eLife assessment

This important study describes an atypical role of the odorant binding protein Obp56g in mating plug formation in Drosophila melanogaster suggesting that Obps may play roles in reproduction in addition to their originally described roles in olfaction. Mutant males lacking Obp56g fail to induce the formation of a mating plug in the female reproductive tract-leading to ejaculate loss and reduced sperm storage. The evidence supporting the claims of the authors is solid and compelling. The work will be of interest to biologists studying Obps and seminal fluid protein function and their evolution.

Reviewer #1 (Public Review):

This study characterized the role of the Drosophila odorant-binding protein Obp56g in mediating post-mating responses in females. The authors show that this Obp56g is expressed in the male ejaculatory bulb, use genetic approaches to disrupt Obp56g, and show that males with disrupted Obp56g fail to form mating plugs in their mates.

Despite the fact that Obp56g deficient males generate and transfer sperm normally, the lack of formation of a functional mating plug leads to sperm loss in females mated to these males and thus a failure to induce normal post-mating responses.

This study identifies an unexpected role for an Obp and raises new questions about the function of this class of protein and the variety of roles that they may play in sensory function and reproduction.

Reviewer #2 (Public Review):

Here, Brown and colleagues report a valuable finding on the function and evolution of the seminal odorant-binding protein Obp56g in Drosophila melanogaster. Previous studies have shown that this family of proteins is highly expressed in olfactory tissues like the antennae and maxillary palps. Some of these proteins have been shown to mediate behavioural responses to specific odorants-hence the general moniker odorant binding proteins. This slightly misleading historical naming convention implies an exclusive role in olfaction-however, many of these proteins are expressed in other tissues of the animal, including the male reproductive system. In addition, seminal fluid proteins exhibit a fascinating evolutionary history, with rapid evolution and turnover across taxa.

The authors suggest that the Obp56g protein may have been co-opted for a reproductive role in Drosophila melanogaster during evolution. The authors show that Obp56g is required for male fertility and the induction of the post-mating response in females. Mutant males lacking Obp56g fail to form a mating plug in the female reproductive tract-leading to ejaculate loss and reduced sperm storage. The experimental evidence supporting the claims of the authors is solid and compelling. The data were collected and analyzed using solid and validated methodologies. The author's findings can be used as a starting point for understanding the mechanistic roles of this family of proteins in mating plug coagulation. The work will interest biologists studying non-sperm seminal fluid protein function and evolution.

Reviewer #3 (Public Review):

Male seminal fluid proteins play a crucial role in fertility and influence female physiology and behavior after mating. Brown et al. have discovered a new reproductive function for odorant-binding proteins (Obps) in Drosophila. The study shows that Obp56g is expressed in male reproductive tissues that produce seminal fluid proteins and is required for the formation of the mating plug in the mated female. The study demonstrates that RNAi knockdown and CRISPR/Cas9-generated mutations in Obp56g result in a defective mating plug, reduced sperm storage, and subsequent effects on female post-mating responses. The research also suggests that Obp56g has been co-opted for a reproductive function over evolutionary time, as supported by functional and comparative RNAseq data across Drosophila species. Finally, the study reports expression shifts, duplication, and divergence in the evolution of these seminal protein genes.

Overall, the study represents a significant contribution to our understanding of seminal proteins and their reproductive function. The creation of novel Obp mutants using CRISPR/Cas9 technology is a valuable resource for future research in the Drosophila community. The manuscript successfully conveys the key findings and their potential implications for the field. However, to reinforce the study's conclusions, more quantitative data is necessary. Furthermore, improving the statistical analysis and incorporating additional genetic controls would enhance the quality of the study and provide a stronger foundation for its conclusions.

eLife assessment

This work is relevant to understanding how people represent uncertain events in the world around them and make decisions, with broad applications to economic behavior. It addresses a long-standing empirical puzzle from a novel perspective, where the authors propose that sequential effects in perceptual decisions may emerge from rational choices under cognitive resource constraints rather than adjustments to changing environments. Two new computational models have been constructed to predict behavior under two different constraints, among which the one assuming higher cost for more precise beliefs is better supported by new experimental data. The conclusion may be further strengthened by comparison with alternative models and (optionally) evidence from additional data.

Reviewer #1 (Public Review):

In this paper, the authors develop new models of sequential effects in a simple Bernoulli learning task. In particular, the authors show evidence for both a "precision-cost" model (precise posteriors are costly) and an "unpredictability-cost" model (expectations of unpredictable outcomes are costly). Detailed analyses of experimental data partially support the model predictions.

Strengths:<br /> - Well-written and clear.<br /> - Addresses a long-standing empirical puzzle.<br /> - Rigorous modeling.

Weaknesses:<br /> - No model adequately explains all of the data.<br /> - New empirical dataset is somewhat incremental.<br /> - Aspects of the modeling appear weakly motivated (particularly the unpredictability model).<br /> - Missing discussion of some relevant literature.

Reviewer #2 (Public Review):

This paper argues for an explanation of sequential effects in prediction based on the computational cost of representing probability distributions. This argument is made by contrasting two cost-based models with several other models in accounting for first- and second-order dependencies in people's choices. The empirical and modeling work is well done, and the results are compelling.

The main weaknesses of the paper are as follows:

1. The main argument is against accounts of dependency based on sensitivity to statistics (ie. modeling the timeseries as having dependencies it doesn't have). However, such models are not included in the model comparison, which makes it difficult to compare these hypotheses.

2. The task is not incentivized in any way. Since incentives are known to affect probability-matching behaviors, this seems important. In particular, we might expect incentives would trade off against computational costs - people should increase the precision of their representations if it generates more reward.

3. The sample size is relatively small (20 participants). Even though a relatively large amount of data is collected from each participant, this does make it more difficult to evaluate the second-order dependencies in particular (Figure 6), where there are large error bars and the current analysis uses a threshold of p < .05 across a large number of tests hence creating a high false-discovery risk.

4. In the key analyses in Figure 4, we see model predictions averaged across participants. This can be misleading, as the average of many models can produce behavior outside the class of functions the models themselves can generate. It would be helpful to see the distribution of raw model predictions (ideally compared against individual data from humans). Minimally, showing predictions from representative models in each class would provide insight into where specific models are getting things right and wrong, which is not apparent from the model comparison.

Reviewer #3 (Public Review):

This manuscript offers a novel account of history biases in perceptual decisions in terms of bounded rationality, more specifically in terms of finite resources strategy. Bridging two works of literature on the suboptimalities of human decision-making (cognitive biases and bounded rationality) is very valuable per se; the theoretical framework is well derived, building upon the authors' previous work; and the choice of experiment and analysis to test their hypothesis is adequate. However, I do have important concerns regarding the work that do not enable me to fully grasp the impact of the work. Most importantly, I am not sure whether the hypothesis whereby inference is biased towards avoiding high precision posterior is equivalent or not to the standard hypothesis that inference "leaks" across time due to the belief that the environment is not stationary. This and other important issues are detailed below. I also think that the clarity and architecture of the manuscript could be greatly improved.

1. At this point it remains unclear what is the relationship between the finite resources hypothesis (the only bounded rationality hypothesis supported by the data) and more standard accounts of historical effects in terms of adaptation to a (believed to be) changing environment. The Discussion suggests that the two approaches are similar (if not identical) at the algorithmic level: in one case, the posterior belief is stretched (compared to the Bayesian observer for stationary environments) due to precision cost, in other because of possible changes in the environment. Are the two formalisms equivalent? Or could the two accounts provide dissociable predictions for a different task? In other words, if the finite resources hypothesis is not meant to be taken as brain circuits explicitly minimizing the cost (as stated by the authors), and if it produces the same type of behavior as more classical accounts: is the hypothesis testable experimentally?

2. The current analysis of history effects may be confounded by effects of the motor responses (independently from the correct response), e.g. a tendency to repeat motor responses instead of (or on top of) tracking the distribution of stimuli.

3. The authors assume that subjects should reach their asymptotic behavior after passively viewing the first 200 trials but this should be assessed in the data rather than hypothesized. Especially since the subjects are passively looking during the first part of the block, they may well pay very little attention to the statistics.

4. The experiment methods are described quite poorly: when is the feedback provided? What is the horizontal bar at the bottom of the display? What happens in the analysis with timeout trials and what percentage of trials do they represent? Most importantly, what were the subjects told about the structure of the task? Are they told that probabilities change over blocks but are maintained constant within each block?

Author Response

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

Reviewer #1 (Public Review):

This work challenges previously published results regarding the presence and abundance of 6mA in the Drosophila genome, as well as the claim that the TET or DMAD enzyme serves as the "eraser" of this DNA methylation mark and its roles in development. This information is needed to clarify these questions in the field. I am less familiar with the biochemical approaches in this work, so my comments are mainly on the genetic analyses. Generally speaking, the methods for fly husbandry and treatment seem to be in accordance with those established in the field.

Response : We thank the reviewer for his/her work and positive assessment of our manuscript.

Reviewer #2 (Public Review):

DNA adenine methylation (6mA) is a rediscovered modification that has been described in a wide range of eukaryotes. However, 6mA presence in eukaryote remains controversial due to the low abundance of its modification in eukaryotic genome. In this manuscript, Boulet et al. re-investigate 6mA presence in drosophila using axenic or conventional fly to avoid contaminants from feeding bacteria. By using these flies, they find that 6mA is rare but present in the drosophila genome by performing LC/MS/MS. They also find that the loss of TET (also known as DMAD) does not impact 6mA levels in drosophila, contrary to previous studies. In addition, the authors find that TET is required for fly development in its enzymatic activity-independent manner.

The strength of this study is, that compared to previous studies of 6mA in drosophila, the authors employed axenic or conventional fly for 6mA analysis. These fly strains make it possible to analyze 6mA presence in drosophila without bacterial contaminant. Therefore, showing data of 6mA abundance in drosophila by performing LC-MS/MS in this manuscript is more convincing as compared with previous studies. Intriguingly, the authors find that the conserved iron-binding motif required for the catalytic activity of TET is dispensable for its function. This finding could be important to reveal TET function in organisms whose genomic 5mC levels are very low.

The manuscript in this paper is well written but some aspects of data analysis and discussion need to be clarified and extended.

1. It is convincing that an increase in 6mA levels is not observed in TETnull presented in Fig1. But it seems 6mA levels are altered in Ax.TET1/2 compared with Ax.TETwt and Ax.TETnull presented in Fig1f (and also WT vs TET1/2 presented in Fig1g). Is it sure that no statistically significant were not observed between Ax.TET1/2 and Ax.TETwt?

2. The representing data of in vitro demethylation assay presented in Fig.3 is convincing, but it is not well discussed and analyzed why these results are contrary to previous reports (Yao et al., 2018 and Zhang et al., 2015).

We thank the reviewer for his/her work and positive assessment of our manuscript.

(1) We repeated our statistical analyses and confirmed that there is no significant difference between wildtype and tet1/2 mutant embryos in axenic conditions (Welch two sample t-test : p=0.075).

(2) We added some elements in the revised manuscript to discuss the possible reasons for the discrepancies with previous reports. Notably both studies performed the in vitro demethylation assays over a much longer time course and with different sources of recombinant proteins. Zhang et al. purified TET catalytic domain from human cells (HEK293T) and observed around 2.5% of 6mA demethylation at 30 min and less than 25% after 10 hours of incubation as measured by HPLC-MS/MS analyses. Yao et al. incubated recombinant TET catalytic domain with 6mA DNA for 3h and observed a 25% decrease in 6mA levels as measured by dot blot. These results suggest that drosophila TET may oxidize 6mA, but with a much lower affinity than 5mC since with observed a near complete oxidation of 5mC after 1 minute and no decrease in 6mA levels after 30 minutes of reaction (for identical concentrations of substrate and enzyme). It is possible too that the preparation of TET catalytic domain in different systems changes its enzymatic activity, potentially in relation with distinct post-translational modifications. Still, as already mentioned in our manuscript, extensive biochemical analyses of the distant TET homolog from the fungus Coprinopsis cinerea (Mu et al., Nature Chem Biol 2022) strongly argue that TET enzymes do not harbor the residues required to serve as 6mA demethylase.

Reviewer #1 (Recommendations For The Authors):

Here are one comment (#1) and a couple of questions (#2-3) that could be addressed in the future, in order to understand the roles of 6mA and TET. Even though #2 and #3 are likely beyond the scope of this paper, #1 should be addressed within the scope of this work and compared with previous reports.

1. The phenotypic analyses in Fig. 4 should use tet_null/Deficiency and tet_CD/Deficiency for their potential phenotypes. This needs to be addressed since both the tet_null and the tet_CD were generated using the same starting fly line (GFP knock-in). Using a deficiency chromosome and testing these alleles in hemizygotes would be helpful to eliminate any secondary effects due to genetic background issues.

Thanks for this comment. Actually, tet_null and tet_CD were not generated using the same starting lines. Whereas tet_cd was generated (by CRISPR) using the tet-GFP knock-in line, tet_null was generated by FRT site recombination between two PBac insertions (Delatte et al. 2016). As for tet1 and tet2 (used in allelic combination in Fig 4 J-L), they correspond to two distinct mutant alleles generated by CRISPR (Zhang et al. 2015). We have clarified this in the M&M (page 9).

1. Regarding the estimated "200 to 400 methylated adenines per haplogenome", is there any insight into where are they located in the genome?

It is an interesting question and we initially used SMRT-seq sequencing to obtain this kind of information. As it turned out that this technique gives a high level of false positive, we should consider with caution the interpretation of these data and we decided not to include them in the manuscript. Still, we characterized the genomic features of the 6mA detected using stringent criteria (mQV>100, cov>25x in the fusion dataset and triplicated across samples of the same genotype). Both in wild type and tet_null, 6mA were dispersed along each chromosome although few of them were found on chromosome X. In both cases there appeared to be a higher accumulation of 6mAs on the histone locus and the transposon-rich tip of chromosome X, but 6mA density remained below 1.3/kb in other genomic regions. Comparisons with annotated genomic regions indicated that 6mA were enriched in long interspersed nuclear elements (LINEs) and satellite repeats, and depleted in 3’UTR and exons, but there was no significant difference in their repartition between the two genetic contexts. Besides, motif analyses showed similar enrichments in both conditions, with GAG triplet accounting for more than one quarter of all the sites. Whether this reflects the specificity of a putative adenine methylase or a technical bias associated the with SMTR-seq technology remains to be established.

1. The TET-GFP and TET-CD-GFP knock-in lines give proper nuclear localization and could be used to identify genomic regions bound with full-length TET and TET-CD using anti-GFP for ChIP-seq or CUT&RUN (or CUT&TAG).

Indeed, this is a line of research that we are following up and will be part of another study. Actually, our ChIP-seq experiments indicate that they bind on the same genomic regions.

Reviewer #2 (Recommendations For The Authors):

• I think the major findings of this paper are showing 6mA present in drosophila by using xenic or conventional breeding conditions and finding that TET function independently of its catalytic activity is essential for fly development. The authors could have been more precise in title and abstract to emphasize these findings.

We have now modified the abstract to try to emphasize these findings.

• The authors claim that any increase of 6mA levels was not observed in both TETnull and TET1/2, but it is not sufficiently convincing. Because it seems 6mA levels were increased in Ax. tet1/2 embryo as compared with in Ax.wt embryo (Fig.1). In this scenario, 6mA abundance in both TETnull and TET1/2 mutant are supposed to be the same. It would be better to re-analyze data carefully and discuss if 6mA levels were significantly increased in TET1/2, and why 6mA levels are different between TETnull and TET1/2. Additionally, the authors describe that the TET null mutant is pupal lethal, while the TET1/2 survivor is available. The text suggests that TET1/2 could have partial functionality on fly development (Fig.4). It would be better to check whether the N-terminus of TET is expressed in the TET1/2 mutant.

Indeed, the increase in 6mA levels in Ax. tet1/2 embryo seems consequent (although it is not statistically significant) and no increase was observed in Ax tet_null embryos. Thus, the putative effect on 6mA levels in tet1/2 embryos may not be directly due to the absence of TET function. We now mention in the revised manuscript (page 6) that “the apparent increase in 6mA levels in tet1/2 axenic embryos was not reproduced in tet_null embryos, suggesting that it does not simply reflect the tet loss of function, and that it was not statistically significant”. Besides, we do not have an antibody to check whether the N-terminus of TET is expressed in the tet1/2 mutants, but the western blot published by Zhang et al 2015 shows that tet2 mutation leads to the expression of TET N-terminal domain. This N-terminal domain could have partial TET functionality and/or interfere with the function of other factors (notably those implicated in 6mA metabolism).

• The authors show that SMRT-seq data did not reveal an increase in 6mA levels in loss of TET (Fig.2). It is convincing that total 6mA abundance was not altered by loss of TET. But were 6mA-accumulated locus/regions observed in WT not altered by loss of TET?

Please refer to our answer to reviewer 1 on that point.

• It remains unclear that the TET proteins the authors prepared do not exhibit 6mA demethylate activity in vitro, contrary to what was reported in previous papers (Fig.3). I think the preparation of recombinant proteins may make different results between this and previous papers. Yao et al., 2018 and Zhang et al., 2015 used recombinant proteins purified from Human cells or insect cells, while the author purified them from E.Coli. Additionally, it's mentioned that VK Rao et al., 2020 demonstrated cdk5-mediated phosphorylation of Tet3 increases its in catalytic activity in vitro. These previous reports suggest modification of TET could change demethylase activity. More analysis and discussion are needed to support the conclusion.

Thanks for your insights. This in an important point and we added the following elements in the revised manuscript to discuss possible reasons for the discrepancies with previous reports (pages 7-8): “Our results contrast with previous reports showing that recombinant drosophila TET demethylates 6mA on dsDNA in vitro (Yao et al. 2018; Zhang et al., 2015a). However, both studies ran much longer reactions (up to 10 hours) and used different sources of recombinant protein (drosophila TET catalytic domain purified from human HEK293T cells). Notably, Zhang et al. (2015a) only found around 2.5% of 6mA demethylation at 30 min and less than 25% after 10 hours of incubation as measured by HPLC-MS/MS analyses. These results suggest that drosophila TET may oxidize 6mA, but with a much lower affinity than 5mC since with observed a near complete oxidation of 5mC after 1 min. and no significant decrease in 6mA levels after 30 min. of reaction (for identical concentrations of substrate and enzyme). It is possible too that the preparation of TET catalytic domain in different systems changes its enzymatic activity, potentially in relation to distinct post-translational modifications.”

eLife assessment

This study investigates the presence of DNA adenine methylation (6mA) and the associated function of TET enzyme, a DNA methylation mark eraser, in Drosophila. The study presents valuable findings on the scarcity of 6mA in the Drosophila genome and challenges previous findings regarding the role of TET in 6mA modification. The evidence supporting the claims is solid, and the paper has the potential to stimulate re-evaluations of the significance and regulatory mechanisms of 6mA DNA modifications in Drosophila.

Reviewer #1 (Public Review)

This work challenges previously published results regarding the presence and abundance of 6mA in Drosophila genome, as well as the claim that the TET or DMAD enzyme serves as the "eraser" of this DNA methylation mark and its roles in development. This information is needed to clarify these questions in the field. Generally speaking, the methods for fly husbandry and treatment seem to be in accordance with those established ones in the field.

Here are a couple of suggestions that could be discussed with the current work and addressed in the future, in order to better understand the roles of 6mA and TET.

1. Regarding the estimated "200 to 400 methylated adenines per haplogenome", some insights regarding where they are enriched in the genome could inform the potential target sites regulated by 6mA.

2. The TET-GFP and TET-CD-GFP knock-in lines give proper nuclear localization and could be used to identify genomic regions bound with full-length TET and TET-CD using anti-GFP for ChIP-seq or CUT&RUN (or CUT&TAG).

Reviewer #2 (Public Review)

DNA adenine methylation (6mA) is a rediscovered modification that has been described in a wide range of eukaryotes. However, 6mA presence in eukaryote remains controversial due to low abundance of its modification in eukaryotic genome. In this manuscript, Boulet et al. re-investigate 6mA presence in drosophila using axenic or conventional fly to avoid contaminant from feeding bacteria. By using these flies, they find that 6mA is rare but present in drosophila genome by performing LC/MS/MS. They also find that the loss of TET (also known as DMAD) does not impact on 6mA levels in drosophila, contrary to previous studies. In addition, the authors find that TET is required for fly development in its enzymatic activity-independent manner.

The strength of this study is, compared to previous studies of 6mA in drosophila, the authors employ axenic or conventional fly for 6mA analysis. These fly strains make it possible to analyze 6mA presence in drosophila without bacterial contaminant. This established method is valuable in this field.

