public review):

Summary:

The study by Yu et al investigated the role of protein N-glycosylation in regulating T-cell activation and functions is an interesting work. By using genome-wide CRISPR/Cas9 screenings, authors found that B4GALT1 deficiency could activate expression of PD-1 and enhance functions of CD8+ T cells both in vitro and in vivo, suggesting the important roles of protein N-glycosylation in regulating functions of CD8+ T cells, which indicates that B4GALT1 is a potential target for tumor immunotherapy.

Strengths:

The strengths of this study are the findings of novel function of B4GALT1 deficiency in CD8 T cells.

Weaknesses:

Although authors have partly addressed my questions, including potential mechanism, however, I found that the impact of B4GALT1 deficiency for T cell function against tumor cells was not very striking, in comparing to other recently identified genes, which may limit its application, such as in adoptive T cell therapy.

Comments on revisions:

Authors have addressed the questions raised in previous review.

Reviewer #2 (Public review):

Summary:

In this study, the authors identify the N-glycosylation factor B4GALT1 as an important regulator of CD8 T-cell function.

Strengths:

The use of complementary ex vivo and in vivo CRISPR screens is commendable and provides a useful dataset for future studies of CD8 T-cell biology.

The authors perform multiple untargeted analyses (RNAseq, glycoproteomics) to hone their model on how B4GALT1 functions in CD8 T-cell activation, as well as the use of a CD8-CD3 to narrow down the effects of B4GALT1, which is a broad-acting enzyme.

B4GALT1 is shown to be important in both in vitro T-cell killing assays and a mouse model of tumor control, reinforcing the authors' claims.

Weaknesses:

The authors did not verify the efficiency of knockout in their single gene KO lines, although they mention a plan to include such data in a future version of the manuscript.

The specific N-glycosylation sites of TCR and CD8 are not identified, and would be helpful for site-specific mutational analysis to further the authors' model.

The study or future studies could benefit from further in vivo experiments testing the role of B4GALT1 other physiological contexts relevant to CD8 T cells, for example autoimmune disease or infectious disease.

Comments on revisions:

The paper improved after revision.

Reviewer #1 (Public review):

Summary:

This manuscript analyzes a large dataset of [NiFe]-CODHs with a focus on genomic context and operon organization. Beyond earlier phylogenetic and biochemical studies, it addresses CODH-HCP co-occurrence, clade-specific gene neighborhoods, and operon-level variation, offering new perspectives on functional diversification and adaptation.

Strengths:

The study has a valuable approach.

Comments on revised version:

I am satisfied that the authors have adequately addressed my previous comments in the revised manuscript.

Reviewer #2 (Public review):

The authors present a comparative genomic and phylogenetic analysis aimed at elucidating the functions of nickel-dependent carbon monoxide dehydrogenases (Ni-CODHs) and hybrid-cluster proteins (HCPs). By examining gene neighborhoods, phylogenetic relationships, and co-occurrence patterns, they propose functional hypotheses for different CODH clades and highlight those with the greatest potential for biotechnological applications.

A major strength of this work lies in its systematic and conceptually clear approach, which provides a rapid and low-cost framework for predicting the functional potential of newly identified CODHs based on sequence data and genomic context. The analysis is careful in minimizing false positives and offers valuable insights into the diversity and distribution of CODH enzyme clades.

Reviewer #1 (Public review):

Summary:

This study compares four models-VALOR (dynamic visual-text alignment), CLIP (static visual-text alignment), AlexNet (vision-only), and WordNet (text-only)-in their ability to predict human brain responses using voxel-wise encoding modeling. The results show that VALOR not only achieves the highest accuracy in predicting neural responses but also generalizes more effectively to novel datasets. In addition, VALOR captures meaningful semantic dimensions across the cortical surface and demonstrates impressive predictive power for brain responses elicited by future events.

Strengths:

The study leverages a multimodal machine learning model to investigate how the human brain aligns visual and textual information. Overall, the manuscript is logically organized, clearly written, and easy to follow. The results well support the main conclusions of the paper.

Comments on revisions:

I am happy with the response letter. I have no further comments on this manuscript.

Reviewer #2 (Public review):

Summary:

Fu and colleagues have shown that VALOR, a model of multimodal and dynamic stimulus features, better predicts brain responses compared to unimodal or static models such as AlexNet, WordNet, or CLIP. The authors demonstrated robustness of their findings from generalizing encoding results to an external dataset. They demonstrated the models' practical benefit by showing that semantic mappings were comparable to another model that required labor-intensive manual annotation. Finally, the authors showed that the model reveals predictive coding mechanisms of the brain, which held meaningful relationship with individuals' fluid intelligence measure.

Strengths:

Recent advances in neural network models that extract visual, linguistic, and semantic features from real-world stimuli have enabled neuroscientists to build encoding models that predict brain responses from these features. Higher prediction accuracy indicates greater explained variance in neural activity, and therefore a better model of brain function. Commonly used models include AlexNet for visual features, WordNet for audio-semantic features, and CLIP for visuo-semantic features; these served as comparison models in the study. Building on this line of work, the authors developed an encoding model using VALOR, which captures the multimodal and dynamic nature of real-world stimuli. VALOR outperformed the comparison models in predicting brain responses. It also recapitulated known semantic mappings and revealed evidence of predictive processing in the brain. These findings support VALOR as a strong candidate model of brain function.

Weaknesses:

The authors argue that this modeling contributes to better understanding how the brain works. However, upon reading, I am less convinced how VALOR's superior performance than other models tell us more about the brain. VALOR is a better model of the audiovisual stimulus because it processes multimodal and dynamic stimuli compared to other unimodal or static models. If the model better captures real-world stimuli, then I almost feel that it has to better capture brain responses, assuming that the brain is a system that is optimized to process multimodal and dynamic inputs from the real world. The authors could strengthen the manuscript if the significance of their encoding model findings is better explained.

In Study 3, the authors show high alignment between WordNet and VALOR feature PCs. Upon reading the method together with Figure 3, I suspect that the alignment almost has to be high, given that the authors projected VALOR features to the Huth et al.'s PC space. Could the authors conduct non-parametric permutation tests, such as shuffling the VALOR features prior to mapping onto Huth et al.'s PC space, and then calculating the Jaccard scores? I imagine that the null distribution would be positively shifted. Still, I would be convinced if the alignment is higher than this shifted null distribution for each PC. If my understanding about this is incorrect, I suggest editing the relevant Method section (line 508) because this analysis was not easy to understand.

In Study 4, the authors show that individuals whose superior parietal gyrus (SPG) exhibited high prediction distance had high fluid cognitive scores (Figure 4C). I had a hard time believing that this was a hypothesis-driven analysis. The authors motivate the analysis that "SPG and PCu have been strongly linked to fluid intelligence (line 304)". Did the authors conduct two analyses only-SPG-fluid intelligence and PCu-fluid intelligence-without relating other brain regions with other individual differences measures? Even if so, the authors should have reported the same r value and p value for PCu-fluid intelligence. If SPG-fluid intelligence indeed hold specificity in terms of statistical significance compared to all possible scenarios that were tested, is this rationally an expected result and could the authors explain the specificity? Also, the authors should explain why they considered fluid intelligence to be the proxy of one's ability to anticipate upcoming scenes during movie watching. I would have understood the rationale better if the authors have at least aggregated predictive scores for all brain regions that held significance into one summary statistics and have found significant correlation with the fluid intelligence measure.