Author Response

1. Reviewer 1 raised the concern that the images shown in the figures seem inconsistent with the quantitative data.

Our provisional response: The quantitative data are based on many samples and the photographs are just supposed to show illustrations of example data. Because of the volume containing P1a cells, is impossible to present a single confocal image that covers all P1a neurons and would therefore correspond more closely to the quantitative data. We chose to illustrate the quantitative data using single confocal images which contain both Hr38+/GFP+ and Hr38-/GFP+ neurons, to demonstate that we can distinguish clearly which P1a neurons are positive or negative for for Hr38 expression. This can be clarified in the figure legends. If it is imperative to show images(s) to reflect the statistics, we can do that but will need to present multiple confocal images for each condition, which could be messy and confusing.

1. Reviewer 2 states: "the major weakness is the calibration of the temporal resolution of HI-CatFISH in Figure 4 and Figure Supplement 4. According to Figure Supplement 4C, close to 100% of the Hr38-positive cells are already labeled with the exonic probe 30min post-stimulation, which is not reflected in Figure 4B (there, the expression level of the exonic probe peaks 60min post-induction)”.

The confusion may arise because we drew the illustration diagram (Fig. 4B) based on the quantitative data in Fig.S4B, which plots the intensity of Hr38 exonic ISH signals, while the reviewer may be comparing the illustration to the time course based on Fig.S4C, which shows the % positive cells, a binary measure. In the illustration (fig.4B), we wrote 'Hr38 expression level', not '%Hr38 positive cells.’ We can clarify this in the figure legend. If the reviewers prefer, we can add a threshold line in the diagram corresponding to the % positive cells at maximum.

eLife assessment

The work addresses an important methodological aspect by optimizing an activity-dependent labelling of neural circuits in behaving flies. The authors provide convincing evidence to support the broad applicability of this method. However, a more comprehensive description of the methodology would greatly enhance its dissemination and adoption. Additionally, the authors successfully implement the method, providing solid evidence for the activity-dependent labelling of P1 neurons during aggression and courtship.

Reviewer #1 (Public Review):

Summary:<br /> The authors have nicely demonstrated the efficiency of the HCR v.3.0 using hr38 mRNA expression as a marker of neuronal activity. This is very important in the Drosophila neuroscience field as in situ hybridization in adult Drosophila brains has been so far very challenging to do and replicate. However, this method has been described before [Choi et al., (2018)] and, if I understand correctly, is now the property of the non-profit organization molecular Technologies, who are the ones responsible for designing the probes (the sequences are not provided). Here the authors present their work as a description of a new method, called HI-FISH. However, I do not consider this as a new method but rather an application, a "proof of principle" that HCR v.3.0 can be done even on challenging tissues such as the adult Drosophila brain. Hence, if HCR v3.0 is sensitive enough and powerful enough to be used as a marker of neuronal activity, we can use it, for other neurobiological purposes, using other gene probes.<br /> To demonstrate the efficiency of HI-FISH, the authors have addressed two biological questions. The first one addressed whether specific groups of neurons, known to trigger specific behaviours (here courtship and/or aggression) are indeed activated by the behavioural context they can promote. In other words: is the behavioural output of these neurons also a trigger for their activation? The second question addressed whether this method is powerful enough to distinguish two subgroups of a class of neurons called P1 known to be involved in the two behaviours considered. In other words, are the same P1 neurons that promote aggression and courtship?

Strengths: The demonstration of the efficiency of the method is very convincing and well-performed. It gives the will for the reader to apply the method to their own subject.

Weakness: The pictures provided for HI-FISH and catFISH do not corroborate with the quantification and therefore I am not convinced about the author's biological interpretation of their data. See below for details.

Previously, using a split-gal4 line to restrict the Gal4 expression to a subset of P1 neurons, the authors have shown that these particular neurons when activated can trigger both aggressivity and courtship behaviour [Hoopfer et al., 2015]. The P1 neurons are composed of about 20 FruM neurons/hemibrain but are part of a bigger group that comprises about the same number of Fru- neurons that seem to be exclusively aggression-promoting neurons [Koganezawa et al., 2016]. Hence, this group of 40 neurons (pC1 neurons) contains aggressive-promoting neurons, courtship-promoting neurons, and perhaps neurons that can do both. Therefore, to address the first question, the authors compared hr38 expression in different groups of neurons, with a focus on subgroups, under different social contexts. While there is no ambiguity concerning the function of the Tk neurons as being exclusively aggressive-promoting neurons [Asahina et al., 2014], it is less clear when we look at the pC1 neurons. This is particularly evident for the P1a neurons which have been shown to be ambiguous as they can promote both aggression and courtship. For example, while optogenetic activation of these neurons promotes hr38 expression (Fig. 3D and fig sup. 4), it is less clear in the pictures to determine whether these specific P1a neurons labeled by the split-gal4 line are specifically activated by an aggressive behavioural context or a courtship behavioural context (Fig1, supp. 2 and fig. 4). Furthermore, the pictures chosen do not reflect the reality of the quantification (Fig. 2 B-D compared to sup. 2 or fig. 4C compared to fig. 4D) and therefore the authors conclusion. Because the drivers used are only expressed by a small subset of a larger population, I believe it would be more informative to separate in the quantification between the Gal4-expressing neurons and the non-expressing ones. Notably, based on the pictures provided, it looks like more P1 neurons (or rather pC1 neurons) are activated by a male-male encounter than by a male-female encounter. On the other hand, the splitGal4+ P1a seem to be more responsive to a courtship behavioural context (2/6 P1a neurons express hr38 in a courtship behavioural context while 0/9 _if we mentally abstract the increase of the background signal compared to the control picture_ seem to express hr38 in an aggression behavioural context). Hence, while activation of this P1asplit-Gal4 can promote both aggressive behaviour and courtship behaviour [Hoopfer et al., 2015], the authors didn't provide clear evidence (pictures not corroborating the quantification) that these specific small subpopulation of neurons are activated by one or the other or both behavioural conditions. Therefore my suggestion of differentiating in the quantification between the Gal4+ neurons from the others in the same local area.

Fig. 3, suppl. 3: In this section the authors addressed the question of whether the HI-FISH can be used to identify the downstream targets of this subpopulation. As positive controls of known downstream targets, the authors looked at the pCd population which they recently published as being an indirect downstream target of the P1a neurons [Jung et al., Neuron 2020]. They identified the Kenyon cells and a group of dopaminergic neurons, the PAM neurons as being activated by the P1a neurons. To confirm the increase of hr38 expression is indeed the result of a neuronal response of these neurons to the P1a activation, the authors used a different strategy used by them and others before. Using Gcamp signal to monitor the neuronal response of the presumably downstream targets the authors activated the P1a neurons using optogenetic (chrimson). It is important to note that they have previously shown that depending on the frequency of the light pulses activation of the P1a neurons can trigger only aggression, both aggression and wing extension or only wing extension [Hoopfer et al., eLife 2015]. Here the authors use 50Hz which is a frequency that leads to wing extension during the stimulation and aggressive behaviour at the offset of the stimulus [Hoopfer et al., eLife 2015]. Interestingly, the Gcamp experiment shows activation of the Kenyon cells and the PAM neurons but this activity is not maintained when the stimulus is turned off, suggesting that these neurons are activated during a courtship context rather than an aggressive behavioural context. I think it would be nice to see in which behavioural context the Kenyon cells and PAM neurons are activated (hr38 expression in the different behavioural context using the corresponding Gal4).<br /> Fig.4 and supp.4: The demonstration that the catFISH can now be done in Drosophila brain with a new in situ method was nicely performed. Notably, the intronic Hr38 probe appears to be an excellent marker for recent neuronal activation. However, while the optogenetic activation of the P1a neurons used to quantify the time lapse for both probes nicely distinguishes between nuclear and cytoplasmic exonic hr38, it is very difficult to use the localization of this probe in the experimental setup the authors used. Also, With their setup, I would simply use the frequency of intronic hr38 as a marker of recent activation correlating or not with the frequency of exonic hr38 marker (present in both conditions first and second encounter). This is important as this experiment addresses the second biological question. Once again, the pictures chosen absolutely do not corroborate the quantification. For example, the picture of the double encounter with the same gender male-male context clearly shows a higher number of cells that are hr38INT positive (and therefore nuclear) than the picture of the female-female context (Fig. 4C), and thus even if we only considered the P1asplit-Gal4 positive cells. In the male-male picture, 5/6 P1a cells have the Hr38INT marker while the presence of this marker is debatable in the female-female context. Especially, in some of the cells these magenta dots appear to be localized in the cytoplasm, suggesting a non-specific signal. Therefore, I would suggest to quantify the percentage of Hr38INT positive cells as the only marker for recent activation and the relative level of Hr38EXN immunostaining, and this among the P1asplit-Gal4 positive cells and the gal4- ones. A high Hr38EXN level associated with the presence of hr38INT would indicate that the cell has been activated during both encounters, while a lower hr38EXN with no hr38INT would suggest only an activation during the 1st behavioural context. Finally, a lower hr38EXN associated with the presence of hr38INT would suggest the opposite, an activation only during the 2nd behaviour.<br /> Overall, by only looking at the pictures provided, I would conclude that the HCR applied with the hr38 probes seems to efficiently work and is usable to address the question of whether a specific group of neurons are indeed activated by a specific social behavioural context. However, I would also conclude that this technique nicely demonstrated that flies are not robots and that even in a "simple" organism model such as Drosophila melanogaster individual variability is present among a group of neurons. Hence, only the quantification of the gal4-expressing neurons in comparison with their neighbor neurons known to belong to the same functional group, would allow a conclusion toward a specificity of contextual response. Therefore, although activation of a small group of neurons can be enough to trigger a specific behaviour that shouldn't happen under a certain environmental context [Hoopfer et al., eLife 2015], the results presented here suggest that we should, using this method, considering the response of the neighbour cells of the Gal4+ ones. Although currently, the quantification of the author's data does not allow such analysis, to strengthen the author's argumentation, I would distinguish in their quantification between gal4+ from the others (Fig. 2 and 4). Furthermore, I am not certain that the distinction between cytoplasmic and nuclear hr38EXN is 100% feasible (based on the pictures provided). I would instead for the hr38EXN marker only use the relative intensity (Fig. 4D).

Reviewer #2 (Public Review):

Summary:<br /> Watanabe et al establish a novel method for the activity-dependent labeling of neural circuits in flies. While activity mapping of neurons that are active during specific behaviors is widespread in rodents, the application of this method to fly circuit neuroscience is limited, mainly due to technological challenges. Thus, the present study addresses a timely problem. To do so, they apply the in situ hybridization amplification method called Hybridization Chain Reaction v. 3.0 (Choi et al. 2018) to the adult fly brain in order to visualize the expression changes of the immediate early gene (IEG) Hr38 under different types of social contexts. The conclusions of this paper are mostly very well supported by data but it would strongly benefit from additional methodological details as well as additional controls, in particular for the HI-catFISH experiments.

Strengths:<br /> The major strength of this method is its versatility and sensitivity. It can be applied to a wide variety of biological questions and assess the dynamic transcriptional regulation of an unlimited number of genes with a high signal-to-noise ratio. It will be therefore useful to many research labs working on different biological questions.

Weaknesses:<br /> Although the paper has great strengths in principle, the major weakness is the calibration of the temporal resolution of HI-CatFISH in Figure 4 and Figure Supplement 4. According to Figure Supplement 4C, close to 100% of the Hr38-positive cells are already labeled with the exonic probe 30min post-stimulation, which is not reflected in Figure 4B (there, the expression level of the exonic probe peaks 60min post-induction) and may have profound implications for the interpretation of the results. The present manuscript would strongly benefit from additional controls, such as the quantification of the intronic and exonic Hr38 probes after either only the 1st or 2nd social context but at the same timepoint than if two consecutive social contexts were tested.

Reviewer #3 (Public Review):

In this study, Yang et al. address a fundamental question of the role of dorsal striatum in neural coding of gait. The authors study the respective roles of D1 and D2 MSNs by linking their balanced activity to detailed gait parameters. In addition, they put in parallel the striatal activity related to whole-body measures such as initiation/cessation of movement or body speed. They are using an elegant combination of high-resolution single-limb motion tracking, identification of bouts of movements, and electrophysiological recordings of striatal neurons to correlate those different parameters. Subpopulations of striatal output neurons (D1 and D2 expressing neurons) are identified in neural recordings with optogenetic tagging. Those complementary approaches show that a subset of striatal neurons have phase-locked activity to individual limbs. In addition, more than a third of MSNs appear to encode all three aspects of motor behavior addressed here, initiation/cessation of movement, body speed, and gait. This activity is balanced between D1 and D2 neurons, with a higher activity of D1 neurons only for movement initiation. Finally, alterations of gait, and the associated striatal activity, are studied in a mouse model of Parkinson's Disease, using 6-OHDA lesions in the medial forebrain bundle (MFB). In the 6OHDA mice, there is an imbalance toward D2 activity.

Strengths:<br /> There is a long-standing debate on the respective role of D1 and D2 MSNs on the control of movement. This study goes beyond prior work by providing detailed quantification of individual limb kinematics, in parallel with whole-body motion, and showing a high proportion of MSNs to be phase-locked to precise gait cycle and also encoding whole-body motion. The temporal resolution used here highlights the preferential activity of D1 MSN at the movement starts, whereas previous studies described a more balanced involvement. Finally, they reveal neural mechanisms of dopamine depletion-induced gait alterations, with a preponderant phase-locked activity of D2 neurons. The results are convincing, and the methodology supports the conclusions presented here.

Weaknesses:<br /> Some more detailed explanations would improve the clarity of the results in the corresponding section. Analysis of the 6OHDA experiments could be expanded to extract more relevant information.

eLife assessment

In this manuscript, the authors recorded the activity of D1- and D2-MSNs in the dorsal striatum and analyzed their firing activity in relation to single-limb gait in normal and 6-OHDA lesioned mice. This important work extends previous studies showing that the striatum multiplexes various aspects of locomotion, including velocity and movement transitions, by demonstrating that striatal neurons also encode single-limb gait. The authors present solid evidence to show that gait deficits induced by severe unilateral dopamine depletion are associated with an imbalance in the gait-modulation of striatal pathways, however, the reviewers also point out that the evidence supporting the conclusion that striatal neurons encode single-limb gait is incomplete.

Reviewer #1 (Public Review):

Summary:<br /> Yang et al combine high-speed video tracking of the limbs of freely moving mice with in vivo electrophysiology to demonstrate how striatal neurons encode single-limb gait. They also examine encoding other well-known aspects of locomotion, such as movement velocity and the initiation/termination of movement. The authors show that striatal neurons exhibit rhythmic firing phase-locked with mouse gait, while mice engage in spontaneous locomotion in an open field arena. Moreover, they describe gait deficits induced by severe unilateral dopamine neuron degeneration and associate these deficits with a relative strengthening of gait-modulation in the firing of D2-expressing MSNs. Although the source and function of this gait-modulation remain unclear, this manuscript uncovers an important physiological correlate of striatal activity with gait, which may have implications for gait deficits in Parkinson's Disease.

Strengths:<br /> While some previous work has looked at the encoding of gait variables in the striatum and other basal ganglia nuclei, this paper uses more careful quantification of gait with video tracking. In addition, few if any papers do this in combination with optically-labeled recordings as were performed here.

Weaknesses:<br /> The data collected has a great richness at the physiological and behavioral levels, and this is not fully described or explored in the manuscript. Additional analysis and display of data would greatly expand the interest and interpretability of the findings.

There are also some caveats to the interpretation of the analyses presented here, including how to compare encoding of gait variables when animals have markedly different behaviors (eg comparing sham and unilaterally 6-OHDA treated mice), or how to interpret the loss of gait modulation when single unit activity is overall very low.

1. The authors use circular analysis to quantify the degree to which striatal neurons are phase-locked to individual limbs during gait. The result of this analysis is shown as the proportion of units phase-locked to each limb, vector length, and vector angle (Fig 2H-K; Fig 4E-F; Fig 6E-F). Given that gait is a cyclic oscillation of the trajectories of all four limbs, one could expect that if one unit is phase-locked to one limb, it will also be phase-locked to the other three limbs but at a different phase. Therefore, it is not clear in the manuscript how the authors determine to which limb each unit is locked, and how some units are locked to more than one limb (Fig 2H). More methodological/analytical detail would be especially helpful.

2. In Figures 2 and 3, the authors describe the modulation of striatal neurons by gait, velocity, and movement transitions (start/end), with most of their examples showing firing rates compatible with rates typical of striatal interneurons, not MSNs. In order to have a complete picture of the relationship between striatal activity and gait, a cell type-specific analysis should be performed. This could be achieved by classifying units into putative MSN, FS interneurons, and TANs using a spike waveform-based unit classification, as has been done in other papers using striatal single-unit electrophysiology. An example of each cell type's modulation with gait, as well as summary data on the % modulation, would be especially helpful.

3. By normalizing limb trajectories to the nose-tail axis, the analysis ignores whether the mouse is walking straight, or making left/right turns. Is the gait-modulation of striatal activity shaped by ipsi- and contralateral turning? This would be especially important to understand changes in the unilateral disease model, given the imbalance in turning of 6-OHDA mice.

4. It looks like the data presented in Figure 4 D-F comes from all opto-identified D1- and D2-MSNs. How many of these are gait-modulated? This information is missing (line 110). Pooling all units may dilute differences specific to gait-modulated units, therefore a similar analysis only on gait-modulated units should be performed.

5. Since 6-OHDA lesions are on the right hemisphere, we would expect left limbs to be more affected than right limbs (although right limbs may also compensate). It is therefore surprising that RF and RR strides seem slightly shorter than LF and LR (Fig 5G), and no differences in other stride parameters (Fig 5H-J). Could the authors comment on that? It may be that this is due to rotational behavior. One interesting analysis would be to compare activity during similar movements in healthy and 6-OHDA mice, eg epochs in which mice are turning right (which should be present in both groups) or walking a few steps straight ahead (which are probably also present in both groups).

6. Multiple publications have shown that firing rates of D1-MSN and D2-MSN are dramatically changed after dopamine neuron loss. Is it possible that changes observed in gait-modulation might be biased by changes in firing rates? For example, dMSNs have exceptionally low overall activity levels after dopamine depletion (eg Parker...Schnitzer, 2018; Ryan...Nelson, 2018; Maltese...Tritsch, 2021); this might reduce the ability to detect modulation in the firing of dMSNs as compared to iMSNs, which have similar or increased levels of activity in dopamine depleted mice. Does vector length correlate with firing rate? In addition, the normalization method used (dividing firing rate by minimum) may amplify very small changes in absolute rates, given that the firing rates for MSN are very low. The authors could show absolute values or Z-score firing rates (Figure 6 A, D).

7. The analysis shown in Fig 3C should also be done for opto-identified D1- and D2-MSNs (and for waveform-based classified units as noted above).

8. Discussion: the origin of the gait-modulation as well as the possible mechanisms driving the alterations observed in 6-OHDA mice should be discussed in more detail.

Reviewer #2 (Public Review):

Summary:<br /> Yang et al. recorded the activity of D1- and D2-MSNs in the dorsal striatum and analyzed their firing activity in relation to single-limb gait in normal and 6-OHDA lesioned mice. Although some of the observations of striatal encoding are interesting, the novelty and implications of this firing activity in relation to gait behavior remain unclear. More specifically, the authors made two major claims. First, the striatal D1- and D2-MSNs were phase-locked to the walking gait cycles of individual limbs. Second, dopamine lesions led to enhanced phase-locking between D2-MSN activity and walking gait cycles. The second claim was supported by the increase of vector length in D2-MSNs after unilateral 6-OHDA administration to the medial forebrain bundle. However, for the first claim, the authors failed to convincingly demonstrate that striatal MSNs were more phase-locked to gait with single-limb and step resolution than to the global gait cycles.

Strengths:<br /> It is a technically advanced study.

Weaknesses:<br /> 1. The authors focused on striatal encoding of gait information in current studies. However, it remains unclear whether the part of the striatum for which the authors performed neuronal recording is really responsible for or contributing to gait control. A lesion or manipulation experiment disrupting the part of the striatum recorded seems a necessary step to test or establish its relationship to gait control.