Comments on revisions:

The revision has addressed these concerns.

Reviewer #3 (Public review):

Summary:

In this work, the authors aim to improve neural encoding models for naturalistic video stimuli by integrating temporally aligned multimodal features derived from a deep learning model (VALOR) to predict fMRI responses during movie viewing.

Strengths:

The major strength of the study lies in its systematic comparison across unimodal and multimodal models using large-scale, high-resolution fMRI datasets. The VALOR model demonstrates improved predictive accuracy and cross-dataset generalization. The model also reveals inherent semantic dimensions of cortical organization and can be used to evaluate the integration timescale of predictive coding.

This study demonstrates the utility of modern multimodal pretrained models for improving brain encoding in naturalistic contexts. While not conceptually novel, the application is technically sound, and the data and modeling pipeline may serve as a valuable benchmark for future studies.

Weaknesses:

The overall framework of using data-driven features derived from pretrained AI models to predict neural response has been well studied and accepted by the field of neuroAI for over a decade. The demonstrated improvements in prediction accuracy, generalization, and semantic mapping are largely attributable to the richer temporal and multimodal representations provided by the VALOR model, not a novel neural modeling framework per se. As such, the work may be viewed as an incremental application of recent advances in multimodal AI to a well-established neural encoding pipeline, rather than a conceptual advance in modeling neural mechanisms.

Within this setup, the finding that VALOR outperforms CLIP, AlexNet, and WordNet is somewhat expected. VALOR encodes rich spatiotemporal information from videos, making it more aligned with movie-based neural responses. CLIP and AlexNet are static image-based models and thus lack temporal context, while WordNet only provides coarse categorical labels with no stimulus-specific detail. Therefore, the results primarily reflect the advantage of temporally-aware features in capturing shared neural dynamics, rather than revealing surprising model generalization. A direct comparison to pure video-based models, such as Video Swin Transformers or other more recent video models, would help strengthen the argument.

Moreover, while WordNet-based encoding models perform reasonably well within-subject in the HCP dataset, their generalization to group-level responses in the Short Fun Movies (SFM) dataset is markedly poorer. This could indicate that these models capture a considerable amount of subject-specific variance, which fails to translate to consistent group-level activity. This observation highlights the importance of distinguishing between encoding models that capture stimulus-driven representations and those that overfit to individual heterogeneities.

Reviewer #1 (Public review):

Summary:

Huang et al. examined ACC response during a novel discrimination-avoid task. The authors concluded that ACC neurons primarily encode post-action variables over extended periods, reflecting the animal's preceding actions rather than the outcomes or values of those actions. The authors have made considerable revisions to address the raised concerns. However, it appears that some important issues remain unresolved.

Strengths:

The inclusion of new figures and analyses in response to the reviews is appreciated, such as Fig. 2 and 5.

Weaknesses:

Motion related signal in ACC: the new Fig. 2E looks good, but it is hard to visualize how it is just a reordering of the old Fig. 5C.

All categories in the new Fig. 4D appear to respond to shuttle initiation, with less than 1s latency. For example, type 2a/2b consists of 40% of the population and their response to movement onset is apparent. Thus, it is not clear whether most neurons respond to shuttle crossing as described in the manuscript.

Could the authors use relatively simple analysis, such as comparing spike rate before and after crossing, or before and after initiation, to quantify the response properties of each neuron? This could also help validate the classification analysis performed in Fig. 4.

Reviewer #2 (Public review):

Summary:

Huang et al recorded anterior cingulate cortex activity in mice while they performed a shuttle escape task. The task utilized two auditory cues, each of which informed the mice to stay or escape depending on which side they were on, and incorrect responses were punished by shock administration. Analyses focused on ACC neurons that fired when mice crossed the shuttle box in either direction (A-->B or B-->A), coined "action state", or when mice crossed in one direction but not the other, coined "action content". The authors characterized these populations, and ACC firing changes mostly occurred around the time of shuttle crossing. This work will likely be of broad interest to those who are interested in neocortical neurophysiology broadly, anterior cingulate cortex specifically, and their contributions to learning about actions. The task is well-designed and provides a nice background for neurophysiological recordings. The authors leveraged these strengths in characterizing the neural populations that fire to shuttle crossings in both directions vs one direction.

Strengths:

The factorial design nicely controls for sensory coding and value coding, since the same stimulus can signal different actions and values.

The figures are well presented, labeled, and easy to read.

Additional analyses, such as the 2.5/7.5s windows and place-field analysis, are nice to see and indicate that the authors were careful in their neural analyses.

The n-trial + 1 analysis where ACC activity was higher on trials that preceded correct responses is a nice addition, since it shows that ACC activity predicts future behavior, well before it happens.

The authors identified ACC neurons that fire to shuttle crossings in one direction or to crossings in both directions. This is very clear in the spike rasters and population scaled color images. While other factors such as place fields, sensory input, and their integration can account for this activity, the authors discuss this and provide additional supplemental analyses.

Weaknesses:

Some of the neural analyses could use the necessary and sufficient comparisons to strengthen the authors' claims.

Comment on revised version:

I think the authors did a very admirable job revising the manuscript. It is much improved. However, I believe a formal analysis of action-state versus action-content neurons on A-->B versus B-->A crossing is still warranted. I appreciate the fact that this analysis may not be as reliable with smaller ensemble sizes, but with careful pseudo-ensemble and resampling approaches, such an analysis would go a long way towards increasing the strength of evidence.

Reviewer #3 (Public review):

Summary:

The authors record from the ACC during a task in which animals must switch contexts to avoid shock as instructed by a cue. As expected, they find neurons that encode context, with some encoding of actions prior to the context, and encoding of neurons post-action. The primary novelty is dynamic encoding of action-outcome in a discrimination-avoidance domain, while this is traditionally done using operant methods.

Comments on revised version:

I appreciate the considerable work done on review, and additional details added throughout. I also noted the additional sessions included in analyses, and additional behavioral data in response to R1 and R2's insightful comments.

The only remaining comment that was not addressed pertains to anatomy and recording details. Some electrodes appear to be clearly in M2 (Fig 2A), and the tetrodes were driven each day. I would strongly suggest that this be included as a further limitation, particularly given the statement on line 178.

Reviewer #1 (Public review):

Summary:

An ongoing controversy in the field of learning and memory is the specific neural mechanism that maintains long-term memory (LTM). A prominent hypothesis proposed by Sacktor and Fenton and their colleagues is that LTM is maintained by the ongoing activity of the atypical PKC isoform PKMζ. Early evidence in support of this hypothesis came from experiments showing that an inhibitory peptide, ZIP, whose activity was purported to be specific for PKMζ, blocked late-phase hippocampal LTP (L-LTP) and LTM. However, in 2013, two articles reported that LTM was normal in PKMζ knockout mice and that ZIP erased LTM in the knockout mice, indicating that ZIP lacked specificity for PKMζ. In response, Sacktor and Fenton and colleagues reported in 2016 that in PKMζ null mice, there is an increase in the expression of PKC𝜾/λ, a related isoform of atypical PKC, and this increased expression can compensate for PKMζ; their data indicated that the upregulation of PKC 𝜾/λ mediates L-LTP and LTM in the PKMζ. In the present article, the authors provide additional support for this idea. They replicate the finding of an upregulation of PKC 𝜾/λ expression in the hippocampus of PKMζ knockout mice; in addition, they show that the expression of several other PKC isoforms is upregulated in the knockouts. They find that down-regulation of PKC𝜾/λ expression in the hippocampus using the Cre-LoxP technology, the 2016 paper merely used an inhibitor to block the activity of PKC𝜾/λ-blocks L-LTP. Finally, the authors demonstrate that, although LTM is preserved in the single PKMζ knockout mouse, it is eliminated in the PKMζ/PKC𝜾/λ double knockout mouse.

Strengths:

The experiments appear to have been carefully executed, the results reliable, and the paper well-written. Overall, the article provides significant additional support for the idea that the activity of PKMζ is critical for the maintenance of hippocampal L-LTP and LTM. The article uses genetic methods, rather than simply pharmacological ones, to demonstrate that when PKMζ is genetically deleted, PKC𝜾/λ, compensates for the missing PKCζ.

Weaknesses:

The paper sets up what I believe is probably a false dichotomy between a structural explanation - a change in the number of synaptic connections among neurons - and the persistent kinase activity explanation for memory maintenance. Why are these two explanations necessarily antithetical? It is possible that an increase in synaptic connections and the ongoing activity of PKMζ both contribute substantially to memory maintenance. The authors certainly don't provide any evidence that the number of synapses in the hippocampus remains unchanged after the induction of L-LTP or LTM. Indeed, I see no reason why persistent PKMζ activity could not be a mechanism for the maintenance of an enhanced number of synaptic connections following the induction of LTP/LTM. To the best of my knowledge, this possibility has not yet been explored. Consequently, I don't see why the present results would lead one to favor a biochemical explanation over a structural one for memory maintenance. Given the significant experimental evidence that LTM involves persistent structural changes in neurons, both explanations are equally plausible at present.

Reviewer #2 (Public review):

Summary:

The authors are attempting to advance understanding of the role of unconventional PKCs, PKCM𝛇, and PKC𝜄/𝝀 in maintenance of late-phase LTP. Their results help to clarify the interplay between "structural" and "biochemical/enzymatic" mechanisms of LTP and learning in the hippocampus.

Strengths:

A strength is the use of conditional knock-outs of PKCM𝛇 and PKC𝜄/𝝀 to assess the role of these two enzymes in maintaining long-term potentiation and in compensating for each other when one of them is conditionally knocked out in the adult.

Weaknesses:

The paper is extremely difficult to read because the abstract does not clearly state the advances made over earlier studies by the use of conditional KO mutation. For example, in line nine of the abstract, the authors state, "Here, we found PKC𝜄/𝝀 persists in LTP and long-term memory when PKM𝛇 is genetically deleted." This is confusing because it sounds as though the experiments have repeated earlier published experiments in which the gene encoding PKM𝛇 is deleted in the embryo. The authors are not clear throughout the manuscript that they are using conditional KO of the two enzymes in the adult animal, rather than deletion of the gene. The term "genetically deleted" does not mean "conditionally deleted in the adult." The final sentences of the abstract are: "Whereas deleting PKM𝛇 and PKC𝜄/𝝀 individually induces compensation, deleting both aPKCs abolishes hippocampal late-LTP. Hippocampal 𝜄/𝝀-𝛇 -double-knockout eliminates spatial long-term memory but not short-term memory. Thus, in the absence of PKM𝛇 , a second persistent biochemical process compensates to maintain late-LTP and long-term memory." These sentences do not convey a clear logical conclusion. The Discussion does a better job of stating the importance of the experiments.

Reviewer #3 (Public review):

Summary:

The manuscript addresses an important, yet unresolved and long-debated, question: whether atypical protein kinase C is required for the maintenance of late-long-term synaptic potentiation (L-LTP) and long-term memory (LTM). The authors confirm previous findings that persistent activity of PKMζ is required for hippocampal L-LTP and spatial memory. They demonstrate that genetically deleting PKCι/λ and PKMζ individually induces compensatory upregulation, whereas deleting both atypical PKCs abolishes hippocampal L-LTP spatial long-term memory. The study uses an elegant combination of immunoblots, electrophysiology, and behavioral assays. The use of Cre-recombinase to target specific hippocampal regions and neurons adds to the rigor of the findings.

Strengths:

The manuscript addresses an important, yet unresolved and long-debated, question; whether PKMζ is required for the maintenance of L-LTP and LTM. The study demonstrates that PKCι/λ, which was previously shown to be critical for the initial generation of the early phase of LTP and short-term memory, becomes persistently active in L-LTP and LTM in a PKMζ knock-out model, compensating for the loss of PKMζ. Furthermore, when the compensation mechanisms are eliminated by simultaneous deletion of both PKMζ and PKCι/λ, maintenance of LTP and long-term spatial memory, but not of short-term memory, is diminished. The strength of this study is that the authors used a double-knockout strategy to directly address the controversy concerning the roles of PKMζ in memory formation. By showing that PKCι/λ compensates when PKMζ is deleted, the authors provided a compelling explanation for previous contradictory findings.

Weaknesses:

(1) The authors should provide the numerical values for all data.

(2) It appears that blind procedures were only used for the behavioral experiments. Some explanation is warranted.

(3) The description of the immunoblotting procedures lacks sufficient detail. The authors state that immunoblots were stained with multiple antisera to visualize multiple PKCs on the same immunoblot. To conserve antisera, the immunoblots were cut to isolate the relevant proteins based on molecular weight. Isoforms with similar molecular weights were either stained with antisera of different species or on separate blots. Despite this explanation, it is unclear how immunoblotting was performed in practice. For example, in Figure 1B, the authors compared the changes of four conventional PKC isoforms. Because all four antibodies are mouse monoclonal antibodies recognizing proteins of similar molecular weights, each probing should presumably have its own actin loading controls. However, these controls are missing from the figure. Some clarification is warranted.

(4) The statement in the legend to Figure 4B, that the increases of maximum avoidance time from pretraining to trial 1 are not different, indicates both groups of mice successfully established short-term memory, which is not correct. The analysis only reveals that there is no difference between the two groups. No differences could be due to both groups learning the same, as the authors suggest, or alternatively to no learning in either group.

(5) The labeling on some of the illustrations (e.g., Figure 2B) is unreadable.

(6) In Figure 4B, only the single statistical comparison between "pretaining" and "1 trial" is shown. The other comparisons described in the legend should also be illustrated.

(7) There is no documentation to support the statement that "The prevailing textbook mechanism for how memory is retained asserts that stable structural changes at synapses, the result of initial protein synthesis and growth, sustain memory without the need for ongoing biochemical activity dedicated to storing information" or for the statement in the Discussion that the structural model of memory storage is the standard account.

Reviewer #1 (Public review):

Summary:

Ducrocq et al. present research exploring the genetic link between simple multicellular group formation (ace2Δ/ace2Δ) and its interaction with cell-cycle progression mutants (e.g., cln3Δ/cln3Δ), demonstrating that this combination can provide fitness benefits during fluctuating resource conditions, resulting in a rapid increase in the fraction of multicellular cell-cycle mutants over unicellular yeast without selection for multicellular size. Because both the multicellular phenotype and the regulatory link enabling faster escape from the stationary phase are controlled by the Ace2 transcription factor, this work demonstrates that multicellularity can arise as a side-effect of a completely independent fitness advantage unrelated to the benefits of group formation itself. As a "passenger phenotype," multicellularity could thus emerge for other selective reasons, potentially facilitating a later transition to more entrenched multicellularity if novel conditions arise where group formation becomes directly beneficial.