2. The authors attributed one of the major novelties to phase-locking of striatal neural activities with single-limb gait cycles. The claim was not clearly supported, as the authors did not demonstrate that phase-locking to single-limb gaits was more significant than phase-locking to global walking gait cycles. In rhythmic walking, the LR and RF limbs were roughly anti-phase with the LF and RR limbs (Fig. 1D, E). In line with this relationship, striatal neurons were mainly in-phase with LR and RF limbs and anti-phase with LF and RR limbs (Fig. 2J, K). One could instead interpret this as the striatal neurons spanned all the phases of the global walking gait cycles (Fig. 3D). To demonstrate phase-locking with individual limb movements, the authors need to show that neural activities were better correlated with a specific limb than to the global gait cycles.

3. The observation of the enhancement of coupling between D2 MSN firing and the gait cycles was interesting, but the physiological interpretation was not clear (as the authors also noted in the Discussion), which hampers the significance of the observation.

4. Due to the lack of causality experiments as mentioned in the first comment above, the observations of coupling between striatal neuronal activity and gait control might well result from a third brain region/factor serving as the common source to both, whether in normal or dopamine lesioned brain. If this is the case, the significance and implications of current findings will be greatly limited.

eLife assessment

This important paper provides web based interface for cross-tissue analysis of omics datasets from – so far – two different human populations, with compelling evidence that the tool can be used to make meaningful scientific discoveries. Conceptually, these analyses are relevant for any systems biologist or bioinformatician who is interested in integrating large population datasets. Currently, the resource is already of use for scientists studying the HMDP or using GTEx data, and we hope to see updates in the coming years that incorporate more populations and more datatypes, which could make it a general tool for a wide community.

Author Response

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

We thank the reviewers and editors for their time and careful consideration of this study. Nearly every comment proved to be highly constructive and thoughtful, and as a result, the manuscript has undergone major revisions including the title, all figures, associated conclusions and web app. We feel that the revised resource provides a more systematic and comprehensive approach to correlating inter-individual transcript patterns across tissues for analysis of organ cross-talk. Moreover, the manuscript has been restructured to highlight utility of the web tool for queries of genes and pathways, as opposed to focused discrete examples of cherry-picked mechanisms. A few key revisions include:

• Manuscript: All figures have been revised to place to explore broad pathway representation. These analyses have replaced the previous circadian and muscle-hippocampal figures to emphasize ability to recapitulate known physiology and remove the discovery portion which has not been validate experimentally.

• Manuscript: The term “genetic correlation” or “genetically-derived” has been replaced throughout with “transcriptional”, “inter-individual”, or mostly just “correlations”.

• Manuscript: A new figure (revised fig 2) has been added to evaluate the innate correlation structure of data used for common metabolic pathways, in addition an exploration of which tissues generally show more co-correlation and centrality among correlations.

• Manuscript: A new figure (revised fig 4) has been added to highlight the utility of exploring gene ~ trait correlations in mouse populations, where controlled diets can be compared directly. These highlight sex hormone receptor correlations with the large amount of available clinical traits, which differ entirely depending on the tissue of expression and/or diet in mouse populations.

• Web tool: Addition of a mouse section to query expression correlations among diverse inbred strains and associated traits from chow or HFHS diet within the hybrid mouse diversity panel.

• Web tool: Overrepresentation analysis for pathway enrichments have been replaced with score-based gene set enrichment analyses and including network topology views for GSEA outputs.

• Web tool: Associated github repository containing scripts for apps now include a detailed walk-through of the interface and definitions for each query and term.

Public Reviews:

Reviewer #1 (Public Review):

Zhou et al. have set up a study to examine how metabolism is regulated across the organism by taking a combined approach looking at gene expression in multiple tissues, as well as analysis of the blood. Specifically, they have created a tool for easily analyzing data from GTEx across 18 tissues in 310 people. In principle, this approach should be expandable to any dataset where multiple tissues of data were collected from the same individuals. While not necessary, it would also raise my interest to see the "Mouse(coming soon)" selection functional, given that the authors have good access to multi-tissue transcriptomics done in similarly large mouse cohorts.

Summary

The authors have assembled a web tool that helps analyze multiple tissues' datasets together, with the aim of identifying how metabolic pathways and gene regulation are connected across tissues. This makes sense conceptually and the web tool is easy to use and runs reasonably quickly, considering the size of the data. I like the tool and I think the approach is necessary and surprisingly under-served; there is a lot of focus on multi-omics recently, but much less on doing a good job of integrating multi-tissue datasets even within a single omics layer.

What I am less convinced about is the "Research Article" aspect of this paper. Studying circadian rhythm in GTEx data seems risky to me, given the huge range in circadian clock in the sample collection. I also wonder (although this is not even remotely in my expertise) whether the circadian rhythm also gets rather desynchronized in people dying of natural causes - although I suppose this could be said for any gene expression pathway. Similarly for looking at secreted proteins in Figure 4 looking at muscle-hippocampus transcript levels for ADAMTS17 doesn't make sense to me - of all tissue pairs to make a vignette about to demonstrate the method, this is not an intuitive choice to me. The "within muscle" results look fine but panels C-E-G look like noise to me...especially panel C and G are almost certainly noise, since those are pathways with gene counts of 2 and 1 respectively.

I think this is an important effort and a good basis but a significant revision is necessary. This can devote more time and space to explaining the methodology and for ensuring that the results shown are actually significant. This could be done by checking a mix of negative controls (e.g. by shuffling gene labels and data) and a more comprehensive look at "positive" genes, so that it can be clearly shown that the genes shown in Fig 1 and 2 are not cherry-picked. For Figure 3, I suspect you would get almost an identical figure if instead of showing pan-tissue circadian clock correlations, you instead selected the electron transport chain, or the ribosome, or any other pathway that has genes that are expressed across all tissues. You show that colon and heart have relatively high connectivity to other tissues, but this may be common to other pathways as well.

Response: We are thankful to the reviewer in their detailed assessment of the manuscript. The comments raised in both the public and suggested reviews clearly improved the revised study and helped to identify limitations. In general, we have removed data suggesting “discovery” using these generalized analyses, such as removing figures evaluating circadian rhythm genes and muscle-hippocampus correlations. These have been replaced with more thorough investigations of tissue correlation structure and potentially identified regions of data sparsity which are important for users to consider. Also, we have added a similar full detailed pipeline of mouse (HMDP) data and highlighted in the manuscript by showing transcript ~ trait correlations of sex hormone receptor genes which differ between organs and diets. Further responses to individual points are also provided below.

Reviewer #2 (Public Review):

Summary:

Zhou et al. use publicly available GTEx data of 18 metabolic tissues from 310 individuals to explore gene expression correlation patterns within-tissue and across-tissues. They detect signatures of known metabolic signaling biology, such as ADIPOQ's role in fatty acid metabolism in adipose tissue. They also emphasize that their approach can help generate new hypotheses, such as the colon playing an important role in circadian clock maintenance. To aid researchers in querying their own genes of interest in metabolic tissues, they have developed an easy-to-use webtool (GD-CAT).

This study makes reasonable conclusions from its data, and the webtool would be useful to researchers focused on metabolic signaling. However, some misconceptions need to be corrected, as well as greater clarification of the methodology used.

Strengths:

GTEx is a very powerful resource for many areas of biomedicine, and this study represents a valid use of gene co-expression network methodology. The authors do a good job of providing examples confirming known signaling biology as well as the potential to discover promising signatures of novel biology for follow-up and future studies. The webtool, GD-CAT, is easy to use and allows researchers with genes and tissues of interest to perform the same analyses in the same GTEx data.

Weaknesses:

A key weakness of the paper is that this study does not involve genetic correlations, which is used in the title and throughout the manuscript, but rather gene co-expression networks. The authors do mention the classic limitation that correlation does not imply causation, but this caveat is even more important given that these are not genetic correlations. Given that the goal of their study aligns closely with multi-tissue WGCNA, which is not a new idea (e.g., Talukdar et al. 2016; https://doi.org/10.1016/j.cels.2016.02.002), it is surprising that the authors only use WGCNA for its robust correlation estimation (bicor), but not its latent factor/module estimation, which could potentially capture cross-tissue signaling patterns. It is possible that the biological signals of interest would be drowned out by all the other variation in the data but given that this is a conventional step in WGCNA, it is a weakness that the authors do not use it or discuss it.

Response: Thank you for the helpful and detailed suggestions regarding the study. The review raised some important points regarding methodological interpretations (ex. bicor-exclusive application as opposed to module-based approaches), as well as clarification of “genetic” inferences throughout the study. The comparison to module-based approaches has also now been discussed directly, pointing our considerations and advantages to each. We hope that the reviewer with our corrections to the misconceptions posed, many of which we feel were due to our insufficient description of methodological details and underlying interpretations. The revised manuscript, web portal and associated github provide much more detail and many more responses to specific points are provided below.

Reviewer #3 (Public Review):

Summary: A useful and potentially powerful analysis of gene expression correlations across major organ and tissue systems that exploits a subset of 310 humans from the GTEx collection (subjects for whom there are uniformly processed postmortem RNA-seq data for 18 tissues or organs). The analysis is complemented by a Shiny R application web service.

The need for more multisystems analysis of transcript correlation is very well motivated by the authors. Their work should be contrasted with more simple comparisons of correlation structure within different organs and tissues, rather than actual correlations across organs and tissues.

Strengths and Weaknesses: The strengths and limitations of this work trace back to the nature of the GTEx data set itself. The authors refer to the correlations of transcripts as "gene" and "genetic" correlations throughout. In fact, they name their web service "Genetically-Derived Correlations Across Tissues". But all GTEx subjects had strong exposure to unique environments and all correlations will be driven by developmental and environmental factors, age, sex differences, and shared and unshared pre- and postmortem technical artifacts. In fact we know that the heritability of transcript levels is generally low, often well under 25%, even studies of animals with tight environmental control.

This criticism does not comment materially detract for the importance and utility of the correlations-whether genetic, GXE, or purely environmental-but it does mean that the authors should ideally restructure and reword text so as to NOT claim so much for "genetics". It may be possible to incorporate estimates of chip heritability of transcripts into this work if the genetic component of correlations is regarded as critical (all GTEx cases have genotypes).

Appraisal of Work on the Field: There are two parts to this paper: 1. "case studies" of cross-tissue/organ correlations and 2. the creation of an R/Shiny application to make this type of analysis much more practical for any biologist. Both parts of the work are of high potential value, but neither is fully developed. My own opinion is that the R/Shiny component is the more important immediate contribution and that the "case studies" could be placed in the context of a more complete primer. Or Alternatively, the case studies could be their own independent contributions with more validation.

Response: We thank the reviewer for their supportive and helpful comments. The discussion of usage of the term “genetic” has been removed entirely from the manuscript as this point was made by all reviewers. Further, we have revised the previous study to focus on more detailed investigations of why transcript isoforms seemed correlated between tissues and areas where datasets are insufficient to provide sufficient information (ex. Kidney in GTEx). As the reviewer points out, the previous “case studies” were unvalidated and incomplete and as a result, have been replaced. Additional points below have been revised to present a more comprehensive analyses of transcript correlations across tissues and improved web tool.

(Recommendations For The Authors):

As this manuscript is focused on the analytical process rather than the biological findings, the reviewer concerns are not a fundamental issue to subsequent acceptance of the paper, but some of the examples will need to be replaced or double-checked to ensure their biological and statistical relevance. To raise the scope and interest of the method developed, it would be seen very positively to include additional datasets, as the authors seem to have intended to have done, with a non-functional (and highlighted as such) selection for mouse data. Establishing that the authors can easily - and will easily - add additional datasets into their tool would greatly raise the reviewers' confidence in the methodology/resource aspect of this paper. This may also help address the significant concerns that all three reviewers raised with the biological examples, e.g. that GTEx data is so uncontrolled that studying environmentally-influenced traits such as circadian rhythm may be challenging or even impossible to do properly. Adding in a more highly controlled set of cross-tissue mouse data may be able to address both these concerns at once, i.e. the resource concern (can the website easily be updated with new data) and the biological concern (are the results from these vignettes actually statistically significant).

Reviewer #1 (Recommendations For The Authors):

Comments, in approximately reverse order of importance

1. Some figure panels are not referenced in the text, e.g. Fig 1B and Figure 2E. Response: Thank you for pointing this out. We have revised every figure in the manuscript and additionally gone through to make sure every panel is referenced in the text.

2. The authors mention "genetic data" several times but I don't see anything about DNA. By "genetic data" do you mean "transcriptome expression data," or something else?

Response: This is an important point, also raised by all 3 reviewers. We have clarified in the abstract, results and discussion that correlations are between transcripts. As a result, all mentions of “genetics” or “genetic data” has been removed, with the exception of introducing mouse genetic reference panels.

1. For Figure 3, the authors look at circadian clock data, but the GTEx data is from all sorts of different times of day from across the patient cohort depending on when the donor died, and I don't see this metadata actually mentioned anywhere. I see Arntl Clock and all the other circadian genes are highly coexpressed in each tissue (except not so strong in liver) but correlation across tissue seems more random. Also hypothalamus seems to be very strongly negatively correlated with spleen, but this large green block doesn't have significance? That is surprising to me, since the sample sizes are all equivalent I would expect any correlation remotely close to -1.0 to be highly significant.

Response: The reviewer raises several important points with regard to the source of data and underlying interpretations. We have added a revised Fig 2, suggesting that representation of gene expression between tissues can be strongly biased by nature of samples (ex. differences in data that is available for each tissue) and also discussed considerations of the nature of sample origin in the limitations section. We have also used some of these points when introducing rationale for using mouse population data. As a result of comments from this reviewer and others, we have removed the circadian rhythm analysis and muscle-hippocampal figures from the revised study; however, specifically mentioned these cohort differences in the discussion section (lines 294-298). Circadian rhythm terms are also evaluated in Fig 2 and consistent with the reviewers concerns, less overall correlations are observed between transcripts across tissues when compared to other common GO terms assessed.

1. Figure 4, this is all transcript-level data, so it is confusing to see protein nomenclature used, e.g. "expression of muscle ADAMTS17" should be "expression of muscle ADAMTS17" (ADAMTS17 the transcript should be in italics, in case the formatting is removed by the eLife portal). Same for FNDC5. In the figures you do have those in italics, so it is just an issue in the manuscript text. In general please look through the text and make sure whether you are referring really to a "gene," "transcript," or "protein." For instance, Figure 1 legend I think should be "A, All transcripts across the ... with local subcutaneous and muscle transcript expression." I know people still sometimes use "gene expression" to refer to transcripts, but now that proteomics is pretty mainstream, I would push for more careful vocabulary here.

Response: Thank you for pointing these out. While we have replaced Fig 4 entirely as to limit the unvalidated discovery or research aspects of the paper, we have gone through the text and figures to check that the correct formatting is used for references to human genes (capitalized italics) or the newly-included mouse genes (lower-case italics).

1. "Briefly, these data were filtered to retain genes which were detected across individuals where individuals were required to show counts > 0 in 1.2e6 gene-tissue combinations across all data." I don't quite understand the filtering metric here - what is 1.2 million gene-tissue combinations referring to? 20k genes times 18 tissues times 310 people is ~100 million measurements, but for a given gene across 310 people * 18 tissues that is only ~6000 quantifications per gene.

Response: We apologize for this oversight, as the numbers were derived from the whole GTEx dataset in total and not the tissues used for the current study. We have clarified this point in the revised manuscript (methods section in Datasets used) and also removed confusing references to specific numbers of transcripts and tissues unless made clear.

1. Generally I think your approach makes sense conceptually but... for the specific example used in e.g. figure 4, this only makes sense to me if applied to proteins and not to transcripts. Looking at the transcript levels per tissue for genes which are secreted could be interesting but this specific example is confusing, as is the tissue selected. I would not really expect much crosstalk between the hippocampus and the muscle, especially not in terms of secreted proteins.

Response: This is a valid point, also raised by other reviewers. While we wanted to highlight the one potentially-new (ADAMTS7) and two established proteins (FNDC5 and ERFE) and their correlations, the fact that this direct circuit remains to be validated led us to replace the figure entirely. The point raised about inference of protein secretion compared to action; however, has been expanded upon in the results and discussion. We now show that complexities arise when using this approach to infer mechanisms of proteins which are primarily regulated post-transcriptionally. We provide a revised Supplemental Fig 4 showing that this general framework, when applied to expression of INS (insulin), almost exclusively captured pathways leading to its secretion and not action.

1. It's not clear to me how correction for multiple testing is working in the analyses used in this manuscript. You mention q-values so I am sure it was done, I just don't see the precise method mentioned in the Methods section.

Response: We apologize for this oversight and have included a specific mention of qvalue adjustment using BH methods, where our reasoning was the efficiency in run-time (compared to other qvalue methods). In addition, we provide a revised Fig 2 which suggests that innate correlation structure exists between tissues for a variety of pathways which should be considered. We also compare several empirical bicor pvalues and qvalue adjustments directly between these large pathways where much of the innate tissue correlation structure does appear present when BH qvalue adjustments are applied (revised Fig 2A).

1. The piecharts in Figure 1 are interesting - I would actually be curious which tissues generally have closer coexpression. This would be an absolutely massive number of pairwise correlations to test, but maybe there is a smarter way to do it? For instance, for ADIPOQ, skeletal muscle has the best typical correlation, but would that be generally true just that many adipose genes have closer relationship between the two tissues?

Response: This comment inspired us to perform a more systematic query of global gene-gene correlation structures, which is now shown as the revised Fig 2A. With respect to ADIPOQ, the reviewer is correct in that there does appear to be a general pattern of muscle genes showing stronger correlation with adipose genes. We emphasize and discuss there in the revised manuscript to point out that global trends of tissue correlation structure should be taken into account when looking at specific genes. Much of this innate co-correlation structure could be normalized by the BH qvalue adjustment (above); however, strongly correlated pathways like mitochondria showed selective patterns throughout thresholds (revised Fig 2A). Further, we analyze KEGG terms and general correlation structures (revised Fig 2B) to point out the converse, that some tissues are just poorly represented. Interpretation of correlated genes from these organ and pathway combinations should be especially considered in the framework that their poor representation in the dataset clearly impacted the global correlation structures. We have added these points to both results and discussion. In sum, we feel that this was a critical point to explore and attempted to provide a framework to identify/consider in the revised manuscript.

1. The pathway enrichments in Figure 1 are more difficult for me to interpret, e.g. for ADIPOQ, the scWAT pathways make sense, but the enriched skeletal muscle pathways are less clearly relevant (rRNA processing?? Not impossible but no clear relevance either). What are the significances for these pathway enrichments? Is it even possible to select a gene that has no peripheral pathway enrichment, e.g. if you take some random Gm#### or olfactory receptor gene and run the analysis, are you also going to see significant pathways selected, as pathway enrichment often has a trend to overfit? The "within organ" does seem to make sense, but I am also just looking at 4 anecdotes here and it is unclear whether they are cherry picked because they did make sense. That is, it's unclear why you selected ADIPOQ and not APOE or HMGCR or etc. I also don't figure out how I can make these pathway enrichment plots using your website. I do get the pie chart but when I try the enrichment analysis block (NB: typo on your website, it says "Enrich-E-ment Analysis" with an extra E) I always get that "the selected tissue do not contain enough genes to generate positive the enrichment." (Also two typos in that phrase; authors should check and review extensively for improvements to the use of English.) After trying several genes I eventually got it to work. I think there is some significant overfitting here, as I am pretty sure that XIST expression in the white adipose tissue has nothing to do with olfactory signalling pathways, which are the top positive network (but with an n = 4 genes).

Response: Several good points within this comment. 1) the pathway enrichments have been revised completely. The reviewer provided a helpful suggestion of a rank-based approach to query pathways, as opposed to the previous over-representation tests. After evaluating several different pathway enrichment tools based on correlated tissue expression transcripts, a rank- and weight-based test (GSEA) captured the most physiologic pathways observed from known actions of select secreted proteins. Therefore, revised pathway enrichments and web-tool queries unitize a GSEA approach which accounts for the rank and weight determined by correlation coefficient. In implementing these new pathway approaches, we feel that pathway terms perform significantly better at capturing mechanisms. 2) With respect to the selection genes, we wanted to provide a framework for investigating genes which encode secreted proteins that signal as a result of the abundance of the protein alone. This is a group-bias; however, and not necessarily reflective of trying to tackle the most important physiologic mechanisms underlying human disease. We agree with the reviewer in those evaluating genes such as APOE and cholesterol synthesis enzymes present an exciting opportunity, our expertise in interpretation and mechanistic confirmation is limited. 3) We have gone through the revised manuscript and attempted to correct all grammatical and/or spelling mistakes.