Strengths:

This work is novel and exciting for research exploring the very first steps of the transition from unicellularity to simple multicellularity. This is particularly significant because the formation of multicellular groups is almost always assumed to come at a cell-level fitness cost due to reduced reproductive fitness compared to remaining unicellular. This cell-level fitness cost generally needs to be outweighed by the benefits of multicellular group formation (e.g., large size escaping predation) for the multicellular phenotype to be stable, which is true for a large number of cases studied in the literature, where the multicellular phenotype can only evolve over unicellular competitors under strong selection for multicellular groups. However, this study presents an interesting case of a genetic and environmental condition under which individual cells (forming simple multicellular clusters) can actually have higher reproductive fitness than unicellular yeast. This demonstrates that the assumed cost at the single-cell level does not always apply. In summary, this work represents a unique example contrary to common assumptions regarding the costs of multicellular phenotypes, showing that simple multicellular phenotypes can evolve and remain stable without requiring strong selection for multicellular size or other benefits of group formation.

The claims and interpretation of the results align well with the data presented. This is due to the careful and straightforward experimental design testing predictions with a clear, stepwise methodology, ruling out alternative explanations and providing support for the proposed link between the mutations (ace2, cln3, and others), their impact on faster exit from quiescence, and thus earlier entry into reproduction in fresh media, resulting in higher fitness in the snowflake yeast phenotype compared to unicellular yeast.

Weaknesses:

The authors show that the same multicellular phenotype with higher cell-level fitness due to faster exit from the stationary phase can also be observed with alleles found at other loci in non-laboratory yeast strains, implying that the results are likely not specific to a peculiar case genetically engineered in laboratory strains, but that similar phenotypes may be present in nature. However, this remains to be explored further by examining the natural ecology of commercially available or wild yeast isolates and their genomes. This is by no means a weakness of this study and, therefore, not necessarily something the current work can improve. It does mean, however, that the relevance of these findings for early multicellularity in yeast, and even more so for nascent multicellularity in distinct taxa, remains to be explored in the future. Until then, it is difficult to make strong claims about how applicable these results would be for non-laboratory yeast and other taxa. Regardless, this work does its part by representing a very exciting finding.

Reviewer #2 (Public review):

Summary:

Here, the authors attempt to demonstrate that a simple model of multicellularity - snowflake yeast - exhibits key ecologically relevant changes in the regulation of the cell cycle. By examining the effects of the ace2 mutation in environments where multicellularity is not directly selected for or against, and combining it with mutations in key cell cycle regulators, they hope to show that mutations driving simple multicellularity can be selectively favored due to their effects on the release from quiescence rather than their effects on multicellularity itself.

Strengths:

The experiments performed are extensive and thorough. The yeast genotypes examined are judiciously chosen, so as to map out a functional model of the relationship between alterations to cell cycle control and changes to multicellularity phenotypes. Multiple possible interactions are examined, with the causal link and model of the relationship between the multicellular passenger phenotype and the selectable quiescence-release phenotype being well-supported. There are extensive controls demonstrating the separation between the 'passenger' multicellular phenotype and the cell cycle regulation phenotypes examined, including haploid/diploid strains with different multicellular phenotypes but similar cell cycle regulation phenotypes, and phenocopy strains in which downstream enzymes are deleted rather than key central regulators.

Weaknesses:

My only concerns about these results relate to the focus on selection on cell cycle control being examined in a model of multicellularity with key core cell cycle mutations rather than in a wild-type background, as this is a somewhat artificial system.

I believe, however, that the authors convincingly make their case that this work on the multicellular phenotypes of yeast represents a potent proof-of-concept that simple multicellularity can be driven into existence or selected for as a passenger phenotype due to pleiotropic effects of mutations under selection from real-world ecological pressures. They are able to connect this phenotype back to known mutations of particular cell cycle regulators (RB) in other multicellular lineages and demonstrate that ecologically relevant changes to the cell cycle are connected to multicellular phenotypes. As a proof of concept of the connection between these phenotypes, rather than a study of a particular event in the past of a living lineage, it makes a strong case.

A longstanding question in the field of multicellularity is the selective pressures that can drive simple multicellularity into existence and then act on simple multicells to drive their increased size and complexity. This work brings to the table tangible evidence of the possibility that, instead of being selected for on its own, simple multicellularity can be a side-effect of selection on other key phenotypes.

This separates the question of the origins of multicellularity and the forces that drive its further evolution. This separation can reframe how the field is studied, especially in the context of the apparent dichotomy between dozens of origins of 'simple' multicellularity across the tree of life and a few origins of 'complex' multicellularity in the history of Earth. Especially in light of other evidence that multicellularity is connected to changes in cell cycle regulation, I believe that this is an important insight that will alter the way we think about the origins of this key evolutionary transition.

Reviewer #1 (Public review):

In this paper, the authors use a doxycycline-inducible DLD1 cell line expressing a Clover-tagged RNA-binding-defective TDP-43 2KQ mutant that forms nuclear "anisosomes" (TDP-43 shell with HSP70 core) to carry out a small-molecule screen using the LOPAC 1280 library to identify compounds that reduce anisosome number or shift their morphology and dynamics. They also conducted a genome-wide siRNA screen to identify genetic modifiers of anisosome formation and dynamics. From these screens, the authors identify pathways in RNA splicing, translation, proteostasis (proteasome and HSP90), and nuclear transport, including XPO1. They then focus on XPO1 as their primary hit. Pharmacological inhibition of XPO1 using KPT-276, Verdinexor, and Leptomycin B reduces anisosome number while enlarging remaining condensates, which retain liquid-like behavior by FRAP and fusion assays. XPO1 overexpression causes fewer, enlarged TDP-43 puncta, including cytoplasmic puncta, with little or no FRAP recovery, interpreted as gel or solid-like aggregates. Anisosome induction reduces detectable nucleoplasmic XPO1 staining. Finally, the authors examine a homozygous TDP-43 K181E iPSC-derived forebrain organoid model, showing increased cytosolic pTDP-43 in K181E/K181E organoids compared to wild-type controls. Chronic low-dose KPT-276 reduces cytoplasmic pTDP-43 without changing total TDP-43 levels. Bulk RNA-seq shows only a modest fraction of dysregulated genes in K181E/K181E organoids are rescued by KPT-276. They conclude that nuclear export, via XPO1, is a key regulator of TDP-43 liquid-to-solid phase transitions and that cytoplasmic aggregation per se may contribute only modestly to TDP-43 proteinopathy, with RNA-processing defects being dominant.

The study presents well-executed chemical and genome-wide siRNA screens in a DLD1 TDP-43 2KQ anisosome model and follows up on nuclear transport, particularly XPO1, as a modulator of TDP-43 phase behavior and cytoplasmic aggregation. The screens are impressive in scale, and the microscopy and fluorescence recovery after photobleaching (FRAP) work is technically strong. However, the central mechanistic and disease-relevance claims are not yet sufficiently supported. There are major concerns about the heavy reliance on non-physiological, RNA-binding-defective, and acetylation-mimetic TDP-43 (2KQ) and a homozygous TDP-43 K181E organoid model. An underdeveloped and partly contradictory mechanistic link exists between XPO1 and TDP-43 phase transitions in the context of prior work showing TDP-43 is not a canonical XPO1 cargo. The paper also appears to overinterpret organoid data to conclude that cytoplasmic TDP-43 aggregation plays only a minor role in pathology, based largely on pTDP-43 antibody staining with limited sensitivity and relatively modest rescue readouts. A deeper mechanistic analysis and additional, more physiological validation are needed for this to reach the level of rigor and impact implied by the title and abstract. The work feels screen-rich but conceptually underdeveloped, with key claims outpacing the data. A major revision with substantial new data and tempering of conclusions is warranted. I outline several problematic areas below:

(1) The central mechanistic discoveries are derived almost entirely from a DLD1 colon cancer cell line overexpressing an RNA-binding-defective, acetylation-mimetic TDP-43 2KQ mutant and homozygous TDP-43 K181E iPSC-derived organoids. Both systems are far from physiological. The 2KQ mutation is a synthetic double lysine-to-glutamine mutant originally designed to mimic acetylation and disrupt RNA binding. In this study, essentially all cell-based mechanistic data on phase behavior, screens, and XPO1 effects rely on 2KQ. Yet there is no quantification of how much endogenous TDP-43 is acetylated in degenerating human neurons, nor whether a 2KQ-like acetylation state is ever achieved in vivo. It is not established that the phase behavior of 2KQ recapitulates the physiological or pathological phase behavior of wild-type TDP-43 or genuine disease-linked mutants, which may retain partial RNA binding and different post-translational modification patterns. As a result, it is difficult to know whether the modifiers identified here regulate a highly artificial 2KQ condensate or physiologically relevant TDP-43 condensates. To address this concern, the paper would benefit from quantifying endogenous TDP-43 acetylation at the relevant lysines in control and ALS/FTD patient tissue or more disease-proximal models such as heterozygous TARDBP mutant iPSC neurons, which would justify the focus on an acetyl-mimetic mutant. Key phenomena, including XPO1 dependence of phase behavior, effects of proteasome and HSP90 inhibition, and effects of splicing and translation inhibitors, should be tested for wild-type TDP-43 expressed at near-physiological levels and for one or more bona fide ALS/FTD-linked TARDBP mutants that are not acetyl mimetics. At a minimum, the authors should show that endogenous TDP-43 in neuronally differentiated cells exhibits qualitatively similar responses to XPO1 modulation, rather than exclusively relying on DLD1 2KQ overexpression.

(2) The organoid model is based on a homozygous K181E knock-in line. However, in patients, TARDBP mutations are overwhelmingly heterozygous. Homozygosity is thus a severe, arguably non-physiological sensitized background that may exaggerate nuclear RNA mis-splicing and phase defects and alter the relative contribution of cytoplasmic aggregation versus nuclear loss-of-function. In addition, it is not fully clear from this manuscript whether the structures in K181E organoids are bona fide anisosomes as defined in Yu et al. 2021, characterized by HSP70-enriched central liquid cores with TDP-43 shells and similar FRAP and fusion behavior to anisosomes in the DLD1 model. At present, the organoid section is framed as validation of "anisosome-bearing organoids," but the figures in this manuscript mainly show pTDP-43 puncta and total TDP-43 immunostaining, without detailed structural or biophysical characterization. The authors should explicitly compare heterozygous K181E/+ organoids or another heterozygous TARDBP mutant line with homozygous K181E/K181E organoids to assess whether XPO1 inhibition has similar effects in a genotype that more closely resembles patient genetics. They should provide direct evidence that the K181E condensates in organoids are anisosomes through HSP70 core immunostaining, three-dimensional reconstruction, and FRAP measurements, and clarify whether KPT-276 is acting on anisosome-like structures or more generic cytoplasmic aggregates or puncta. Without this, the leap from a DLD1 2KQ cancer cell model to human ALS/FTD-relevant neurons is not convincingly supported.

(3) The title and framing assert that "nuclear export governs TDP-43 phase transitions." However, prior studies such as Pinarbasi et al. 2018 and Duan et al. 2022 indicate that TDP-43 is not a canonical XPO1 cargo and that its export is largely passive, with active nuclear import being the dominant determinant of nuclear localization. The authors cite these studies but still position XPO1 as a central, quasi-direct regulator. The data presented are largely correlative or based on pharmacologic manipulation and overexpression in an overexpression mutant background, with no direct evidence that XPO1 engages TDP-43 in a specific, regulated manner. Even if XPO1 does not engage WT TDP-43, it could still engage the 2KQ variant, which needs to be tested.

(4) The XPO1 perturbations yield somewhat confusing phenotypes. XPO1 inhibition using Leptomycin B, KPT-276, and Verdinexor reduces anisosome number and enlarges remaining anisosomes, which remain liquid-like by FRAP recovery and fusion assays and stay nuclear. XPO1 overexpression causes fewer, enlarged puncta, but these are FRAP-impaired (gel-like) and redistribute to the cytoplasm. Thus, both decreased and increased XPO1 activity reduce anisosome number and enlarge puncta, but with opposite phase behaviors and subcellular localizations. The model presented in Figure 5L is relatively qualitative and does not resolve these issues. Moreover, XPO1 inhibition globally impairs nuclear export of many cargos and profoundly alters the nuclear environment, transcription, RNA processing, and chromatin. It is therefore difficult to conclude that the observed effects are specific to TDP-43 phase regulation as opposed to secondary consequences of broad nuclear export blockade.

(5) The authors show that anisosome induction depletes nucleoplasmic XPO1 signal and that mCherry-XPO1 can be seen in some TDP-43 puncta. However, antibody penetration into anisosomes is limited, so XPO1 depletion from nucleoplasm could reflect sequestration in the anisosome shell or core, but this is not demonstrated. There is no demonstration of physical interaction, even indirect interaction, between XPO1 and TDP-43 or a defined adaptor, nor identification of a specific mutant of XPO1 that selectively disrupts this putative interaction while preserving other functions. The known TDP-43 NES has been shown to be weak and not a functional XPO1-dependent NES in multiple studies. If XPO1 is acting through an adaptor that recognizes 2KQ or K181E specifically, that by itself would bring into question the generality of the mechanism for wild-type TDP-43.

(6) To support a mechanistic claim that nuclear export governs TDP-43 phase transitions, more targeted evidence is needed. The authors should test whether siRNA knockdown or CRISPR interference of XPO1 in the DLD1 2KQ model reproduces the effects seen with Leptomycin B and KPT-276, including FRAP and fusion phenotypes, and verify on-target effects by rescue with an siRNA-resistant XPO1 construct. They should demonstrate that canonical XPO1 cargos behave as expected under the inhibitor conditions used, as a positive control, and that the concentrations used are not grossly toxic. They should attempt to identify or at least constrain candidate adaptors that might enable XPO1-dependent export of TDP-43 through proteomic analysis of XPO1 co-purifying with 2KQ condensates or loss-of-function studies of candidate adaptors from the siRNA screen. Finally, they should test whether a TDP-43 mutant that cannot bind the proposed adaptor still responds to XPO1 manipulation.

(7) Even with these data, what is currently shown is that global modulation of nuclear export capacity can alter the phase behavior and localization of a highly overexpressed RNA-binding-defective TDP-43 mutant and of K181E in organoids. This is important, but it is weaker than asserting that XPO1 directly governs TDP-43 phase transitions in physiological contexts. The title, abstract, and Discussion should be tempered to reflect that nuclear export is one of several pathways, alongside RNA splicing, translation, and proteostasis, that influence TDP-43 phase states in this model, and that the specific mechanism and cargo relationship between XPO1 and TDP-43 remain unresolved and may be indirect.