1. The network figures I get on your website look actually more interesting than the ones you have in Figure 2, which only stay within a tissue. Making networks within a tissue is pretty easy I think for any biologist today, but the cross-tissue analysis is still fairly hard due to the size of the datasets and correlation matrices.

Response: We greatly appreciate the reviewer’s enthusiasm for the network model generation aspect. We have tried to improve the figure generation and expanded the gene size selection for network generation in the web tool, both within and across tissues. We are working toward allowing users to select specific pathway terms and/or tissue genes to include in these networks as well, but will need more time to implement.

1. I get a bug with making networks for certain genes, e.g. XIST - Liver does not work for plotting network graphs. Maybe XIST is a suppressed gene because it has zero expression in males? It is an interesting gene to look at as a "positive control" for many analyses, since it shows that sample sexing is done correctly for all samples.

Response: The reviewer recognized a key consideration in underlying data structure for GTEx. In the revised manuscript, we evaluated tissue representation (or lack thereof) being a crucial factor in driving where significant relationships cannot be observed in tissues such as kidney, liver and spleen (Fig 2). Moreover, the representation of females (self-reported) in GTEx is less-than half of males (100 compared to 210 individuals). We have emphasized this point in the discussion where we specifically pointed out the lack of XIST Liver correlation being a product of data structure/availability and not reflecting real biologic mechanisms. We expanded on this point by highlighting the clear sex-bias in terms of representation.

1. On the network diagram on your website, there doesn't seem to be any way to zoom in on the website itself? You can make a PDF which is nice but the text is often very small and hard to read.

Response: We have revised the web interface plot parameters to create a more uniform graph.

1. On a related note, is it possible to output the raw data and gene lists for the network graph? I would want to know what are those genes and their correlation coefficient.

Response: We have enabled explore as .pdf or .svg graphics for the network and all plots. In addition, following pie chart generation at the top of the web app, users now have the ability to download a .csv file containing the bicor coefficients, regression pvalues and adjusted qvalues for all other gene-tissue combinations.

1. Some functionality issues, e.g. on the "Scatter plot" block, I input a gene name again here. Shouldn't this use the same gene selected already at the top of the page? It seems confusing to again select the gene and tissue here, but maybe there is a reason for that.

Response: It would be more intuitive to only display genes from a given selected tissue for scatterplots; however, we chose to keep all possible combinations with the [perhaps unnecessary] option of reselecting a tissue to allow users to query any specific gene without having to wait to run the pathways for all that correspond to a given tissues.

1. Figure 4H should also probably be Figure 1A.

Response: Good point, the revised Fig 1A is now a summary of the web tool

I realize I have written a fairly critical review that will require most of the figures to be redone, but I think the underlying method is sound and the implementation by and end-user is quite simple, so I think your group should have no trouble addressing these points.

Response: Your comments were really helpful and we feel that the tool has significantly improved as a result. So, we are thankful to the time and effort put toward helping here.

Reviewer #2 (Recommendations For The Authors)

Comments on the use of "genetic correlation"

• The use of "genetic correlation" in title and throughout the manuscript is misleading. Should broadly be replaced with "gene expression correlation". Within genetics, "genetic correlation" generally refers to the correlation between traits due to genetic variation, as would be expected under pleiotropy (genetic variation that affects multiple traits). Here, I think the authors are somewhat conflating "genetic" (normally referring to genetic variation) with "gene" (because the data are gene expression phenotypes). I don't think they perform any genetic analysis in the manuscript. I hope I don't sound too harsh. I think the paper still has merit and value, but it is important to correct the terminology.

Response: This was an important clarification raised by all reviewers. We apologize for the oversight. As a result, all mentions of “genetics” or “genetic data” has been removed, with the exception of introducing mouse genetic reference panels. These have generally been replaced with “transcript correlations”, “correlations” or “correlations across individuals” to avoid confusion.

• The authors note an important limitation in the Discussion that correlations don't imply a specific causal model between two genes, and furthermore note that statistical procedures (mediation and Mendelian randomization) are dependent on assumptions and really only a well-designed experiment can completely determine the relationship. This is a very important point that I greatly appreciate. I think they could even further expand this discussion. The potential relationships between gene A and gene B are more complex than causal and reactive. For example, a genetic variant or environmental exposure could regulate a gene that then has a cascade of effects on other genes, including A and B. They belong to a shared causal pathway (and are potentially biologically interesting), but it's good to emphasize that correlations can reflect many underlying causal relationships, some more or less interesting biologically.

Response: We thank the reviewer for pointing this out. We have expanded both the results and discussion sections to mention specifically how correlation between two genes can be due to a variety of parameters, often and not just encompassing their relationship. We mention the importance of considering genetic and environmental variables in these relationships as well which we feel will be an important “take-home message” for the reader. These points were also explored in the revised Fig 2 in terms of investigating broad pathway gene-gene correlation structures. As noted by the reviewer, contexts such as circadian rhythm or other variables in the data which are not fixed show much less overall significance in terms of broad relationships across organs.

• It would be good for the authors to provide more context for the methods they use, even when they are fully published. For example, stating that biweight midcorrelation (bicor) is an approach for comparing to variables that is more robust to outliers than traditional correlations and is commonly used with gene co-expression correlation.

Response: Thank you for pointing this out. A lack of method description was also an important reason for lack of clarity on other aspects so we have done our best to detail what exact approaches are being implemented and why. In the revised manuscript, we mention the usage if bicor values to limit influence of outlier individuals in driving regressions, but also point out that it is still a generalized linear model to assess relationships. We hope that the revised methods and expanded git repositories which detail each analysis provide much more transparency on what is being implemented.

• Performing a similar analysis based on genetic correlation is an interesting idea, as it would potentially simplify the underlying causal models (removing variation that doesn't stem from genetic variants). I don't expect the authors to do this for this paper because it would be a significant amount of work (fitting and testing genetic correlations are not as straightforward). But still, an interesting idea to think about, and individuals in GTEx are genotyped I believe. Could be mentioned in the Discussion.

Response: Absolutely. While we did not implement and models of genetic correlation (despite misusing the term) in this analysis. We have added to the discussion on how when genetic data is available, these approaches offer another way to tease out potentially causal interactions among the large amount of correlated data occurring for a variety of reasons.

Comments on use of the term "local" and "regression"

• "Local" is largely used to mean within-tissue, so how correlated gene X in tissue Y is with other genes in tissue Y. I think this needs to be defined explicitly early in the manuscript or possibly replaced with something like "within-tissue".

Response: We have replaced al “local” mentions with “within-tissue” or simply name the tissue that the gene is expressed to avoid confusion with other terms of local (ex a transcript in proximity to where it is encoded on the genome).

• "Regression" is also used frequently throughout, often when I think "correlation" would be more accurate. It's true that the regression coefficient is a function of the correlation between X and Y, but I don't think actual regression (the procedure) applies here. The coefficients being used are bicor, which I don't think relates as cleanly to linear regression.

Response: Thank you for pointing this out. A lack of method description was also an important reason for lack of clarity on other aspects so we have done our best to detail what exact approaches are being implemented and why. In the revised manuscript, we mention the usage if bicor values to limit influence of outlier individuals in driving correlations, but also point out that it is still a generalized linear model to assess relationships. Further, we have removed usage of “regression” when referencing bicor values. We hope that the revised methods and expanded git repositories which detail each analysis provide much more transparency on what is being implemented.

• "Further, pan-tissue correlations tend to be dominated by local regressions where a given gene is expressed. This is due to the fact that within-tissue correlations could capture both the regulatory and putative consequences of gene regulation, and distinguishing between the two presents a significant challenge" (lines 219-223). This sentence includes both "local" and "regressions" (and would be improved by my suggested changes I think), but I also don't fully understand the argument of "regulatory and putative consequences". I think the authors should elaborate further. In the examples, the within-tissue correlations do look stronger, suggesting within-tissue regulation that is quite strong and potentially secondary inter-tissue regulation. If that's the idea, I think it can be stated more clearly.

Response: Thank you for pointing this out. We have revised the sentence to state the following:

Further, many correlations tend to be dominated by genes expressed within the same organ. This could be due to the fact that, within-tissue correlations could capture both the pathways regulating expression of a gene, as well as potential consequences of changes in expression/function, and distinguishing between the two presents a significant challenge. For example, a GD-CAT query of insulin (INS) expression in pancreas shows exclusive enrichments in pancreas and corresponding pathway terms reflect regulatory mechanisms such as secretion and ion transport (Supplemental Fig 4).

We feel that this point might not be intuitive, so have included a new figure (Supplemental Fig 4) which contains the tissue correlations and pathways for INS expression in pancreas. These analyses show an example where co-correlation structure seems almost entirely dominated by genes within the same organ (pancreas) and GSEA enrichments highlight many known pathways which are involved in regulating the expression/secretion of the gene/protein. We hope that this makes the point more clearly to the reader.

Additional comments on Results:

• I would break the titled Results sections into multiple paragraphs. For example, the first section (lines 84-129) has a few natural breakpoints that I noticed that would potentially make it feel less over-whelming to the reader.

Response: We have broken up the results section into separate paragraphs in the revised manuscript. In addition, we have gone through to try and make sure that the amount of information per block/sentence focuses on key points.

• "Expression of a gene and its corresponding protein can show substantial discordances depending on the dataset used" (line 224 of Results). This is a good point, and the authors could include citations here of studies that show discordance between transcripts and proteins, of which there are a good number. They could also add some biological context, such as saying differences could reflect post-translational regulation, etc.

Response: Thank you for the supportive comment. We have referenced several comprehensive reviews of the topic, each of which contain tables summarizing details of mRNA-protein correlation. The revised discussion sentence is as follows:

Expression of a gene and its corresponding protein can show substantial discordances depending on the dataset used. These have been discussed in detail39–41, but ranges of co-correlation can vary widely depending on the datasets used and approaches taken. We note that for genes encoding proteins where actions from acute secretion grossly outweigh patterns of gene expression, such as insulin, caution should be taken when interpreting results. As the depth and availability of tissue-specific proteomic levels across diverse individuals continues to increase, an exciting opportunity is presented to explore the applicability of these analyses and identify areas when gene expression is not a sufficient measure.

1. Liu, Y., Beyer, A. & Aebersold, R. On the Dependency of Cellular Protein Levels on mRNA Abundance. Cell 165, 535–550 (2016).

2. Maier, T., Güell, M. & Serrano, L. Correlation of mRNA and protein in complex biological samples. FEBS Letters 583, 3966–3973 (2009).

3. Buccitelli, C. & Selbach, M. mRNAs, proteins and the emerging principles of gene expression control. Nat Rev Genet 21, 630–644 (2020).

• In many ways, this work has similar goals to many studies that have performed multi-tissue WGCNA (e.g., Talukdar et al. 2016; https://doi.org/10.1016/j.cels.2016.02.002). In this manuscript, WGCNA's conventional approach to estimating robust correlations (bicor) is used, but they do not use WGCNA's data reduction/clustering functionality to estimate modules. Perhaps the modules would miss the signaling relationships of interest, being sort of lost in the presence of stronger signals that aren't relevant to the biological questions here. But I think it would be good for the authors to explain why they didn't use the full WGCNA approach.

Response: This is an important point and we also feel that the previous lack of methodological details and discussion did a poor job at distinguishing why module-based approaches were not used. We wanted to be careful not to emphasize one approach being superior/inferior to another, rather point out the different considerations and when a direct correlation might inform a given question. As the reviewer points out, our general feeling is that adopting a simple gene-focused correlation approach allows users to view mechanisms through the lens of a single gene; however, this is limited in that these could be influenced by cumulative patterns of correlation structure (for example mitochondria in revised Fig 2A) which would be much more apparent in a module-based approach. This comment, in combination with the other listed above, was our motivation in exploring cumulative patterns of gene-gene correlations in the revised Fig 2. In the revised manuscript, we expanded on the results and discussion section to highlight utility of these types of approaches compared to module-based methods:

The queries provided in GD-CAT use fairly simple linear models to infer organ-organ signaling; however, more sophisticated methods can also be applied in an informative fashion. For example, Koplev et al generated co-expression modules from 9 tissues in the STARNET dataset, where construction of a massive Bayesian network uncovered interactions between correlated modules6. These approaches expanded on analysis of STAGE data to construct network models using WGCNA across tissues and relating these resulting eigenvectors to outcomes42. The generalized approach of constructing cross-tissue gene regulatory modules presents appeal in that genes are able to be viewed in the context of a network with respect to all other gene-tissue combinations. In searching through these types of expanded networks, individuals can identify where the most compelling global relationships occur. One challenge with this type of approach; however, is that coregulated pathways and module members are highly subjective to parameters used to construct GRNs (for example reassignment threshold in WGCNA) and can be difficult in arriving at a “ground truth” for parameter selection. We note that the WGCNA package is also implemented in these analyses, but solely to perform gene-focused correlations using biweight midcorrelation to limit outlier inflation. While the midweight bicorrelation approach to calculate correlations could also be replaced with more sophisticated models, one consideration would be a concern of overfitting models and thus, biasing outcomes.

Additional comments on Discussion:

• In the second paragraph of the Discussion (lines 231-244), the authors mention that GD-CAT uses linear models to compare data between organs and point to other methods that use more complex or elaborate models. It's good to cite these methods, but I think they could more directly state that there are limitations to high complexity models, such as over-fitting.

Response: Thank you for this suggestion. We have added a line (above) mentioning the overfitting concern.

Comments on Methods:

• The described gene filtration in the Methods of including genes with non-zero expression for 1.2e6 gene-tissue combinations is confusing. If there are 310 individuals and 18 tissues, for a given gene, aren't there only 5,580 possible data points? Might be helpful to contextualize the cut-off in terms of like the average number of individuals with non-zero expression within a tissue.

Response: We apologize for this error. This number was pasted from a previous dataset used and not appropriate for this manuscript. In general, we have removed specific mentions of total number of gene_tissue correlation combinations, as these numbers reflect large but almost meaningless quantifications. Instead, we expanded the methods in terms of how individuals and genes filtered.

• More details should be given about the gene ontology/pathway enrichment analysis. I suspect that a set-based approach (e.g., hypergeometric test) was used, rather than a score-based approach. The authors don't state what universe of genes were used, i.e., the overall set of genes that the reduced set of interest is compared to. Seems like this could or should vary with the tissues that are being compared. A score-based approach could be interesting to consider (https://www.biorxiv.org/content/10.1101/060012v3), using the genetic correlations as the score, as this would remove the unappealing feature of sets being dependent on correlation thresholds. This isn't something that I would demand of the published paper, but it could be an appealing approach for the authors to consider and confirm similar results to the set-based analysis.

Response: This is an important point. Following this suggestion, we evaluated several different rank- and weight-based pathway enrichment tools, including FGSEA and others. Ultimately, we concluded that GSEA performed significantly better at 1) recapitulating known biology of select secreted protein genes and 2) leveraging the large numbers of genes occurring at qvalue cutoffs without having to further refine (ex. in the previous overrepresentation tests). For this reason, all pathway enrichments in the web tools and manuscripts not contain GSEA outputs and corresponding pathway enrichments or network graph visualizations. Thank you for this suggestion.

Comments on figures:

• I think there is a bit of a missed opportunity to use the figures to introduce and build up the story for readers. For example, in Figure 1, plotting ADIPOQ expression against a correlated gene in adipose (local) as well as peripheral tissues. This doesn't need to be done for every example, but I think it would help readers understand what the data are, and what's being detected before jumping into higher level summaries.

Response: Thank you, this point also builds on others which recommended to restructure the manuscript and figures. In the revised manuscript, we first introduce the web tool (which was last previously), and immediately highlight comparisons of within- and across-organ correlations, such as ADIPOQ. We feel that the revised manuscript presents a superior structure in terms of demonstrating the key points and utility of looking at gene-gene correlations across tissues.

• Figures 1 and 4 are missing the color scale legend for the bar plots, so it's impossible to tell how significant the enrichments are.

Response: We apologize for the oversight. The pathways in the revised Fig 1 detail pathway network graphs among the top pathways which should make interpretation more intuitive. We have also gone through and made sure that GSEA enrichment pvalues are now present for all figures including pathways (revised Fig 1, Fig 3 and supplemental Fig 4).

• The Figure 2 caption says that edges are colored based on correlation sign? Are there any negative correlations (red)? They all look blue to me. The caption could also state that edge weight reflects correlation magnitude (I assume). It would be ideal to include a legend that links a range of the depicted edge weights to their genetic correlation, though I don't know how feasible that may be depending on the package being used to plot the networks.

Response: Good catch. We included in the revised manuscript the network edge parameters: Network edges represent positive (blue) and negative (red) correlations and the thicknesses are determined by coefficients. They are set for a range of bicor=0.6 (minimum to include) to bicor=0.99

Related to seeing a dominant pattern of positive correlations, we agree that this observation is fascinating and gene-gene correlations being dominated by positive coefficients will be the topic of a closely-following manuscript from the lab

• Figure 4A would be more informative as boxplots, which could still include Ssec score. This would allow the reader to get a sense of the variation in correlation p-value across all hippocampus transcripts.

Response: Related to comments from this reviewer and others, we have removed the previous Fig 4 entirely from the manuscript to emphasize the ability of these gene-gene correlations to capture known biology and limit the extend of unvalidated “suggested” new mechanisms.

Comments on GD-CAT

• The online webtool worked nicely for me. It was easy to use and produce figures like in the manuscript. One suggestion is show data points in the scatter plot rather than just the regression line (if that's possible currently, I didn't figure it out). A regression line isn't that interesting to look at, but seeing how noisy the data look around it is something humans can usually interpret intuitively.

Response: Thank you so much. We are excited that the web tool works sufficiently. We have also revised the individual gene-gene correlation tab to show individual data points instead of simple regression lines.

Minor comments:

Response: Thank you for these detailed improvements

• This sentence is awkwardly constructed: "Here, we surveyed gene-gene genetic correlation structure for ~6.1x10^12 gene pairs across 18 metabolic tissues in 310 individuals where variation of genes such as FGF21, ADIPOQ, GCG and IL6 showed enrichments which recapitulate experimental observations" (lines 68-70). It's an important sentence because it's where in the Abstract/Introduction the authors succinctly state what they did, thus I would re-work it to something like: "Here, we surveyed gene expression correlation structure..., identifying genes, such as FGF21, ADIPOQ, GCG and IL6, that possess correlation networks that recapitulate known biological pathways."

Response: The numbers of pairs examined and dataset size have been removed for clarity and we have revised this statement and results as a whole

• Prefer swapping "signal" for "signaling" in line 53 of Abstract/Introduction.

Response: Done

• Remove extra period in line 208 of Results.

Response: Removed

• Change "well-establish" to "well-established" in line 247 of Discussion.

Response: Replaced

• Missing commas in line 302 of Methods.

Response: added

• Missing comma in line 485 of Figure 3 caption.

Response: The previous Fig 3 has been removed

• Typo in title of Figure 3E (change "Perihperal" to "Peripheral")

Response: Thank you, changed

• Add y-axis label to y-axis labels (relative cell proportions) to Supplemental Figures 1-3.

Response: These labels have been added

Reviewer #3 (Recommendations For The Authors):

Minor technical comment: The authors refer to correlations between genes when they actually mean correlations between GTEX transcript isoform models. It is exceedingly important to keep this distinction clear in the reader's mind, a fact that is emphasized by the authors themselves when they comment on the potential value of similar proteomic assays to evaluate multiorgan system communication. GTEx has tried to do proteomics but I do not know of any open data yet.

Response: Thank you for this point. We have gone through the manuscript and replaced “gene correlations” with “transcript” or other similar mentions. Related to the comment on GTEx proteomics, this is an important point as well. As the reviewer mentions, proteomics has been performed on GTEx data; however, given that this dataset contains only 6 sparsely-represented individuals, analyses such as the ones highlighted in our study remain highly limited. We have added the following to the discussion: As the depth and availability of tissue-specific proteomic levels across diverse individuals continues to increase, an exciting opportunity is presented to explore the applicability of these analyses and identify areas when gene expression is not a sufficient measure. For example, mass-spec proteomics was recently performed on GTEx42; however, given that these data represent 6 individuals, analyses utilizing well-powered inter-individual correlations such as ours which contain 310 individuals remain limited n applications.

The R/Shiny companion application: The community utility of this application would be greatly improved by a link to a primer and more basic functionality. The Github site is a "work in progress" and does not include a readme file or explanation (that I could find) on the license.