(8) The authors conclude that cytoplasmic TDP-43 aggregation plays only a modest role in TDP-43 proteinopathies because in homozygous K181E organoids, chronic KPT-276 treatment almost abolishes cytoplasmic pTDP-43 puncta, yet bulk RNA-seq shows only a relatively small fraction of dysregulated genes are rescued. There are several issues with this inference. Relying primarily on pTDP-43 antibody staining to define cytoplasmic TDP-43 aggregation is limiting. pTDP-43 antibodies label only phosphorylated species and may miss non-phosphorylated, oligomeric, or amorphous TDP-43 species that could still be toxic. Different pTDP-43 antibodies vary in epitope accessibility depending on aggregate conformation and subcellular location. More sensitive approaches, such as high-affinity TDP-43 RNA aptamer probes developed by Gregory and colleagues, biochemical fractionation for SDS-insoluble and urea-soluble TDP-43, and filter-trap assays, would provide a more quantitative assessment of cytoplasmic aggregation and its reduction by KPT-276. Without these, it is not safe to assume that cytoplasmic aggregation has been eliminated, as opposed to one antigenic subclass.

(9) The treatment window, spanning from day 87 to 122 with 20 nanomolar KPT-276, may be too late or too mild to reverse entrenched nuclear RNA-processing defects, even if cytoplasmic inclusions are cleared. Once widespread cryptic exon inclusion and alternative polyadenylation misregulation are established, many downstream changes may become self-sustaining or only partially reversible. Moreover, XPO1 inhibition will massively rewire nucleocytoplasmic transport of many transcription factors, splicing factors, and RNA-binding proteins. Thus, the lack of full transcriptomic rescue cannot be cleanly interpreted as evidence that cytoplasmic aggregates are only modest contributors. It may instead reflect that nuclear dysfunction is primary and XPO1 inhibition does not correct, and may even exacerbate, certain nuclear defects.

(10) To support a causal statement about the modest contribution of cytoplasmic aggregates, one would want more direct measures of neuronal health and function, such as cell death, neurite complexity, synaptic markers, and electrophysiology before and after KPT-276, not only transcriptomics. A way to selectively reduce cytoplasmic aggregation without globally inhibiting nuclear export would allow comparison of outcomes.

(11) Given these caveats, the concluding statements that cytoplasmic TDP-43 aggregation is only a modest contributor should be substantially softened. A more defensible interpretation is that in this homozygous K181E organoid model, chronic global XPO1 inhibition reduces pTDP-43-positive cytoplasmic puncta but only partially normalizes the steady-state transcriptome, suggesting that persistent nuclear RNA-processing defects and other pathways continue to drive pathology.

(12) The screens are a major strength but need more rigorous validation for key hits, especially nuclear transport factors. For the siRNA screen, hits are filtered by anisosome number per nucleus, but there is no direct demonstration in the main text that XPO1 or CSE1L knockdown is efficient at the messenger RNA or protein level. For the highlighted genes, Western blot or quantitative polymerase chain reaction validation and phenotypic rescue would strengthen confidence. For small-molecule hits, it is not systematically shown that anisosome modulation is independent of changes in total TDP-43 2KQ expression or gross toxicity. Translation inhibitors are tested for this, but for many other hits, including proteasome, HSP90, and kinase inhibitors, expression and general nuclear structure should be monitored. Given the reliance on anisosome count as a readout, secondary screens that specifically distinguish changes in TDP-43 expression levels, changes in nuclear morphology or cell cycle, and specific changes in anisosome phase behavior, including FRAP and fusion for top hits, would greatly increase interpretability.

(13) The classification of condensates as liquid versus gel-like or solid is based almost entirely on FRAP recovery or lack thereof. While FRAP is appropriate, interpretations could be made more robust by including half-region-of-interest bleach controls and assessing mobile fractions and recovery kinetics more quantitatively across conditions. Complementing FRAP with other phase-behavior assays such as sensitivity to 1,6-hexanediol, shape relaxation after deformation, and coarsening behavior over longer timescales would strengthen the analysis. At present, some assignments, such as that XPO1 overexpression drives a gel-like transition, are reasonable but somewhat qualitative.

(14) For the Leptomycin B and KPT-276 experiments in cells and organoids, it would be important to confirm that canonical XPO1 cargo proteins accumulate in the nucleus and that the concentrations used are within a range that is not overtly toxic over the experimental timeframe. Assessing nuclear morphology, chromatin condensation, and general transcriptional activity through global RNA synthesis or key reporter genes would ensure that observed effects are not secondary to severe global nuclear export collapse.

(15) In the organoid section, it is not clear how many independent iPSC clones and organoid batches were used per condition, nor whether batch effects were assessed in the bulk RNA-seq analysis. This should be fully specified and ideally controlled with isogenic wild-type and K181E clones. For transcriptional rescue, it is important to know whether the changes in wild-type organoids treated with KPT-276 are negligible. A direct wild-type comparison with or without KPT-276 is important to disentangle general drug effects from K181E-specific rescue. More detailed quantification of total TDP-43 and pTDP-43 in both nuclear and cytoplasmic fractions, including biochemical fractionation if possible, would strengthen the assertion that KPT-276 specifically reduces cytosolic pTDP-43 aggregates while sparing nuclear TDP-43.

(16) Beyond the core issues above, several additions could greatly enhance the impact. The manuscript currently emphasizes XPO1, but the genetic and chemical data clearly implicate RNA splicing, translation, and proteostasis as equally strong or stronger regulators of TDP-43 phase states. A more integrated model that explains how these pathways intersect, for example, how splicing factor availability, ribosome loading, and proteasome capacity co-govern anisosome nucleation, growth, and hardening, would be valuable.

(17) A key unresolved question is whether XPO1 is acting directly on TDP-43, or instead primarily regulates anisosomes by exporting other factors that more proximally control TDP-43 phase behavior. Given that TDP-43 is not a canonical XPO1 cargo and prior work indicates that its nuclear export is largely passive, it seems at least as plausible that XPO1 inhibition alters the nuclear concentration or localization of splicing factors, RNA-binding proteins, chaperones, or other modifiers identified in the screens, and that changes in these proteins secondarily reshape anisosome dynamics. In other words, XPO1 may be exporting a more direct regulator of anisome formation and hardening, rather than exporting TDP-43 itself in a specific, regulated way. The current data do not distinguish between these possibilities. Systematic identification of XPO1-dependent cargos that colocalize with or biochemically associate with anisosomes, combined with targeted perturbation of their nuclear export, would be needed to determine whether the relevant XPO1 substrate in this system is actually TDP-43 or an upstream modulator of its phase behavior.

(18) Testing whether identified modifiers converge on nuclear TDP-43 concentration would be informative. Since phase separation is concentration-dependent, measuring nuclear versus cytoplasmic TDP-43 levels across key perturbations, including splicing inhibition, translation inhibition, proteasome inhibition, HSP90 inhibition, and XPO1 modulation, would help determine whether modifiers mainly work by changing nuclear TDP-43 concentration or by altering interaction networks and the material properties of condensates.

(19) Examining other ALS-relevant RNA-binding proteins would be valuable. Given the role of XPO1 and other hits, it would be informative to briefly test whether similar principles apply to FUS, hnRNPA1, or other ALS-relevant RNA-binding proteins in the same cellular context, to argue for generality versus TDP-43-specific idiosyncrasies of the 2KQ system.

(20) The Introduction sometimes implies that anisosomes are common and well-established intermediates en route to pathology. It would be helpful to more clearly state that, to date, anisosomes are primarily observed in overexpression and mutant systems and have not yet been unequivocally demonstrated in human patient tissue. The link between PDGFRβ, PAK4, GSK-3β, and YAP and TDP-43 phase dynamics is intriguing but only briefly mentioned. The authors should either expand on this or tone down the emphasis in the Results section.

(21) In the organoid methods, the authors should consider clarifying whether doxycycline is continuously used, which might alter TDP-43 expression and nuclear transport in a non-negligible way.

(22) For statistical methods, it would be beneficial to indicate whether multiple-comparison corrections were applied for the many FRAP, anisosome count, and size comparisons beyond DESeq2 internal corrections for RNA-seq.

(23) Some figure legends could more clearly indicate whether the images shown are single z-planes or maximum intensity projections and how the thresholding for anisosome detection was performed.

(24) In its current form, the manuscript contains an impressive set of screens and some nicely executed imaging of TDP-43 condensates, highlighting nuclear export among other pathways as a modulator of TDP-43 phase behavior. However, the physiological relevance is undercut by heavy reliance on an acetylation-mimetic, RNA-binding-defective TDP-43 mutant and a homozygous K181E organoid model. The mechanistic link between XPO1 and TDP-43 remains largely inferential and partly at odds with prior work. The conclusion that cytoplasmic TDP-43 aggregation is only a modest contributor to disease is not firmly supported by the available data.

(25) With substantial additional mechanistic work, particularly around XPO1, rigorous validation in more physiological TDP-43 contexts, more sensitive detection of cytoplasmic TDP-43 aggregates, and a tempering of the central claims, this study could make a meaningful contribution to understanding how nucleocytoplasmic transport and other cellular pathways influence TDP-43 phase transitions and aggregation. The work should be reframed as an important screening study that identifies nuclear export as one among several cellular processes that modulate TDP-43 phase behavior in a model system, rather than as a definitive demonstration that nuclear export governs pathological TDP-43 aggregation in disease.

Reviewer #2 (Public review):

Summary:

This manuscript addresses an important and timely question in TDP-43 biology by systematically identifying regulators of TDP-43 anisosome formation, with a particular focus on nuclear export via XPO1. Using a combination of unbiased chemical screening, genetic perturbation, and advanced imaging approaches, the authors propose that inhibition of nuclear export modulates the abundance and biophysical properties of TDP-43 anisosomes. The study is conceptually innovative and has potential relevance for neurodegenerative diseases characterized by TDP-43 pathology. However, significant concerns regarding experimental controls, reporting transparency, and model translatability currently limit the strength of the conclusions and the interpretability of several key findings.

Strengths:

(1) The study employs an unbiased, hypothesis-free compound screen to identify regulators of TDP-43 anisosome formation, which is a major strength and reduces confirmation bias.

(2) The authors combine chemical and genetic screening approaches, providing orthogonal validation of key pathways and increasing confidence in the biological relevance of top hits.

(3) The focus on biophysical properties of TDP-43 assemblies, assessed through imaging and FRAP, moves beyond simple presence/absence of aggregates and provides mechanistic insight into the biophysical states of TDP-43.

(4) The use of multiple experimental modalities, including live-cell imaging, FRAP, pharmacological perturbation, and transcriptomic analysis, reflects a technically sophisticated and ambitious study design.

(5) The authors attempt to extend findings beyond immortalized cancer cell lines by incorporating organoid models, demonstrating awareness of disease relevance and translational importance.

Overall, the manuscript is clearly written and logically structured, making complex experimental workflows accessible and the central hypotheses easy to follow.

Weaknesses:

Despite its strengths, the manuscript has several major limitations that affect data interpretation and confidence in the conclusions.

(1) Lack of appropriate controls for overexpression experiments:

A central concern is the absence of proper controls for TDP-43 and XPO1 overexpression. Prior studies (including those cited by the authors, Archbold et al.2018) show that overexpression of WT TDP-43 alone is toxic to neurons. Thus, the experimental system itself may induce anisosome formation independently of the mechanisms under study. Similarly, XPO1 overexpression lacks a suitable control (e.g., mCherry alone or mCherry fused to a protein known to be independent of TDP-43). The near-complete colocalization of XPO1 with TDP-43 anisosomes upon overexpression raises the possibility that these structures reflect non-physiological protein accumulation rather than regulated assemblies.

2) Insufficient experimental and analytical transparency:

The manuscript frequently lacks clear reporting of experimental details. In multiple figures, the stated number of independent experiments does not match the number of data points shown, making it difficult to assess statistical validity. Concentrations used in the compound screen are not clearly defined, nor is it stated whether multiple concentrations were tested. It is unclear how many wells, cells, or independent cultures were analyzed. The criteria used to reduce 1,533 screening hits to 211 candidates via STRING analysis are not explained. Knockdown and overexpression efficiencies are not reported.

(3) RNA-seq concerns:

The RNA-seq experiments are particularly problematic. The number of biological replicates per condition is not stated, and heatmaps suggest that only one sample per group may have been used, which would preclude statistical analysis. No baseline comparison between WT and mutant TDP-43 is shown. Given that TDP-43 is an RNA-binding protein, splicing analyses would be far more informative than gene expression alone, yet no splicing data are presented. Moreover, nuclear retention of TDP-43 does not preclude nuclear aggregation, which may still impair its splicing function.

(4) Limited translatability to neuronal biology:

All anisosome analyses are performed in a cancer cell line, raising concerns about relevance to post-mitotic neurons. While organoids are used as a secondary model, the assays performed do not overlap with those used in cancer cells, making it difficult to assess whether anisosome-related mechanisms are conserved. Neuronal toxicity, a critical outcome given known TDP-43 biology, is not assessed. Prior work has shown that WT TDP-43 overexpression alone is toxic to neurons, yet this is not addressed.

(5) Conceptual and interpretational gaps:

The authors quantify anisosome number but also report conditions in which anisosome number decreases while size increases. The biological interpretation of larger anisosomes is not discussed, and whether this reflects improvement or worsening of pathology is unclear. Compounds targeting the same mechanism (e.g., nuclear export inhibition) are inconsistently used across experiments (KPT compounds, verdinexor, leptomycin B), raising concerns about reproducibility. In organoids, the experimental paradigm shifts to long-term treatment (35 days vs. 16 hours), further complicating interpretation.

(6) Overinterpretation of rescue effects:

Although the authors state that they aim to test whether nuclear export inhibition rescues neuronal defects, no functional neuronal readouts are provided (e.g., viability, morphology, axon outgrowth, or electrophysiological measures). RNA-seq alone is insufficient to support claims of rescue.

(7) Finally, the model does not appear to exhibit cytosolic TDP-43 aggregation at baseline. It remains unclear whether longer induction would produce cytosolic gel-like assemblies and whether these would be prevented by nuclear export inhibition. Long-term data are shown only in organoids, yet anisosome formation is not assessed there.

Reviewer #3 (Public review):

Summary:

TDP-43 proteinopathy is broadly found in neurodegenerative diseases. This manuscript investigates how nuclear export influences the biophysical properties of TDP-43. The authors use a combination of chemical screening and genome-wide siRNA screening to identify pathways that modulate TDP-43 liquid-to-solid transitions. Overall, the study employs a broad array of approaches and addresses an important question in TDP-43 pathobiology. The identification of nuclear export as a central regulator is compelling and conceptually aligns with the emerging view that TDP-43 nucleocytoplasmic trafficking is a major defect in neurodegeneration.