Response: Thank you, we are excited that the apps operate sufficiently. We have revised the github repository entirely to contain a full walk-through of app details and parameter selections. These are meant to walk users through each step of the pipeline and discuss what is being done at each step. We agree that this updated github repository allows users to understand the details of the R/Shiny app in much more detail. We also made all the app scripts, datasets, markdown/walkthrough files and docker image fully available to enhance accessibility.

Reviewer #1 (Public Review):

Zhou et al. have slightly expanded and improved their web tool from the previous submission, fixing some small issues and adding in additional sets of data from HMDP mice. Essentially, the authors have created a tool that facilitates the integrated analysis of omics datasets (particularly transcriptomics, but could be easily adapted to include proteomics) across tissues.

The strength is that this is new; as far as I know, any other multi-tissue analysis software is relatively ad hoc and it is not easily supported by e.g. SRA/GEO, but rather you'd need to download the multiple datasets and DIY. The authors have now shown some statistically significant (albeit expected from literature) results created using their pipeline. Whether the method will be generally useful for the community depends on its further development and support, but of course whether a project is supported also depends on whether its first publication is accepted - somewhat of a Catch-22 for a reviewer. Right now, the results shown are a convincing proof-of-concept that would likely be of utility mostly to the hosting laboratory and their direct collaborators, but which, with continued development at a similar level of effort, could be more generally useful for the growing number of groups interested in cross-tissue analysis.

Reviewer #2 (Public Review):

Summary:<br /> Zhou et al. have revised their previous manuscript, which has greatly improved the quality of the work. Zhou et al. use publicly available GTEx data of 18 metabolic tissues from 310 individuals to explore gene expression correlation patterns within-tissue and across-tissues. Furthermore, they have added an analysis of data from a diverse panel of inbred mouse strains, which allows them to also incorporate data on physiological phenotypes relevant to metabolic signaling between tissues. They now focus on validating their approach to exploring signal in gene co-expression rather than emphasizing unvalidated discoveries. They provide a webtool (GD-CAT) to allow users to explore these data. Focusing more on known biology does result in the study making stronger conclusions from its data. The webtool is also improved, expanded with the mouse data, and of value to the scientific community. Their revision has also corrected key misconceptions from the initial submission and provides greater clarification of the methodologies used.

Strengths:<br /> GTEx as well as the hybrid diversity mouse panel are powerful resource for many areas of biomedicine, and this study represents a valid use of gene co-expression network methodology. They have greatly improved its description and contextualization within the gene co-expression studies. The authors previously did a good job of providing examples confirming known signaling biology and have further improved these. They have largely removed the sections on discovery of novel biology, which is potentially for the better given a lack of follow-up validation, which could be beyond the scope of this manuscript anyway. The webtool, GD-CAT, is easy to use and allows researchers with genes and tissues of interest to perform the same analyses in the GTEx and HMDP data.

Weaknesses:<br /> With the previous version, the primary weaknesses for me were key misconceptions and lack of detail in the methods, which have all been greatly improved. The manuscript could be considered more of a "Resource" than "Research", though there is value in showing how the known biology is reflected in the correlation data and could presumably be paired with validation to discover new biology. Finally, there are sentences here and there that could be rephrased to improve clarity, but overall it is greatly improved.

eLife assessment

This fundamental work significantly advances our understanding of the regulation of neurotransmitter and hormone secretion by exploring the mechanisms by which the protein complexin interacts with the release machinery and the calcium sensor synaptotagmin. The authors identify structural requirements within the protein for complexin's dual role in preventing premature vesicle release and enhancing evoked exocytosis. The evidence supporting the author's conclusions is compelling and the findings are of broad interest to neuroscientists and cell biologists.

Reviewer #1 (Public Review):

Summary:

Using chromaffin cells as powerful model systems for studying secretion, the authors study the regulatory role of complexin in secretion. Complexin is still enigmatic in its regulatory role, as it both provides inhibitory and facilitatory functions in release. The authors perform an extensive structure-function analysis of both the C- and N-terminal regions of complexin. There are several interesting findings that significantly advance our understanding of cpx/SNARe interactions in regulating release. C-terminal amphipathic helix interferes with SNARE complex assembly and thus clamps fusion. There are acidic residues in the C-term that may be seen as putative interaction partners for Synaptotagmin. The N-terminus of Complexin promoting role may be associated with an interaction with Syt1. In particular, the putative interaction with Syt1 is of high interest and supported by quite strong functional and biochemical evidence. The experimental approaches are state-of-the-art, and the results are of the highest quality and convincing throughout. They are adequately and intelligently discussed in the rich context of the standing literature. Whilst there are some concerns about whether the facilitatory actions of complexion have to be tightly linked to Syt1 interactions, the proposed model will significantly advance the field by providing new directions in future research.

I have only minor comments related to the interpretation of the data:

Fig 5 While the data very nicely show that CPX and Syt1 have interdependent interactions in the chromaffin neurons, this seems to be not the case in neurons, where the loss of complexins and synaptotagmins have additive effects, suggesting independent mechanisms (eg Xue et al., 2010). This would be a good opportunity to discuss some possible differences between secretion in endocrine cells vs neurons.

Fig 8 Shows an apparent shift in Ca sensitivity in N-terminal mutants suggesting a modification of Ca sensitivity of Syt1. Could there be also an alternative mechanism, that explains this phenotype which is based on a role of the n-term lowering the energy barrier for fusion, that in turn shifts corresponding fusion rates to take place at lower Ca saturation levels?

Reviewer #2 (Public Review):

Summary:

Complexin (Cplx) is expressed at nearly all chemical synapses. Mammalian Cplx comes in four different paralogs which are differentially expressed in different neuron types, either selectively or in combination with one or two other Cplx isoforms. Cplx binds with high affinity to assembled SNARE complexes and promotes AP-evoked release by increasing vesicle fusogenicity. Cplx is assumed to preclude premature SV fusion by preventing full SNARE assembly, thereby arresting subsequent SNARE-driven fusion ("fusion-clamp" theory). The protein has multiple domains, the functions of which are controversially discussed. Cplx's function has been studied in a variety of model organisms including mice, flies, worms, and fish with seemingly conflicting results which led to partly contradicting conclusions.

Makee et al. study the function of mammalian Cplx2 by making use of chromaffin cells derived from Cplx2 ko mice as a system to overexpress and functionally characterize mutant Cplx2 forms. This work is an important extension of previous studies of the same lab using similar techniques. The main conclusion of the present study are:

The hydrophobic character of the amphipathic helix in Cplx's C-terminal domain is essential for inhibiting premature vesicle fusion at a [Ca2+]i of several hundreds of nM (pre-flash [Ca2+]i). The Cplx-mediated inhibition of fusion under these conditions does not rely on the expression of either Syt1 or Syt7.

Slow-down of exocytosis by N-terminally truncated Cplx mutants in response to a [Ca2+]i of several µM (peak flash [Ca2+]i) occurs regardless of the presence or absence of Syt7 demonstrating that Cplx2 does not act as a switch favoring preferential assembly of the release machinery with Syt1,2 rather than the "slow" sensor Syt7.

Cplx's N-terminal domain is required for the Cplx2-mediated increase in the speed of exocytosis and faster onset of exocytosis which likely reflect an increased apparent Ca2+ sensitivity and faster Ca2+ binding of the release machinery.

Strengths:

The authors perform systematic truncation/mutational analyses of Cplx2 by making use of chromaffin cells derived from Cplx2 ko mice. They analyze the impact of single and combined deficiencies for Cplx2 and Syt1 to establish interactions of both proteins.

State-of-the-art methods are employed: Vesicle exocytosis is assayed directly and with high resolution using capacitance measurements. Intracellular [Ca2+] is controlled by loading via the patch-pipette and by UV-light-induced flash-photolysis of caged [Ca2+]. The achieved [Ca2+ ] is measured with Ca2+ -sensitive dyes.

The data is of high quality and the results are convincing.

Weaknesses:

The authors provide a "chromaffin cell-centric" view of the function of mammalian Cplx in vesicle fusion. With the exception of mammalian retinal ribbon synapses (and some earlier RNAi knockdown studies that had off-target effects), there is very little evidence for a "fusion-clamp"-like function of Cplxs in mammalian synapses. At conventional mammalian synapses, genetic loss of Cplx (i.e. KO) consistently decreases AP-evoked release, and generally either also decreases spontaneous release rates or does not affect spontaneous release, which is inconsistent with a "fusion-clamp" theory. This is in stark contrast to invertebrate (D. m. and C. e.) synapses where genetic Cplx loss is generally associated with strong upregulation of spontaneous release, providing support for Cplx acting as a "fusion-clamp".

The authors use a Semliki Forest virus-based approach to express mutant proteins in chromaffin cells. This strategy leads to a strong protein overexpression (~7-8fold, Figure 3 Suppl. 1). Therefore, experimental findings under these conditions may not necessarily be identical to findings with normal protein expression levels.

Measurements of delta Cm in response to Ca2+ uncaging by ramping [Ca2+ ] from resting levels up to several µM over a time period of several seconds were used to establish changes in the release rate vs [Ca2+ ]i relationship. It is not clear to this reviewer if and how concurrently occurring vesicle endocytosis together with a possibly Ca2+-dependent kinetics of endocytosis may affect these measurements.

It should be pointed out that an altered "apparent Ca2+ affinity" or "apparent Ca2+ binding rate" does not necessarily reflect changes at Ca2+-binding sites (e.g. Syt1).

There are alternative models on how Cplx may "clamp" vesicle fusion (see Bera et al. 2022, eLife) or how Cplx may achieve its regulation of transmitter release without mechanistically "clamping" fusion (Neher 2010, Neuron). Since the data presented here cannot rule out such alternative models (in this reviewer's opinion), the authors may want to mention and briefly discuss such alternative models.

Some parts of the Discussion are quite general and not specifically related to the results of the present study. The authors may want to consider shortening those parts.

Last but not least, the presentation of the results could be improved to make the data more accessible to non-specialists, this concerns providing necessary background information, choice of colors, and labeling of diagrams.

eLife assessment

This important study proposes a new method for tracking neurons recorded with Neuropixel electrodes across days. The methods and the strength of the evidence are convincing, but the authors do not adequately address whether their approach can be generalized to other brain areas, species, behaviors, or tools. Overall, this method will be potentially useful to many neuroscientists who want to study long-term activity changes of individual neurons in the brain.

Reviewer #1 (Public Review):

The brain's code is not static. Neuronal activity patterns change as a result of learning, aging, and disease. Reliable tracking of activity from individual neurons across long time periods would enable detailed studies of these important dynamics. For this reason, the authors' efforts to track electrophysiological activity across days without relying on matching neural receptive fields (which can change due to learning, aging, and disease) are very important.

By utilizing the tightly-spaced electrodes on Neuropixels probes, they are able to measure the physical distance and the waveform shape 'distance' between sorted units recorded on different days. To tune the matching algorithm and validate the results, they used the visual receptive fields of neurons in the mouse visual cortex (which tend to change little over time) as ground truth. Their approach performs quite well, with a high proportion of neurons accurately matched across multiple weeks. This suggests that the method may be useable in other cases where the receptive fields can't be used as ground truth to validate the tracking. This potential extendibility to tougher applications is where this approach holds the most promise.

The main caveat (and disappointment) is that this paper does not address generalizability to other experimental conditions. Because it only looks at one brain area (visual cortex), in one species (mouse), using one type of spike sorter (Kilosort), and one type of behavioral prep (head-fixed), it is not clear if this approach is overfit to those conditions or if it will perform equally well in other conditions. Most importantly, in brain areas where neuronal receptive fields are more dynamic and can't be used as a ground truth diagnostic, it isn't clear how to apply the technique outlined in this study, since many of the parameters are tuned to a very specific set of conditions using visual receptive fields as ground truth.

Reviewer #2 (Public Review):

The manuscript presents a method for tracking neurons recorded with neuropixels across days, based on the matching of cells' spatial layouts and spike waveforms at the population level. The method is tested on neuropixel recordings of the visual cortex carried over 47 days, with the similarity in visual receptive fields used to verify the matches in cell identity.

This is an important tool as electrophysiological recordings have been notoriously limited in terms of tracking individual neuron's fate over time, unlike imaging approaches. The method is generally sound and properly tested but I think some clarifications would be helpful regarding the implementation of the method and some of the results.

1) Page 6: I am not sure I understand the point of the imposed drift and how the value of 12µm is chosen.<br /> Is it that various values of imposed drift are tried, the EMDs computed to produce histograms as in Fig2c, values of rigid drifts estimated based on the histogram modes, and then the value associated with minimum cost selected? The corresponding manuscript section would need some clarification regarding this aspect.

2) The EMD is based on the linear sum, with identical weight, of cell distance and waveform similarity measures. How performance is affected by using a different weighting of the 2 measures (for instance, using only cell distance and no waveform similarity)? It is common that spike waveforms associated with a given neuron appear differently on different channels of silicon probes (i.e. the spike waveform changes depending on the position of recording sites relative to the neuron), so I wonder if that feature is helping or potentially impeding the tracking.

3) Fig.5: I assume the dots represent time gaps for which cell tracking is estimated. The 3 different groups of colors correspond to the 3 mice used. For a given mouse, I would expect to always see 3 dots (for ref, putative, and mixed) for a given tracking gap. However, for mouse AL036 for instance, at a tracking duration of 8 days, a dot is visible for mixed but not for ref and putative. How come this is happening?

4) Matched visual responses are measured by the sum of the correlation of visual fingerprints, which are vectors of cells' average firing rate across visual stimuli, and the correlation of PSTHs, which are implemented over all visual stimuli combined. I believe that some information is lost from combining all stimuli in the implementation of PSTHs (assuming that PSTHs show specificity to individual visual stimuli). The authors might consider, as an alternative measure of matched visual responses, a correlation of the vector concatenations of all stimulus PSTHs. Such a simpler measure would contain both visual fingerprint and PSTH information, and would not lose the information of PSTH specificity across visual stimuli.

eLife assessment

This study provides valuable insights into how chromatin-bound PfMORC controls gene expression in the asexual blood stage of Plasmodium falciparum. By interacting with key nuclear proteins, PfMORC appears to affect expression of genes relating to host invasion and subtelomeric var genes. Correlating transcriptomic data with in vivo chromatin insights, the study provides solid evidence for the central role of PfMORC in epigenetic transcriptional regulation through modulation of chromatin compaction.

eLife assessment

The findings presented by Huff and colleagues describe different motor patterns of swallowing following optogenetic activation of the Postinspiratory Complex (PiCo) in a group of mice exposed to Chronic Intermittent Hypoxia (CHI). The presented results are important, and the experimental procedures are rigorous and technically remarkable, but drawing meaningful conclusions is currently not obvious due to some bias in statistical comparisons that require consideration. The strength of the evidence is currently incomplete and would benefit from additional experiments. Overall this work would be of interest to the field of respiratory physiology and pathophysiology since a disruption of swallowing and possibly discoordination with breathing may be involved in diseases characterized by the presence of hypoxic conditions such as obstructive sleep apnea.

Author Response

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

Reviewer #1

The study provides a complete comparative interactome analysis of α-arrestin in both humans and drosophila. The authors have presented interactomes of six humans and twelve Drosophila α-arrestins using affinity purification/mass spectrometry (AP/MS). The constructed interactomes helped to find α-arrestins binding partners through common protein motifs. The authors have used bioinformatic tools and experimental data in human cells to identify the roles of TXNIP and ARRDC5: TXNIP-HADC2 interaction and ARRDC5-V-type ATPase interaction. The study reveals the PPI network for α-arrestins and examines the functions of α-arrestins in both humans and Drosophila.

Comments

I will like to congratulate the authors and the corresponding authors of this manuscript for bringing together such an elaborate study on α-arrestin and conducting a comparative study in drosophila and humans.

Introduction:

The introduction provides a rationale behind why the comparison between humans and Drosophila is carried out.

• Even though this is a research manuscript, including existing literature on similar comparison of α-arrestin from other articles will invite a wide readership.

Results:

The results cover all the necessary points concluded from the experiments and computational analysis.

1) The authors could point out the similarity of the α-arrestin in both humans and Drosophila. While comparing α-arrestin in both humans and Drosophila If percentage homology between α-arrestin of both Drosophila and humans needs to be calculated.

Thank you for your insightful feedback. As suggested by reviewer, we determined percentage homology of α-arrestin protein sequences from human and Drosophila using Clustal Omega. This homology is now illustrated as a heatmap in revised Figure S5. Please note that only the values with percentage homology of 40% or higher are selectively labeled.

• Citing the direct connecting genes from the network in the text will invite citations and a wider readership.

Figures:

The images are elaborate and well-made.

2) The authors could use a direct connected gene-gene network that pointing interactions. This can be used by other readers working on the same topic and ensure reproducibility and citations.

We appreciate your valuable comment. Based on the reviewer’s suggestion, we have developed a new website in which one can navigate the gene-gene networks of α-arrestins. These direct connected gene-gene networks are housed in the network data exchange (NDEx) project. Additionally, we have included gene ontology and protein class details for α-arrestins’ interactors in these set of networks, offering a more comprehensive view of α-arrestins’ interactomes.

On page 24 lines 15-18, we have revised the manuscript to introduce the newly developed website, as follows.

“Lastly, to assist the research community, we have made comprehensive α-arrestin interactome maps on our website (big.hanyang.ac.kr/alphaArrestin_PPIN). Researchers can search and download their interactomes of interest as well as access information on potential cellular functions and protein class associated with these interactomes.”  

3-1) The co-expression interactions represented as figures should reveal interaction among the α-arrestin and other genes. Which are the sub-network genes does the α- arrestin interact to/ with from the sub-network? The arrows are only pointing at the sub-networks. The figures do not reveal their interaction. Kindly reveal the interaction in the figure with the proper nodes in the figure.

3-2) Figure 2: the network attached in both human and drosophila is well represented. The green lines from α-arrestin indicate the strength of the interaction. Several smaller expression networks are seen. But "α-arrestin" in both organisms seems highly disconnected from all the genes. Connected genes have edges, not arrows. If α-arrestin can be shown connected to these gene-gene networks will help in identifying which genes connect with which gene through α-arrestin. This can be used by other readers working on the same topic and ensure reproducibility and citations.

Thank you for your valuable comment. In response to the reviewer’s recommendation, we’ve added supplementary figure, Figure S4, which illustrates direct interaction between α-arrestin and protein components of clustered complexes (or sub-networks) in addition to the associations shown between α-arrestins and the clustered complexes in Figure 2. We believe that this newly incorporated information regarding direct protein interactions will invite citations and wider readership as the reviewer pointed out.

On page 12 line 27 to page 13 line 5, we have revised the manuscript to cite the direction interactions between ARRDC3 and proteins involved in ubiquitination-dependent proteolysis, as follows.

“While the association of ARRDC3 with these ubiquitination-dependent proteolysis complexes is statistically insignificant, ARRDC3 does interact with individual components of these complexes such as NEDD4, NEDD4L, WWP1, and ITCH (Figure S4A). This suggest their functional relevance in this context, as previously reported in both literatures and databases (Nabhan et al., 2010; Shea et al., 2012; Szklarczyk et al., 2015; Warde-Farley et al., 2010) (Puca & Brou, 2014; Xiao et al., 2018).”

Direct interaction between α-arrestins and protein components of clustered complexes are illustrated in the newly added figure, Figure S4.

4-1) Figure 4. The Protein blot image was blurred. Kindly provide a higher-resolution image.

4-2) Figure 5. B. - The authors can provide images with higher resolution blot images. The bands were not visible.

We appreciate for valuable comment. Unfortunately, the protein blot image was scanned from the original film and the images we provided in the figure represent the highest resolution that we have obtained to date. Raw, uncropped images are shown in Author response image 1 and 2.

Author response image 1.

Raw image of Figure 4B

Author response image 2.

Raw image of Figure 5B

5) Figure: 5. A. - I see non-specific amplifications in the gel images. Are these blotting images? or the gel images that were changed to "Grayscale"? Non-specific amplification may imply that the experiment was not repeated and standardized. Was it gel images or blot images?

We appreciate your insightful comment. The images in Figure 5A represent western blot bands from co-immunoprecipitation assay for analysis of the interaction between TXNIP and HDAC2 proteins. Since immunoblotting using immunoprecipitates can usually detect some non-specific bands from heavy (~ 50 kDa) and light (~25 kDa) chains of the target antibody or from multiple co-immunoprecipitated proteins, we assume that the vague non-specific bands in Figure 5A might be a heavy chain of TXNIP or HDAC2 antibody or an unclear non-specific band. Because target bands showed strong intensity and very clear pattern compared to the non-specific bands in the co-immunoprecipitation assay, we believe that this data is sufficient to support the interaction of TXNIP with HDAC2. Finally, In the revised Figure 5A, we’ve modified the labeling for different experimental conditions, namely siCon and siTXNIP treatments, and added expected size of proteins (kDa), as shown below.

6) Figure 5. A. RT-PCR analysis: What was your expected size of the amplifications? the ladder indicated is in KDa. Is that right?

We appreciate your insightful questions. As mentioned above, Figure 5A shows the blotting images of co-immunoprecipitation analysis, and the ladder indicates the molecular weight (kDa) of protein markers. For clearer interpretation, the expected size of target proteins has been added in Figure 5A in the revised manuscript.

7) How were the band intensities determined?

Thank you for your question. For quantification of immunoblot results, the densities of target protein bands were analyzed with Image J, as we described in the Materials and Methods.

Discussion:

The authors have utilized and discussed the conclusion they draw from their study. But could highlight more on ARRDCs and why it was selected out of the other arrestins. The authors have provided future work directions associated with their work.

8) Why were only ARRDCs presented amongst all the arrestin in the main part of the manuscript?

We’re grateful for your valuable feedback. The reason we focused on α-arrestins was that α-arrestins have been discovered relatively recently, especially when compared to more established visual/ β-arrestin proteins in the same arrestin family but the biological functions of many α-arrestins remain largely unexplored, with notable exceptions in the budding yeast model and a few α-arrestins in mammals and invertebrate species. Most importantly, comparative study highlighting the shared or unique features of α-arrestins is yet to be undertaken. To gain a more comprehensive understanding of these unexplored α-arrestins across multiple species, we’ve centered our research on the ARRDCs within the arrestin protein family.

On page 21 lines 8-17, we’ve edited the manuscript to emphasize the importance of a comparative study on α-arrestins, as detailed below.

“According to a phylogenetic analysis of arrestin family proteins, α-arrestins were shown to be ubiquitously conserved from yeast to human (Alvarez, 2008). However, compared to the more established visual/ β-arrestin proteins, α-arrestins have been discovered more recently and much of their molecular mechanisms and functions remain mostly unexplored except for budding yeast model (Zbieralski & Wawrzycka, 2022). Based on the high-confidence interactomes of α-arrestins from human and Drosophila, we identified conserved and specific functions of these α-arrestins. Furthermore, we uncovered molecular functions of newly discovered function of human specific α-arrestins, TXNIP and ARRDC5. We anticipate that the discovery made here will enhance current understanding of α-arrestins.”

9) The discussion could be elaborated more by utilizing the data.

We appreciate your insightful feedback. Based on the reviewer’s suggestion, we’ve enhanced the discussion in the manuscript to provide a clearer interpretation of our results. First, we’ve added description of conserved protein complexes significantly associated with α-arrestins, stated on page 22 lines 5-12 and lines 23-26.

Page 22 lines 5-12: “The integrative map of protein complexes also highlighted both conserved and unique relationships between α-arrestins and diverse functional protein complexes. For instance, protein complexes involved in ubiquitination-dependent proteolysis, proteasome, RNA splicing, and intracellular transport (motor proteins) were prevalently linked with α-arrestins in both human and Drosophila. To more precisely identify conserved PPIs associated with α-arrestins, we undertook ortholog predictions within the α-arrestins’ interactomes. This revealed 58 orthologous interaction groups that were observed to be conserved between human and Drosophila (Figure 3).”

Page 22 lines 23-26: “Additionally, interaction between α-arrestins and entities like motor proteins, small GTPase, ATP binding proteins, and endosomal trafficking components were identified to be conserved. Further validation of these interactions could unveil molecular mechanisms consistently associated with these cellular functions.”

Secondly, we’ve added description of role of ARRDC5 in osteoclast maturation, as stated on page 23 lines 22-24.

“Conversely, depletion of ARRDC5 reduces osteoclast maturation, underscoring the pivotal role of ARRDC5 in osteoclast development and function (Figure S9A and B).”

Lastly, we examined the association between α-arrestins’ interactomes and human diseases, incorporating our findings into the discussion. The newly introduced figure based on the result is Figure S10.

On page 24 lines 10-14, we’ve added discussion on Figure S10 as follows.

“We further explored association between α-arrestins’ interactomes and disease pathways (Figure S10). Notably, the interactomes of α-arrestins in human showed clear links to specific diseases. For instance, ARRDC5 is closely associated with disease resulting from viral infection and cardiovascular conditions. ARRDC2, ARRDC4, and TXNIP share common association with certain neurodegenerative diseases, while ARRDC1 is implicated in cancer.”

Supplementary figures:

The authors have a rigorous amount of work added together for the success of this manuscript.

10) The reference section needs editing before publication. Maybe the arrangement was disturbed during compiling.

Thank you for your valuable comment. Based on the reviewer’s suggestion, we have rearranged the reference section to enhance its clarity. Below are excerpts from the update reference section in the manuscript.

“Adenuga, D., & Rahman, I. (2010). Protein kinase CK2-mediated phosphorylation of HDAC2 regulates co-repressor formation, deacetylase activity and acetylation of HDAC2 by cigarette smoke and aldehydes. Arch Biochem Biophys, 498(1), 62-73. doi:10.1016/j.abb.2010.04.002

Adenuga, D., Yao, H., March, T. H., Seagrave, J., & Rahman, I. (2009). Histone Deacetylase 2 Is Phosphorylated, Ubiquitinated, and Degraded by Cigarette Smoke. American Journal of Respiratory Cell and Molecular Biology, 40(4), 464-473. doi:10.1165/rcmb.2008-0255OC

Akalin, A., Franke, V., Vlahovicek, K., Mason, C. E., & Schubeler, D. (2015). Genomation: a toolkit to summarize, annotate and visualize genomic intervals. Bioinformatics, 31(7), 1127-1129. doi:10.1093/bioinformatics/btu775

Alvarez, C. E. (2008). On the origins of arrestin and rhodopsin. BMC Evol Biol, 8, 222. doi:10.1186/1471-2148-8-222”

11) many important references were missing.

We appreciate and agree with the reviewer’s comment. In response to the reviewer’s recommendation, we’ve thoroughly reviewed the manuscript and below are sections of the manuscript where around 20 new references have been added.

On page 8 lines 12-14:

“Utilizing the known affinities between short linear motifs in α-arrestins and protein domains in interactomes(El-Gebali et al., 2019; UniProt Consortium, 2018) “

On page 8 lines 19-22:

“One of the most well-known short-linear motifs in α-arrestin is PPxY, which is reported to bind with high affinity to the WW domain found in various proteins, including ubiquitin ligases (Ingham, Gish, & Pawson, 2004; Macias et al., 1996; Sudol, Chen, Bougeret, Einbond, & Bork, 1995)”

On page 9 lines 3-6:

“Next, we conducted enrichment analyses of Pfam proteins domains (El-Gebali et al., 2019; Huang da, Sherman, & Lempicki, 2009b) among interactome of each α-arrestin to investigate known and novel protein domains commonly or specifically associated (Figure S3A; Table S5).”

On page 9 lines 7-10:

“HECT and C2 domains are well known to be embedded in the E3 ubiquitin ligases such as NEDD4, HECW2, and ITCH along with WW domains (Ingham et al., 2004; Melino et al., 2008; Rotin & Kumar, 2009; Scheffner, Nuber, & Huibregtse, 1995; Weber, Polo, & Maspero, 2019)”

On page 10 lines 12-16:

“In fact, the known binding partners, NEDD4, WWP2, WWP1, and ITCH in human and CG42797, Su(dx), Nedd4, Yki, Smurf, and HERC2 in Drosophila, that were detected in our data are related to ubiquitin ligases and protein degradation (C. Chen & Matesic, 2007; Ingham et al., 2004; Y. Kwon et al., 2013; Marin, 2010; Melino et al., 2008; Rotin & Kumar, 2009) (Figure 1E; Figure S2F).”

On page 13 lines 20-21:

“Given that α-arrestins are widely conserved in metazoans (Alvarez, 2008; DeWire, Ahn, Lefkowitz, & Shenoy, 2007), “

On page 14 lines 12-17:

“The most prominent functional modules shared across both species were the ubiquitin-dependent proteolysis, endosomal trafficking, and small GTPase binding modules, which are in agreement with the well-described functions of α-arrestins in membrane receptor degradation through ubiquitination and vesicle trafficking (Dores et al., 2015; S. O. Han et al., 2013; Y. Kwon et al., 2013; Nabhan et al., 2012; Puca & Brou, 2014; Puca et al., 2013; Shea et al., 2012; Xiao et al., 2018; Zbieralski & Wawrzycka, 2022) (Figure 3).”  

Reviewer #2

In this manuscript, the authors present a novel interactome focused on human and fly alpha-arrestin family proteins and demonstrate its application in understanding the functions of these proteins. Initially, the authors employed AP/MS analysis, a popular method for mapping protein-protein interactions (PPIs) by isolating protein complexes. Through rigorous statistical and manual quality control procedures, they established two robust interactomes, consisting of 6 baits and 307 prey proteins for humans, and 12 baits and 467 prey proteins for flies. To gain insights into the gene function, the authors investigated the interactors of alpha-arrestin proteins through various functional analyses, such as gene set enrichment. Furthermore, by comparing the interactors between humans and flies, the authors described both conserved and species-specific functions of the alpha-arrestin proteins. To validate their findings, the authors performed several experimental validations for TXNIP and ARRDC5 using ATAC-seq, siRNA knockdown, and tissue staining assays. The experimental results strongly support the predicted functions of the alpha-arrestin proteins and underscore their importance. `

I would like to suggest the following analyses to further enhance the study:

1) It would be valuable if the authors could present a side-by-side comparison of the interactomes of alpha-arrestin proteins, both before and after this study. This visual summary network would demonstrate the extent to which this work expanded the existing interactome, emphasizing the overall contribution of this study to the investigation of the alpha-arrestin protein family.

We greatly appreciate your insightful feedback. In response to the reviewer’s suggestion, we’ve depicted a network of known PPIs associated with α-arrestins (Figure S2C and D). Furthermore, by comparing our high-confidence PPIs to these known sets, we found that the overlaps are statistically significant and the high-confidence PPIs of α-arrestins broaden the existing interactome (Figure S2E).

From page 7 line 26 to page 8 line 8, we’ve detailed this side-by-side comparisons of existing interactome and newly discovered high-confidence PPIs of α-arrestins, as outline below.

“As a result, we successfully identified many known interaction partners of α-arrestins such as NEDD4, WWP2, WWP1, ITCH and TSG101, previously documented in both literatures and PPI databases (Figure S2C-F) (Colland et al., 2004; Dotimas et al., 2016; Draheim et al., 2010; Mellacheruvu et al., 2013; Nabhan et al., 2012; Nishinaka et al., 2004; Puca & Brou, 2014; Szklarczyk et al., 2015; Warde-Farley et al., 2010; Wu et al., 2013). Additionally, we greatly expanded repertoire of PPIs associated with α-arrestins in human and Drosophila, resulting in 390 PPIs between six α-arrestins and 307 prey proteins in human, and 740 PPIs between twelve α-arrestins and 467 prey proteins in Drosophila (Figure S2E). These are subsequently referred to as ‘high-confidence PPIs’ (Table S3).”

2) While the authors conducted several analyses exploring protein function, there is a need to further explore the implications of the interactome in human diseases. For instance, it would be beneficial to investigate the association of the newly identified interactome members with specific human diseases. Including such investigations would strengthen the link between the interactome and human disease contexts.

Thank you for your valuable comment. As suggested by the reviewer, we examined the association between α-arrestins’ interactomes and human diseases, incorporating our findings into the discussion. The newly introduced figure based on the result is Figure S10.

On page 24 lines 10-14, we’ve added discussion on Figure S10 as follows.

“We further explored association between α-arrestins’ interactomes and disease pathways (Figure S10). Notably, the interactomes of α-arrestins in human showed clear links to specific diseases. For instance, ARRDC5 is closely associated with disease resulting from viral infection and cardiovascular conditions. ARRDC2, ARRDC4, and TXNIP share common association with certain neurodegenerative diseases, while ARRDC1 is implicated in cancer.”

Reviewer #3:

Lee, Kyungtae and colleagues have discovered and mapped out alpha-arrestin interactomes in both human and Drosophila through the affinity purification/mass spectrometry and the SAINTexpress method. They found the high confident interactomes, consisting of 390 protein-protein interactions (PPIs) between six human alpha-arrestins and 307 preproteins, as well as 740 PPIs between twelve Drosophila alpha-arrestins and 467 prey proteins. To define and characterize these identified alpha-arrestin interactomes, the team employed a variety of widely recognized bioinformatics tools. These included protein domain enrichment analysis, PANTHER for protein class enrichment, DAVID for subcellular localization analysis, COMPLEAT for the identification of functional complexes, and DIOPT to identify evolutionary conserved interactomes. Through these analyses, they confirmed known alpha-arrestin interactors' role and associated functions such as ubiquitin ligase and protease. Furthermore, they found unexpected biological functions in the newly discovered interactomes, including RNA splicing and helicase, GTPase-activating proteins, ATP synthase. The authors carried out further study into the role of human TXNIP in transcription and epigenetic regulation, as well as the role of ARRDC5 in osteoclast differentiation. This study holds important value as the newly identified alpha-arrestin interactomes are likely aiding functional studies of this group of proteins. Despite the overall support from data for the paper's conclusions, certain elements related to data quantification, interpretation, and presentation demand more detailed explanation and clarification.

1) In Figure 1B, it is shown that human alpha-arrestins were N-GFP tagged (N-terminal) and Drosophila alpha-arrestins were C-GFP (C-terminal). However, the rationale of why the authors used different tags for human and fly proteins was not explained in the main text and methods.

We appreciate your valuable comment. Both N- and C-terminally tagged α-arrestins have been used previously. Given that our study aims to increase the repertoire of α-arrestin interacting proteins, where GFP is added might not be a concern. We note that GFP is a relatively bulky tag, and tagging a protein with GFP can potentially abolish the interaction with some of the binding proteins. Follow-up studies utilizing different approaches for detecting protein-protein interactions, such as BioID and yeast two-hybrid, will allow us to build more comprehensive α-arrestin interactomes.

2) In Figure 2A, there seems to be an error for labeling the GAL4p/GAL80p complex that includes NOTCH2, NOTCH1 and TSC2.

Thank you for comment. We double-checked COMPLEAT (protein COMPLex Enrichment Analysis Tool) database for the name of protein complex consisting of NOTCH1, NOTCH2, AND TSC2. The database indeed labeled this complex as the “GAL4p/GAL80p complex”. However, given the potential for mis-annotation (since we could not ascertain the relevance of these proteins to the “GAL4p/GAL80p complex”), we chose to exclude this protein complex from the network. The update protein complex network is illustrated in the revised Figure 2A.

3) In Figure 5, given that knockdown of TXNIP did not affect the levels and nuclear localization of HDAC2, the authors suggest that TXNIP might modulate HDAC2 activity. However, the ChiP assay suggest a different model - TXNIP-HDAC2 interaction might inhibit the chromatin occupancy of HDAC2, reducing histone deacetylation and increasing global chromatin accessibly. The authors need to propose a model consistent with these sets of all data.

We greatly appreciate your detailed feedback. Our data indicates a global decrease in chromatin accessibility (Figure 4C-G) and a diminished interaction between TXNIP and HDAC2 under depletion of TXNIP (Figure 5A). Additionally, we observed an increased occupancy of HDAC2 and subsequent histone deacetylation at TXNIP-target promoter regions (Figure 5C) without any changes in the HDAC2 expression level (Figure 5A) in TXNIP- knockdown cells. From these observations, we infer that the interaction between TXNIP-HDAC2 might suppress the function of HDAC2, a major gene silencer affecting the formation of condensed or accessible chromatin by deacetylating activity. Although we checked whether TXNIP could induce cytosolic retention of HDAC2 to inhibit nuclear function of HDAC2, TNXIP knockdown did not alter its subcellular localization (Figure 5B).

To elucidate the mechanism by which TXNIP inhibits the function of HDAC2, we further investigated the effect of TXNIP on the levels of HDAC2 phosphorylation, which is known to be crucial for its deacetylase activity and the formation of transcriptional repressive complex. However, as shown in the Figure S8C and D, the knockdown of TXNIP did not affect the HDAC2 phosphorylation status, as well as the interaction between HDAC2 and other components in NuRD complex in the immunoblotting and co-IP assays, respectively. The results suggest that TXNIP may inhibit the function of HDAC2 independently of these factors.

Following the reviewer’s suggestion, we carefully provided a proposed model describing the possible role of TXNIP in transcriptional regulation through interaction with HDAC2 and co-repressor complex in Figure S8E.

Description of these newly added figures can be found in the revised manuscript from page 18 line 7 to 27, as outlined below.

“HDAC2 typically operates within the mammalian nucleus as part of co-repressor complexes as it lacks ability to bind to DNA directly (Hassig, Fleischer, Billin, Schreiber, & Ayer, 1997). The nucleosome remodeling and deacetylation (NuRD) complex is one of the well-recognized co-repressor complexes that contains HDAC2 (Kelly & Cowley, 2013; Seto & Yoshida, 2014) and we sought to determine if depletion of TXNIP affects interaction between HDAC2 and other components in this NuRD complex. While HDAC2 interacted with MBD3 and MTA1 under normal condition, the interaction between HDAC2 and MBD3 or MTA1 was not affected upon TXNIP depletion (Figure S8C). Next, given that HDAC2 phosphorylation is known to influence its enzymatic activity and stability (Adenuga & Rahman, 2010; Adenuga, Yao, March, Seagrave, & Rahman, 2009; Bahl & Seto, 2021; Tsai & Seto, 2002), we tested if TXNIP depletion alters phosphorylation status of HDAC2. The result indicated, however, that phosphorylation status of HDAC2 does not change upon TXNIP depletion (Figure S8D). In summary, our findings suggest a model where TXNIP plays a role in transcriptional regulation independent of these factors (Figure S8E). When TXNIP is present, it directly interacts with HDAC2, a key component of transcriptional co-repressor complex. This interaction suppresses the HDAC2 ‘s recruitment to target genomic regions, leading to the histone acetylation of target loci possibly through active complex including histone acetyltransferase (HAT). As a result, transcriptional activation of target gene occurs. In contrast, when TXNIP expression is diminished, the interaction between TXNIP and HDAC2 weakens. This restores histone deacetylating activity of HDAC2 in the co-repressor complex, leading to subsequent repression of target gene transcription.”

4) The authors showed that ectopic expression of ARRDC5 increased osteoclast differentiation and function. Does loss of ARDDC5 lead to defects in osteoclast function and fate determination?

We appreciate your valuable comment. We have confirmed the endogenous expression of ARRDC5 in osteoclasts and conducted a loss-of-function study using shARRDC5. As determined by qPCR, ARRDC5 was endogenously expressed very low in osteoclasts. Even during RANKL-induced osteoclast differentiation, the CT value (29-31) for ARRDC5 expression was high in osteoclasts compared to the CT value (17-24) for the expression of marker genes Cathepsin K, TRAP, and NFATc1. Even though its endogenous expression was very low, we generated ARRDC5 knockdown cells by infecting BMMs with lentivirus expressing shRNA of ARRDC5 and subsequently differentiated the cells into mature osteoclasts. After five days of differentiation, we observed a significant decrease in the total number of TRAP-positive multinucleated cells (No. of TRAP+ MNCs) in shARRDC5 cells compared to that in the control cells. This result indicates that the loss of ARRDC5 leads to defects in osteoclast differentiation. Result of this loss-of-function study using shARRDC5 is depicted in Figure S9A and B.

In the revised manuscript, following sentence explaining Figure S9A and B was added on page 19 lines 15-17 as follows.

“Depletion of ARRDC5 using short hairpin RNA (shRNA) impaired osteoclast differentiation, further affirming its crucial role in this differentiation process (Figure S9A and B).”

5) From Figure 6D, the authors argued that ARRDC5 overexpression resulted in more V-ATPase signals: however, there is no quantification. Quantification of the confocal images will foster the conclusion. Also, western blots for V-ATPase proteins will provide an alternative way to determine the effects of ARRDC5.

We appreciate your insightful feedback. As suggested by the reviewer, we quantified V-type ATPase signals using confocal images, which were shown in Figure 6D. The ImageJ program was employed for integrated density measurements, and the integrated density of GFP-GFP overexpressing osteoclasts was set to 1 for relative comparison. The result in the revised Figure 6D revealed a significant increase in V-type ATPase signals in GFP-ARRDC5 overexpressing osteoclasts compared to that in GFP-GFP overexpressing osteoclasts, as outlined below.