Strengths:

This work integrates chemical and genetic screening to identify novel modifiers. The candidates were validated in both reporter cell lines and iPS-differentiated organoids. The findings support the nucleocytoplasmic transport is important for the biophysical properties of TDP-43.

Weaknesses:

The mechanisms underlying the connection between nuclear export and phase transition need further clarification. Broader consequences of XPO1 inhibition are not addressed.

Reviewer #1 (Public review):

I enjoyed reading this long but compelling account of the new (generalised) version of the Hierarchical Gaussian filter (HGF). Effectively, it describes an extension of the HGF to accommodate the influence of latent states on volatility - and vice versa. This paper describes a generalisation that has been made available to the community via the TAPAS software. This contribution will be of special interest to people in computational psychiatry, where the application of the HGF has been the most prevalent.

I thought the background, motivation, description and illustration of the scheme were excellent. The paper is rather long; however, it serves as a useful technical reference.

There are two issues that I think the authors need to address.

(1) The first is the failure to properly relate the current scheme to standard implementations of Bayesian filtering under hierarchical state-space models.

(2) The second is that whilst the paper is well-written, some of the mathematical notation is cluttered. Furthermore, I think that the authors need to motivate the otherwise overengineered description of the requisite variational message passing and decomposition into update steps.

I think that the authors can address both of these issues by including a technical section in the introduction, relating the HGF to state-of-the-art in the broader field of Bayesian filtering and predictive coding. They can then explain the benefits of the particular generative model - to which the HGF is committed - by drilling down on the update scheme and its implementation in the remainder of the paper.

I was underwhelmed by the account of predictive coding and its relationship to Bayesian filtering. I think that the authors should suppress the references to predictive coding in the recent machine learning literature. Rather, the presented narrative should emphasise the fact that predictive coding and Bayesian filtering are the same thing. The authors could then explain where the hierarchical Gaussian filter fits within Bayesian filtering and why its particular form lends itself to the variational updates they subsequently derive.

The authors could add something like the following to the introduction (accompanying PDF has the equations). There is a summary of what follows in the Wikipedia entry on generalised filtering, in particular, its relationship to predictive coding (https://en.wikipedia.org/wiki/Generalized_filtering).

Relationship to Existing Work

Technically, the hierarchical Gaussian filter is a Bayesian filter under a hierarchical state-space model. The most general form of these models can be expressed as stochastic differential or difference equations as follows, c.f., Equation 9 in (Feldman and Friston, 2010):

This functional form implies a hierarchical decomposition into hierarchical levels (l) that are linked through latent causes (v), with dynamics among latent states (x) at each level. From the perspective of the HGF, the state-dependency of state (z) and observation (e) noise at each level is a key feature. The variance (i.e., inverse precision) of the random fluctuations z is known as volatility, which - in a hierarchical setting - can depend upon latent causes and states at higher levels. The variational inversion of these models - sometimes called variational or generalised filtering - finds a number of important applications: a key example here is dynamic causal modelling, typically in the analysis of imaging timeseries. In this setting, unknown or latent states, parameters and precisions are updated in variational steps by minimising variational free energy (a variational bound on negative log marginal likelihood).

In engineering, the simplest form of generalised filtering is known as a Kalman filter, in which all the equations are linear, and volatility is assumed to be constant. In neurobiology, there is an intimate relationship between generalised filtering and predictive coding: predictive coding was originally introduced for timeseries analysis and compression of sound files (Elias, 1955). Subsequently, the implicit filtering or compression scheme was considered as a description of neuronal processing in the retina (Srinivasan et al., 1982) and then cortical hierarchies (Mumford, 1992; Rao, 1999; Rao and Ballard, 1999). The formal equivalence between predictive coding and Kalman filtering was noted in (Rao, 1999). Kalman filtering itself was then recognised as a special case of generalised filtering that could be read as predictive coding in the brain (Friston and Kiebel, 2009). The estimation of precision in these predictive coding schemes has been associated with endogenous (Feldman and Friston, 2010) and exogenous (Kanai et al., 2015) attention; i.e., with and without state dependency, respectively. Subsequently, precision estimation or uncertainty quantification has become a key focus in computational psychiatry.

In machine learning, there have been recent attempts to implement predictive coding via the minimisation of variational free energy under generative models with the functional form of conventional neural networks: e.g., (Millidge et al., 2022; Salvatori et al., 2022). However, much of this work is nascent and does not deal with dynamics or volatility. There is an interesting exception in machine learning, namely, transformer architectures, where the attention heads can be read as implementing a form of Kalman gain, namely, estimating state-dependent precision, e.g., (Buckley and Singh, 2024).

Within this general setting, the HGF emphasises the importance of precision estimation or uncertainty quantification by committing to a particular functional form for the generative model that can be summarised as follows:

"We will unpack this form below and show how it leads to a remarkably compact and efficient Bayesian belief updating scheme. We will appeal implicitly to variational message passing on factor graphs (Dauwels, 2007; Friston et al., 2017; Winn and Bishop, 2005) to decompose message passing between nodes and, crucially, within-node computations. These computations furnish a scalable and flexible form of generalised Bayesian filtering. In principle, this scheme inherits all the biological plausibility of belief propagation and variational message passing in cortical hierarchies (Friston et al., 2017)."

It might be worth the authors [re-]reading the abstracts of the above papers, for a clearer sense of how those in computational neuroscience and state-space modelling (but not machine learning) think about predictive coding and its relationship to Bayesian filtering. They could then go through the manuscript, nuancing your discussion of the intimate relationship between variational Bayes, generalised filtering, predictive coding and hierarchical Gaussian filtering.

Reviewer #2 (Public review):

Summary:

The authors introduce a generalised HGF featuring (1) volatility coupling (rate of change), value coupling (phasic or autoregressive drift) [and 'noise coupling', which is a volatility parent of an outcome state] (2) parameters: volatility coupling κ, tonic volatility ω, value coupling α, tonic drift ρ, {plus minus}auto-regressive drift λ (3) inputs at irregular intervals (but still discrete time steps, unlike continuous time belief evolution in predictive coding) (4) states with multiple parents or parents with multiple child states (5) value parents by default have a volatility parent, and volatility parents have a value parent (or none) (6) linear or non-linear (including ReLU) functions (7) also beliefs can be any exponential family distribution (incl binary, categorical), hence can also model POMDPs

They describe the 3 steps involved in updating (for both value and volatility): (1) prediction (2) update posterior (entails passing both pwPE and prediction precision from lower to upper node - the latter is not found in other predictive coding schemes) (3) prediction error NB this makes the network modular, so nodes can be added/removed without recomputing all the update equations.

They give some examples of models working using simulated data: (1) sharing of parent nodes can generalise an update from one context to another (2) sharing of child nodes enables multisensory cue combination (e.g. auditory-visual, or interoceptive-exteroceptive).

The authors further discuss a potential shortcoming of the HGF - its discretisation of timesteps - which is less naturalistic but nevertheless makes it very amenable to fitting trial-wise experimental data. They propose to extend the HGF to modelling within-step dynamics in future, which could make testable continuous time neuronal predictions.

Strengths:

Overall, I think the paper is excellent - it contributes an important extension to a popular modelling tool which substantially increases the number of potential applications. It is well written, and I have almost no criticisms to make.

Weaknesses:

The authors state that this generalised HGF will "make it easy to build large networks with considerable hierarchical depth", comparable to neural network architectures. The examples they give are extremely simple; however, it would be good to see a more complex one.