We also agree with the reviewer’s comment that Western blot for V-ATPase proteins will be an alternative way to determine the effects of ARRDC5 in osteoclast differentiation. We have confirmed no different expression of V-type ATPase between GFP-GFP and GFP-ARRDC5 overexpressing osteoclasts using qPCR and western blot analysis. The corresponding western blot result is shown in the revised Figure S9C.

In addition, the corresponding qPCR that measures the expression level of V-type ATPase between GFP-GFP and GFP-ARRDC5 overexpressing osteoclasts is shown in Author response image 3.

Author response image 3.

Moreover, based on the references, the V-type ATPase is localized at the plasma membrane during osteoclast differentiation (Toyomura et al., 2003). Although mRNA and protein expression levels were similar in both cells, localization of V-ATPase in plasma membrane was significantly increased in GFP-ARRDC5 overexpressing osteoclasts compared to that in GFP-GFP osteoclasts, as shown in the revised Figure 6D above.

6) The results from Figure 6D did not support the authors' argument that ARRDC5 might control the membrane localization of the V-ATPase, as bafilomycin is the V-ATPase inhibitor. ARRDC5 knockdown experiments will help to determine whether ARRDC5 can control the membrane localization of the V-ATPase in osteoclast.

Thank you for your insightful comment. V-type ATPase has been reported to play an important role in the differentiation and function of osteoclasts (Feng et al., 2009; Qin et al., 2012). Given that various subunits of the V-type ATPase interact with ARRDC5 (Figure 6A), we speculated that ARRDC5 might be involved in the function of this complex and play a role in osteoclast differentiation and function. As answered above, GFP-ARRDC5 overexpressing osteoclasts showed a similar expression level of V-type ATPase to GFP-GFP cells but exhibited increased V-type ATPase signals at the cell membrane compared to those in GFP-GFP cells (Figure 6D). Additionally, co-localization of ARRDC5 and V-type ATPase was observed in the osteoclast membrane (Figure 6D), as predicted by the human ARRDC5-centric PPI network. On the other side, bafilomycin A1, a V-type ATPase inhibitor, not only blocked localization of V-type ATPase to plasma membrane in GFP-ARRDC5 overexpressing osteoclasts, but also reduced ARRDC5 signals (Figure 6D). These results indicate that ARRDC5 plays a role in osteoclast differentiation and function by interacting with V-type ATPase and promoting the localization of V-type ATPase to plasma membrane in osteoclasts.

V-type ATPase present in osteoclast membrane is important to cell fusion, maturation, and function during osteoclast differentiation (Feng et al., 2009; Qin et al., 2012). GFP-ARRDC5 overexpressing osteoclasts showed a significant increase of V-type ATPase signals in the cell membrane compared to GFP-GFP cells (Figure 6D), and also significantly increased cell fusion (No. of TRAP+ MNCs in Figure 6B) and resorption activity (resorption pit formation in Figure 6C). However, ARRDC5 knockdown in osteoclasts (shARRDC5 cells) showed a significant decrease in No. of TRAP+ MNCs compared to that in the control cells, indicating that the loss of ARRDC5 leads to defects in cell fusion during osteoclast differentiation (Figure S9A and B). As described above, the endogenous expression of ARRDC5 was very low in osteoclasts and could be specifically expressed in a certain timepoint during the differentiation. Therefore, to better understand the interaction with V-type ATPase of ARRDC5 in osteoclasts, ARRDC5 overexpression is more suitable than its knockdown.

Part of the manuscript on page 19 line 21 to page 20 line 6 was edited to support our statement, as outlined below.

“The V-type ATPase is localized at the osteoclast plasma membrane (Toyomura et al., 2003) and its localization is important for cell fusion, maturation, and function during osteoclast differentiation (Feng et al., 2009; Qin et al., 2012). Furthermore, its localization is disrupted by bafilomycin A1, which is shown to attenuate the transport of the V-type ATPase to the membrane (Matsumoto & Nakanishi-Matsui, 2019). We analyzed changes in the expression level and localization of V-type ATPase, especially V-type ATPase V1 domain subunit (ATP6V1), in GFP-GFP and GFP-ARRDC5 overexpressing osteoclasts. The level of V-type ATPase expression did not change in osteoclasts regardless of ARRDC5 expression levels (Figure S9C). GFP signals were detected at the cell membrane when GFP-ARRDC5 was overexpressed, indicating that ARRDC5 might also localize to the osteoclast plasma membrane (Figure 6D; Figure S9D). In addition, we detected more V-type ATPase signals at the cell membrane in the GFP-ARRDC5 overexpressing osteoclasts, and ARRDC5 and V-type ATPase were co-localized at the osteoclast membrane (Figure 6D; Figure S9D).”

7) The tables (excel files) do not have proper names for each table S numbers. Please correct the name of excel files for readers.

We appreciate your valuable comments. In response to the reviewer’s suggestion, we’ve renamed excel files to more appropriate titles for easier readability. List of renamed tables (excel files) are shown below.

Table S1. List of α-arrestins from human and Drosophila Table S2. Evaluation sets of α-arrestins PPIs Table S3. Summary tables of SAINTexpress results Table S4. Protein domains and short linear motifs in the α-arrestin interactomes Table S5. Enriched Pfam domains in the α-arrestin interactomes Table S6. Subcellular localizations of α-arrestin interactomes Table S7. Summary of protein complexes and cellular components associated with α-arrestin Table S8. Orthologous relationship of α-arrestin interactomes between human and Drosophila Table S9. Summary of ATAC- and RNA-seq read counts before and after processing Table S10. Differential accessibility of ACRs and gene expression Table S11. Summary of ATAC-seq peaks located in promoters and gene expression level Table S12. List of primer sequences used in this study

8) http://big.hanyang.ac.kr/alphaArrestin_Fly link does not work. Please fix the link.

We appreciate your comment. In response to the reviewer’s comment, we have made comprehensive α-arrestin interactome maps on our new website (big.hanyang.ac.kr/alphaArrestin_PPIN) and confirmed that users can be re-directed to networks housed in NDEx.

Author response image 4.

Screen shot of the first page of the newly developed website.

Website address: big.hanyang.ac.kr/‌‌‌‌‌‍‍‍‌‌alphaArrestin_PPIN

Author response image 5.

Screen shot of the gene-gene network involving α-arrestin in human.

Reviewer #3 (Public Review):

Lee, Kyungtae and colleagues have discovered and mapped out alpha-arrestin interactomes in both human and Drosophila through the affinity purification/mass spectrometry and the SAINTexpress method. Their work revealed highly confident interactomes, consisting of 390 protein-protein interactions (PPIs) between six human alpha-arrestins and 307 preproteins, as well as 740 PPIs between twelve Drosophila alpha-arrestins and 467 prey proteins.

To define and characterize these identified alpha-arrestin interactomes, the team employed a variety of widely recognized bioinformatics tools. These analyses included protein domain enrichment analysis, PANTHER for protein class enrichment, DAVID for subcellular localization analysis, COMPLEAT for the identification of functional complexes, and DIOPT to identify evolutionary conserved interactomes. Through these assessments, they not only confirmed the roles and associated functions of known alpha-arrestin interactors, such as ubiquitin ligase and protease, but also unearthed unexpected biological functions in the newly discovered interactomes. These included involvement in RNA splicing and helicase, GTPase-activating proteins, and ATP synthase.

The authors carried out further study into the role of human TXNIP in transcription and epigenetic regulation, as well as the role of ARRDC5 in osteoclast differentiation. It is particularly commendable that the authors conducted comprehensive testing of TXNIP's role in HDAC2 in gene expression and provided a compelling model while revising the manuscript. Additionally, the quantification of the immunocytochemistry data presented in Figure 6 convincingly supports the authors' hypothesis.

Overall, this study holds important value, as the newly identified alpha-arrestin interactomes are likely aiding functional studies of this protein group and advance alpha-arrestin research.

eLife assessment

This study provides a valuable resource that documents the protein-protein interactions (PPI) network for alpha-arrestins in both human and Drosophila based on affinity purification/mass spectrometry and the SAINTexpress method followed by a series of bioinformatic and functional assessments. Through these, the authors confirmed the roles of known and novel interactions, including proteins involved in RNA splicing and helicase, GTPase-activating proteins, and ATP synthase. This study represents a convincing example of how to adopt comparative molecular interactions and how to interpret the functional implications.

Reviewer #1 (Public Review):

The study provides a complete comparative interactome analysis of α-arrestin in both humans and drosophila. The authors have presented interactomes of six humans and twelve Drosophila α-arrestins using affinity purification/mass spectrometry (AP/MS). The constructed interactomes helped to find α-arrestins binding partners through common protein motifs. The authors have used bioinformatic tools and experimental data in human cells to identify the roles of TXNIP and ARRDC5: TXNIP-HADC2 interaction and ARRDC5-V-type ATPase interaction. The study reveals the PPI network for α-arrestins and examines the functions of α-arrestins in both humans and Drosophila. The authors have carried out the necessary changes that were suggested.

I would like to congratulate the authors and the corresponding authors of this manuscript for bringing together such an elaborate study on α-arrestin and conducting a comparative study in drosophila and humans.

Reviewer #2 (Public Review):

In this manuscript, the authors present a novel interactome focused on human and fly alpha-arrestin family proteins and demonstrate its application in understanding the functions of these proteins. Initially, the authors employed AP/MS analysis, a popular method for mapping protein-protein interactions (PPIs) by isolating protein complexes. Through rigorous statistical and manual quality control procedures, they established two robust interactomes, consisting of 6 baits and 307 prey proteins for humans, and 12 baits and 467 prey proteins for flies. To gain insights into the gene function, the authors investigated the interactors of alpha-arrestin proteins through various functional analyses, such as gene set enrichment. Furthermore, by comparing the interactors between humans and flies, the authors described both conserved and species-specific functions of the alpha-arrestin proteins. To validate their findings, the authors performed several experimental validations for TXNIP and ARRDC5 using ATAC-seq, siRNA knockdown, and tissue staining assays. The experimental results strongly support the predicted functions of the alpha-arrestin proteins and underscore their importance.

Author Response

eLife assessment

This study presents valuable insights into the epigenetic landscape in adult kidney podocytes. A series of solid experiments demonstrate that genes that are regulated by a key kidney transcription factor, Mafb, are essential for H3K4me3 methylation and recruitment of Wt1 to Nphs1 and Nphs2. This new information provides insights into the potential relationship and coordination of transcription factors in regulating target genes in podocytes in glomerular diseases, although the conclusion that MafB is generally required for Wt1 to bind to podocyte-specific promoters is incomplete and should be extended beyond two or three genes.

We thank the reviewers and editors for critically reading our manuscript and their insightful comments. We will strive to revise

Reviewer #1 (Public Review):

Summary:

In their manuscript, Massa and colleagues provide a map of the epigenetic landscape in podocytes and analyze the role of the transcription factor MafB in podocyte gene expression. They initially map the histone profile in adult podocytes of the mouse by assaying three different histone methylation marks, namely H3K4me3, H3K4me1, and H3K27me3 for active, primed, and repressed states. They then perform Wt1- and MafB-ChIP-Seq analysis to identify respective direct targets of those transcription factors. Subsequently, they employ an inducible MafB knockout model and show that homozygous knockout mice show proteinuria and FSGS, suggesting an important role for MafB in podocyte homeostasis. RNA-Seq analysis in mice two daysafter tamoxifen application identified direct and indirect MafB target genes. Finally, the authors turn to a constitutive MafB knockout model, carry out anti-H3K4me3 and anti-Wt1 ChIP experiments, and examine selected promoters. One main conclusion from this work is that MafB opens chromatin and thus facilitates the binding of other transcription factors like Wt1 to podocyte-specific genes.

Strengths and weaknesses:

The authors have performed an impressive number of experiments and generated very valuable data. They use state-of the-art technology and the data are presented well and are sound. This being said the manuscript contains significant novel data, but also experiments that are already available in some sort. The histone profile in adult mouse podocytes is novel and provides an interesting map of epigenetic marks in this particular cell type. It is maybe not too surprising that podocyte-differentiation genes have different chromatin accessibility than genes associated with general development. The Wt1-ChIP has been done before by several labs but is certainly an important control in this work. The MafB-ChIP is new. The inducible MafB knockout model including the identification of Tcf21 as a target gene has been published by others in 2020 (and is acknowledged by the authors). The experiments addressing the potential role of MafB in chromatin opening are new. I find that the data are certainly compatible with the model put forward by the authors, but they are not compelling.

We agree that additional data on changes in chromatin accessibility in the absence of Mafb would help to support our model and we will be working towards this data for a revised version of the manuscript.

Reviewer #2 (Public Review):

Summary:

The authors investigate the role of MafB in regulating podocyte genes. Mafb is required for podocyte differentiation and maintenance. Mutations of this gene cause FSGS in mice and humans. They profiled MafB binding genome-wide in isolated glomeruli and defined overlap with Wt1. They provide evidence that Mafb is required for Wt1 binding and H3K4me3 methylation at the promoters of two essential podocyte genes, Nphs1 and Nphs2 Understanding how the action of different transcription factors is coordinated to control gene expression - the main goal of this paper - is an important line of investigation.

While the main conclusion of the paper is supported by their data, the scope is limited. Additional ChIP-seq experiments and data analysis are needed to solidify and extend their conclusions.

Strengths:

1) Performing ChIP-seq for histone modifications on isolated podocytes provides valuable cell-type-specific information. Similarly, profiling Mafb and Wt1 in isolated glomeruli provides podocyte-specific binding patterns because these transcription factors (TFs) are not expressed in other cell types in glomeruli. The significant overlap of their Wt1 binding genome-wide withthat of prior published work is reassuring. RNA-seq on isolated podocytes provides the appropriate cell-type specific gene expression data to integrate with ChIP-seq data. Together, the RNA-seq and ChIP-seq data are valuable resources for other investigators examining gene regulation in mouse podocytes.

2) The phenotype analysis of their FSGS model is convincing and well done.

3) Testing how Wt1 binding is affected by loss of Mafb provides insight into how these key podocyte TFs may cooperate to regulate genes.

Weaknesses:

1) The conclusion that Mafb is required for Wt1 binding and H3K4me3 methylation is based solely on ChIP-PCR at two gene promoters (Nphs1, Nphs2). This result should be validated and extended by ChIP-seq. Mafb and Wt1 binding overlap at more than 200 sites. If their model is correct, it is likely that Wt1 binding would be affected at other genomic sites. This result would add strong support to their model of how Wt1 and Mafb cooperate to regulate genes in podocytes. Moreover, ChIP-seq would define whether the dependence of Wt1 on Mafb is also evident at distal regulatory regions (defined H3K4me1, which is typically found at predicted enhancers).

We agree that a genome wide analysis of chromatin accessibility would help corroborating our model and will work towards this data for a revised version.

2) The FSGS model generated by the authors involved conditional deletion of Mafb in podocytes at 8 weeks of age. They found that this resulted in reduced expression of Nphs1 and Nphs2 within 48 hours post-deletion. However, they investigated Wt1 binding and H3K4me3 genomic binding in Mafb homozygous null embryos. While this result provides information about podocyte differentiation, it does not address the maintenance of expression of these essential podocyte genes in the adult kidney. Because post-natal deletion of Mafb led to FSGS and reduced expression of Nphs1/2, ChIP-seq should be performed on the adult conditional mutants in order to provide mechanistic information about the disease.

The fact that the phenotype in Mafb conditional mutant animals is progressive means that epigenetic changes are also likely to be quantitative. Indeed, Nphs1/Nphs2 are still expressed 6 weeks after Mafb deletion, albeit at lower levels. Since ChIP-seq experiments are not necessarily quantitative, we believe it may be difficult to detect statistically significant changes in this model. We will discuss this limitation of our study in a revised version of our manuscript.

3) H3K4me1 binds enhancer regions. The authors performed ChIP-seq to profile H3K4me1 in isolated podocytes. However, there was no analysis reported of these results. It would be valuable to determine if Wt1 and Mafb co-localize at predicted enhancers in podocytes and if Wt1 binding is lost at these regions in Mafb mutant glomeruli.

We well reanalyse the data taking the reviewer’s comments into account.

eLife assessment

This study presents valuable insights into the epigenetic landscape in adult kidney podocytes. A series of solid experiments demonstrate that genes that are regulated by a key kidney transcription factor, Mafb, are essential for H3K4me3 methylation and recruitment of Wt1 to Nphs1 and Nphs2. This new information provides insights into the potential relationship and coordination of transcription factors in regulating target genes in podocytes in glomerular diseases, although the conclusion that MafB is generally required for Wt1 to bind to podocyte-specific promoters is incomplete and should be extended beyond two or three genes.

Reviewer #1 (Public Review):

Summary:<br /> In their manuscript, Massa and colleagues provide a map of the epigenetic landscape in podocytes and analyze the role of the transcription factor MafB in podocyte gene expression. They initially map the histone profile in adult podocytes of the mouse by assaying three different histone methylation marks, namely H3K4me3, H3K4me1, and H3K27me3 for active, primed, and repressed states. They then perform Wt1- and MafB-ChIP-Seq analysis to identify respective direct targets of those transcription factors. Subsequently, they employ an inducible MafB knockout model and show that homozygous knockout mice show proteinuria and FSGS, suggesting an important role for MafB in podocyte homeostasis. RNA-Seq analysis in mice two days after tamoxifen application identified direct and indirect MafB target genes. Finally, the authors turn to a constitutive MafB knockout model, carry out anti-H3K4me3 and anti-Wt1 ChIP experiments, and examine selected promoters. One main conclusion from this work is that MafB opens chromatin and thus facilitates the binding of other transcription factors like Wt1 to podocyte-specific genes.

Strengths and weaknesses:<br /> The authors have performed an impressive number of experiments and generated very valuable data. They use state-of the-art technology and the data are presented well and are sound. This being said the manuscript contains significant novel data, but also experiments that are already available in some sort. The histone profile in adult mouse podocytes is novel and provides an interesting map of epigenetic marks in this particular cell type. It is maybe not too surprising that podocyte-differentiation genes have different chromatin accessibility than genes associated with general development. The Wt1-ChIP has been done before by several labs but is certainly an important control in this work. The MafB-ChIP is new. The inducible MafB knockout model including the identification of Tcf21 as a target gene has been published by others in 2020 (and is acknowledged by the authors). The experiments addressing the potential role of MafB in chromatin opening are new. I find that the data are certainly compatible with the model put forward by the authors, but they are not compelling.

Reviewer #2 (Public Review):

Summary:<br /> The authors investigate the role of MafB in regulating podocyte genes. Mafb is required for podocyte differentiation and maintenance. Mutations of this gene cause FSGS in mice and humans. They profiled MafB binding genome-wide in isolated glomeruli and defined overlap with Wt1. They provide evidence that Mafb is required for Wt1 binding and H3K4me3 methylation at the promoters of two essential podocyte genes, Nphs1 and Nphs2. Understanding how the action of different transcription factors is coordinated to control gene expression - the main goal of this paper - is an important line of investigation.

While the main conclusion of the paper is supported by their data, the scope is limited. Additional ChIP-seq experiments and data analysis are needed to solidify and extend their conclusions.

Strengths:<br /> 1) Performing ChIP-seq for histone modifications on isolated podocytes provides valuable cell-type-specific information. Similarly, profiling Mafb and Wt1 in isolated glomeruli provides podocyte-specific binding patterns because these transcription factors (TFs) are not expressed in other cell types in glomeruli. The significant overlap of their Wt1 binding genome-wide with that of prior published work is reassuring. RNA-seq on isolated podocytes provides the appropriate cell-type specific gene expression data to integrate with ChIP-seq data. Together, the RNA-seq and ChIP-seq data are valuable resources for other investigators examining gene regulation in mouse podocytes.

2) The phenotype analysis of their FSGS model is convincing and well done.

3) Testing how Wt1 binding is affected by loss of Mafb provides insight into how these key podocyte TFs may cooperate to regulate genes.

Weaknesses:<br /> 1) The conclusion that Mafb is required for Wt1 binding and H3K4me3 methylation is based solely on ChIP-PCR at two gene promoters (Nphs1, Nphs2). This result should be validated and extended by ChIP-seq. Mafb and Wt1 binding overlap at more than 200 sites. If their model is correct, it is likely that Wt1 binding would be affected at other genomic sites. This result would add strong support to their model of how Wt1 and Mafb cooperate to regulate genes in podocytes. Moreover, ChIP-seq would define whether the dependence of Wt1 on Mafb is also evident at distal regulatory regions (defined H3K4me1, which is typically found at predicted enhancers).

2) The FSGS model generated by the authors involved conditional deletion of Mafb in podocytes at 8 weeks of age. They found that this resulted in reduced expression of Nphs1 and Nphs2 within 48 hours post-deletion. However, they investigated Wt1 binding and H3K4me3 genomic binding in Mafb homozygous null embryos. While this result provides information about podocyte differentiation, it does not address the maintenance of expression of these essential podocyte genes in the adult kidney. Because post-natal deletion of Mafb led to FSGS and reduced expression of Nphs1/2, ChIP-seq should be performed on the adult conditional mutants in order to provide mechanistic information about the disease.

3) H3K4me1 binds enhancer regions. The authors performed ChIP-seq to profile H3K4me1 in isolated podocytes. However, there was no analysis reported of these results. It would be valuable to determine if Wt1 and Mafb co-localize at predicted enhancers in podocytes and if Wt1 binding is lost at these regions in Mafb mutant glomeruli.

Author Response

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

For the final Version of Record the following changes will be included: 1. Figure 4: Example traces replaced with a more representative simulation run that is more similar to the mean. 2. Methods: Description of the alignment procedure expanded to explain the algorithm steps better.

---

The following is the authors’ response to the previous reviews

We are grateful for the positive and insightful feedback from the editors and reviewers. These constructive comments have contributed to the enhancement of our work. We have revised the manuscript, addressing each of the comments raised. In addition, based on the commentary provided, we have introduced two new figures that offer a deeper understanding of our research findings:

In new Figure 7, we present the analysis of the difference in onset times between motion and flash responses. This figure also includes a simple illustration elucidating the origins of these differences, highlighting the varying engagement of receptive fields by these stimuli. The data presented in this figure were initially featured in the main text of the original manuscript. Figure 11 offers a detailed comparison of the temporal and spatial characteristics of the synthetic presynaptic signals driving optimal DS in SACs. We compare these characteristics with the properties extracted from recorded glutamate release. Our analysis suggests that the sluggish dynamics observed in biological signals impede effective directional integration. Below are the detailed point-by-point responses to reviewers comments.

Reviewer #1 (Public Review):

Summary:

Direction selectivity (DS) in the visual system is first observed in the radiating dendrites of starburst amacrine cells (SACs). Studies over the last two decades have aimed to understand the mechanisms that underlie these unique properties. Most recently, a 'space-time' model has garnered special attention. This model is based on two fundamental features of the circuit. First, distinct anatomical types of bipolar cells (BCs) are connected to proximal/distal regions of each of the SAC dendritic sectors (Kim et al., 2014). Second, that input across the length of the starburst is kinetically diverse, a hypothesis that has been only recently demonstrated experimentally using iGluSnFR imaging (Srivastava et al., 2022). However, the stark kinetic distinctions, i.e., the sustained/transient nature of BC input to SACs dendrites appear to be present mainly in responses to stationary stimuli. When BC receptive field properties are probed using white noise stimuli, the kinetic differences between BCs are relatively subtle or nonexistent (Gaynes et al., 2022; Strauss et al., 2022, Srivastava et al., 2022). Thus, if and how BCs contribute to direction selectivity driven by moving spots that are commonly used to probe the circuit remains to be clarified. To address this issue, Gaynes et al., combine evolutionary computational modeling (Ankri et al., 2020) with two-photon iGluSnFR imaging to address to what degree BCs contribute to the generation of direction selectivity in the starburst dendrites in response to stimuli that are commonly used experimentally.

Strengths:

Combining theoretical models and iGluSnFR imaging is a powerful approach as it first provides a basic intuition on what is required for the generation of robust DS, and then tests the extent to which the experimentally measured BC output meets these requirements.

The conclusion of this study builds on the previous literature and comprehensively considers the diverse BC receptive field properties that may contribute to DS (e.g. size, lag, rise time, decay time).

By 'evolving' bipolar inputs to produce robust DS in a model network, these authors provide a sound framework for understanding which kinetic properties could potentially be important for driving downstream DS. They suggest that response delay/decay kinetics, rather than the center/surround dynamics are likely to be most relevant (albeit the latter could generate asymmetric responses to radiating/looming stimuli).

Weaknesses:

Finally, these authors report that the experimentally measured BC responses are far from optimal for generating DS. Thus, the BC-based DS mechanism does not appear to explain the robust DS observed experimentally (even with mutual inhibition blocked). Nevertheless, I feel the comprehensive description of BC kinetics and the solid assessment of the extent to which they may shape DS in SAC dendrites, is a significant advancement in the field.

Reviewer #2 (Public Review):

Summary:

In this study, the authors sought to understand how the receptive fields of bipolar cells contribute to direction selectivity in starburst amacrine cell (SAC) dendrites, their post synaptic partners. In previous literature, this contribution is primarily conceptualized as the 'space-time wiring model', whereby bipolar cells with slow-release kinetics synapse onto proximal dendrites while bipolar cells with faster kinetics synapse more distally, leading to maximal summation of the slow proximal and fast distal depolarizations in response to motion away from the soma. The space-time wiring contribution to SAC direction selectivity has been extensively tested in previous literature using connectomic, functional, and modeling approaches. However, the authors argue that previous functional studies of bipolar cell kinetics have focused on static stimuli, which may not accurately represent the spatiotemporal properties of the bipolar cell receptive field in response to movement. Moreover, this group and others have recently shown that bipolar cell signal processing can change directionally when visual stimuli starts within the receptive field rather than passing through it, complicating the interpretation of moving stimuli that start within a bipolar cell of interest's receptive field (e.g. stimulating only one branch of a SAC or expanding/contracting rings). Thus, the authors choose to focus on modeling and functionally mapping bipolar cell kinetics in response to moving stimuli across the entire SAC dendritic field.

General Comments

There have been several studies that have addressed the contribution of space-time wiring to SAC process direction selectivity. The impact of this project is to show that this contribution is limited. First, the optimal solution obtained by the evolutionary algorithm to generate DS processes is slow proximal and fast distal inputs - exactly what is predicted by space-time wiring, which is exactly what is required of the HRC model. Hence, this result seems expected and it's not clear what the alternative hypothesis is. Second, the experimental results based on glutamate imaging to assess the kinetics of glutamate release under conditions of visual stimulation across a large region of retina confirm previous observations but were important to test. Third, by combining their model model with this experiment data, they conclude that even the optimal space-time wiring is not sufficient to explain the SAC process DS. The results of this approach might be more impactful if the authors come to some conclusion as to what factors do determine the direction selectivity of the SAC process since they have argued that all the current models are not sufficient.

Reviewer #3 (Public Review):

Gaynes et al. investigated the presynaptic and postsynaptic mechanisms of starburst amacrine cell (SAC) direction selectivity in the mouse retina by computational modeling and glutamate sensitivity (iGluSnFR) imaging methods. Using the SAC computational simulation, the authors initially tested bipolar cell contributions (space-time wiring model, presynaptic effect) and SAC axial resistance contributions (postsynaptic effect) to the SAC DS. Then, the authors conducted two-photon iGluSnFR imaging from SACs to examine the presynaptic glutamate release, and found seven clusters of ON-responding and six clusters of OFF-responding bipolar cells. They were categorized based on their response kinetics: delay, onset phase, decay time, and others. Finally, the authors generated a model consisting of multiple clusters of bipolar cells on proximal and distal SAC dendrites. When the SAC DS was measured using this model, they found that the space-time wiring model accounted for only a fraction of SAC DS.

The article has many interesting findings, and the data presentation is superb. Strengths and weaknesses are summarized below.

Major Strengths:

• The authors utilized solid technology to conduct computational modeling with Neuron software and a machine-learning approach based on evolutionary algorithms. Results are effectively and thoroughly presented.

• The space-time wiring model was evaluated by changing bipolar cell response properties in the proximal and distal SAC dendrites. Many response parameters in bipolar cells are compared, and DSI was compared in Figure 3.

• Two-photon microscopy was used to measure the bipolar cell glutamate outputs onto SACs by conducting iGluSnFR imaging. All the data sets, including images and transients, are elegantly presented. The authors analyzed the response based on various parameters, which generated more than several response clusters. The clustering is convincing.

Major Weaknesses:

• In Figure 9, the authors generated the bipolar cell cluster alignment based on the space-time wiring model. The space-time wiring model has been proposed based on the EM study that distinct types of bipolar cells synapse on distinct parts of SAC dendrites (Green et al 2016, Kim et al 2014). While this is one of the representative Reicardt models, it is not fully agreed upon in the field (see Stincic et al 2016). While the authors' approach of testing the space-time wiring model and conclusions is interesting and appreciated, the authors could address more issues: mainly two clusters were used to generate the model, but more numbers of clusters should be applied. Although the location of each cluster on the SAC dendrites is unknown, the authors should know the populations of clusters by iGluSnFR experiments. Furthermore, the authors could provide more suggestive mechanisms after declining postsynaptic factors and the space-time wiring model.

The reviewer is correct that the proximal and more distal SAC dendrites sample from different IPL depths. It should be theoretically possible to match the functional clusters we measured with anatomical bipolar cell identities. However, the stratifications of these cells have significant overlaps (Figure 6-S2), and previous attempts to match iGluSnFR signals to anatomy proved to be challenging (Franke et al., 2017; Gaynes et al., 2022; Matsumoto et al., 2019; Srivastava et al., 2022; Strauss et al., 2022). In the revised version of the manuscript, we reorder the functional clusters based on their transiency, which has a higher correlation to stratification depth (Franke et al., 2017).

We have examined a scenario in which the presynaptic population comprises more than two clusters. We constructed synthetic models whose input structure was as in Figure 10 (old Figure 9). The optimal configuration for the most proximal and distal inputs closely resembled the proximal-distal model reported in Figure 2. However, we observed a nearly linear variation in the shape of the optimal mid-range inputs, transitioning from proximal-like to distal-like responses as the distance increased. We consider this outcome to be expected based on the structure of the space-time wiring model (Kim et al., 2014). Interestingly, this was not the case with models incorporating physiologically recorded signals. As we show in Figure 10, the most common optimal directional tuning was seen when the bipolar drive consisted of two main populations, both in the ON and OFF SACs.

Finally, we believe that uncovering additional mechanisms that underlie directional selectivity in SACs represents a crucial challenge for the field to tackle. It is highly probable that achieving directional selectivity involves a complex interplay of multiple factors. This includes the organization of the presynaptic circuit, which we have partially addressed in this study, as well as the influence of postsynaptic active conductances and feedback loops involving other SACs and presynaptic cells. We have expanded the discussion section to describe the possible mechanisms

• The computational modeling demonstrates intriguing results: SAC dendritic morphology produces dendritic isolation, and a massive input overcomes the dendritic isolation (Figure 1). This modeling seems to be generated by basic dendritic cable properties. However, it has been reported that SAC dendrites express Kv3 and voltage-gated Ca channels. It seems to be that these channels are not incorporated in this model.

The reviewer's observation is accurate; the model depicted in Figure 1 did not include voltage-gated channels. Our goal was to study electrotonic isolation, which is often measured in passive models. However, while we did not incorporate voltage-gated potassium channels implicitly in the models, our simulations are rooted in previous models that were fine-tuned using empirical data. As potassium channels are expected to influence the experimentally recorded input resistance, we have indirectly accounted for their impact on the interdendritic signal propagation.

In subsequent model iterations, we have integrated voltage-gated calcium channels into our simulations to assess the signal responsible for driving synaptic release. We show that nonlinear voltage dependence of the calcium currents enhances compartmentalization of the local calcium levels (Figure 2), but did not significantly influence local voltages. Therefore, calcium channels do not appear to have a major impact on electrotonic distances.

• In Figure 5B, representative traces are shown responding to moving bars in horizontal directions. These did not show different responses to two directional stimuli. It is unclear whether directional preference was not detected, which was shown by Yonehara's group recently (Matsumoto et al 2021). Or that was not investigated as described in the Discussion.

Indeed, we observed no discernible directional differences in bipolar responses. This phenomenon can be primarily attributed to the fact that the signals originating from the limited number of directionally-tuned release sites are overshadowed by the release from non-directionally-tuned units (Matsumoto et al., 2021). In the revised discussion, we have acknowledged this limitation in our recorded data.

• The authors found seven ON clusters and six OFF clusters, which are supposed to be bipolar cell terminals. However, bipolar cells reported to provide synaptic inputs are T-7, T-6, and multiple T-5s for ON SACs and T-1, T-2, and T-3s for OFF SACs. The number of types is less than the number of clusters. Potentially, clusters might belong to glutamatergic amacrine cells. These points are not fully discussed.

We have expanded the discussion section to address these points.

Reviewer #1 (Recommendations For The Authors):

Major comments

1. One of the main conclusions of this study is that diverse BC kinetics contribute to DS (Fig. 9). The authors nicely demonstrate using modeling that the experimentally measured BC kinetics are far from ideal. However, this conclusion is based on a model that almost exclusively relies on just two of the 7 putative BC types (e.g., C1 & C6 for On SACs) placed optimally along the dendrites, which raises two important caveats.

First, given that other BC types are likely to contribute, the effects of two distinct types are likely to be diluted. Thus, the contribution of BCs to DS is likely to be significantly overestimated. Second, given that the dendrites of 10-30 SACs cross each point in the honeycomb, for the given model to work, each BC would need to connect extremely selectively to SACs. i.e., at a given point, a sustained input must only connect to the more proximal dendritic segments, while avoiding entirely the distal segments of overlapping SAC dendrites. Thus, their model requires extremely selective wiring for which there is no evidence. In fact, there is evidence to the contrary provided by Ding et al. 2016, which showed that the type 7 (proximally biased) and type 5 (distally biased) populations had a substantial overlap (assuming these BC types correspond to kinetically diverse clusters).

We wholeheartedly concur with the reviewer's perspective that our findings have led to an overestimation of the space-time wiring mechanism's role in SAC directional selectivity (DS). We have adjusted our discussion to emphasize this point. In light of this, our assertion that, even with the most favorable distribution of synaptic inputs, the space-time wiring model still does not fully account for the experimentally-determined directional tuning in SAC, remains valid.

With regard to the model, it would also be worth comparing results to previous starburst models (e.g., Tukker et al,. 2004), which demonstrated a robust DS in SAC dendrites in the absence of kinetically diverse BC input. Why is the cell-intrinsic DS so weak in the present model?

We have directly explored this question in the synthetic model (Figures 2, 3). Despite variances in the anatomy of SACs and the distribution of bipolar inputs between our model and the study by (Tukker et al., 2004), we observed remarkably similar levels of directional selectivity index computed from the voltage response (approximately 10%, as shown in Figure 3, 'Identical BCs').

The primary distinction emerged in the degree of DS amplification mediated by calcium currents. Tukker et al., 2004 reported considerably higher DS compared to our findings, despite employing similar formulations for voltage-gated calcium channel models. The key factor driving this difference lies in the fact that Tukker et al., 2004 measured amplification in proximity to the threshold of calcium channel activation. Even minor variations in membrane potentials near this threshold can lead to substantial differences in calcium influx, especially when outward stimulation results in a calcium spike. In fact, recently, Robert Smith’s group revisited the threshold-based mechanism and concluded that it often fails to produce robust DS due to the heterogeneity of membrane potentials among different terminal dendrites (Wu et al., 2023).

Our models were trained on five different stimuli velocities whose synaptic integration produced substantially different peak amplitudes. Consequently, the spike threshold alone couldn't reliably distinguish between inward and outward directions across all five conditions, resulting in reduced directional performance in our simulations. In the revised Figure 2-S2 we directly explore the performance of the model with identical BC formulations, trained on a single velocity. We find a dramatic enhancement of calcium DS (DSI=66%) in this condition compared to an identical model trained on 5 velocities (DSI=17%). Thus, evolutionary search is capable of finding the threshold-based solution, but only when the training is performed on a single stimulus velocity (Figure 2-S2). This solution did not generalize to multiple stimuli speeds because, as mentioned above, they lead to different postsynaptic depolarization levels (Figure 2, 2-S1). Instead, the algorithm converged on a set of postsynaptic paraments leading to less nonlinear calcium channel activation over a broader voltage range, ensuring effective DS performance over multiple velocities and heterogenous local potentials (Wu et al., 2023).

1. Functionally distinct responses across different regions of interest (ROIs) were used to classify BC input. ROIs were obtained from multiple scan fields and retinas and combined into a single dataset for functional clustering. However, the consistency of the cluster distribution across these replicates has not been addressed. As BCs can exhibit different functional properties dependant on the state/health of the retina, it is important to know whether certain functional clusters may originate disproportionately from a particular experiment, as it implies that each cluster does not represent a different stable functional/anatomical population.

We acknowledge that the state of the preparation can significantly impact signal dynamics. In response to this important consideration, we have incorporated details about the distribution of functional clusters in various experiments in the revised version of the manuscript (Figure 6-S1, and discussion).

Other comments:

1. Interpreting iGluSnFR signals: Since the sensor is expressed uniformly across the SAC dendrite, it is important to clarify why the measured F signals are considered synaptic responses. Could spillover contribute to the generation of slower responses?

We do not believe spillover can explain slower responses because the sluggish clusters often responded significantly (up to 500ms) sooner to moving bars (Figures 6, 6-S3). We acknowledge and discuss this possibility of spillover in the revised discussion.

1. One striking finding is the diversity of BCs RF sizes (Fig. 7C). Some BCs have RF that are far larger than their dendritic fields. It will be useful to discuss the potential mechanisms that may underlie large BC RFs.

We changed the discussion to address this question.

1. SAC DS is independent of dendritic isolation: The authors claim that dendritic isolation does not significantly impact DS. However, while this might be true for a linear motion through the receptive field, dendritic isolation probably matters for more dynamic stimuli. For example, DSGCs can encode rapid changes in objection direction, as DS is computed over fine spatiotemporal scales relying on SACs (Murphy-Baum et al., 2022). This could not occur if SAC dendrites were not well electrically isolated from each other.

We believe that this is an accurate interpretation of our findings. Our research suggests that dendritic isolation is likely not a critical factor in the space-time wiring mechanism. However, as we demonstrate that this particular mechanism cannot fully account for the observed levels of DS in SACs, other mechanisms must be important. As previous studies revealed that dendritic isolation enhances SAC DS (for example, Koren et al., 2017), dendritic independence likely contributes to directional performance within SACs by these additional mechanisms.

1. Figure 4: From what I understand, the BC inputs for the electrotonic connectivity variations evolved much like they were for the original model without axial resistance constraints. This makes sense, since stronger/weaker inputs with different temporal kernels may be appropriate for each condition, hence why the axial resistance wasn't changed post-evolution, which would have likely caused the DS to drop. If that is the case, however, I wonder how the best DS attainable by the final model which is constrained to the radial arrangement of realistic BC inputs (without being able to fit much more optimal sustained-transient BCs to their circumstance) would be impacted. Is dendritic isolation similarly unimportant when the pre-synaptic story isn't ideal?

We have explored this question directly by allowing the evolutionary algorithm to modify the passive and active characteristics of the postsynaptic SAC. Our findings are summarized in Figure 9-S1. We observed a correlation between DSI levels and membrane/axial resistance values in SACs in the evolved models. Better DS was seen with leaky membranes (higher isolation) and lower axial resistance (lower isolation). While it is clear that postsynaptic parameters can influence synaptic integration, they can not fully compensate for inadequate presynaptic dynamics.

1. BC are shown to contribute to DS across velocities (Fig. 9), which contrasts with results from Srivastava et al., (2022) that showed BCs contribute to DS at lower velocities. However, this discrepancy can easily be explained by the choice of moving spots. In this study, the sweeping bars had dynamic width (targeting pixel dwell time of 2s), which means for higher velocities the bar is significantly wider. While in the previous study, the width of the stimulus was kept constant, and thus for higher velocities, the sustained/transient kinetic differences of BCs are less clear (Srivastava et al., 2021). The author's should discuss this explicitly, to avoid discrepancies between these two studies the reader might otherwise perceive.

We value reveiwer’s feedback, and in response, we have included an additional paragraph in the manuscript addressing the distinctions in directional tuning that arise from the space-time model presented in this work, in comparison to earlier studies.

1. Methods: It will be good to discuss how ROIs sizes and positions were selected (pixel correlations?)

We have included a more detailed explanation of the clustering procedure

• Lines 614 describe whole-cell patch clamp techniques, which are not used in this study.

We used patch-clamp to record the waveforms shown in Figure 2-S2

1. Figure 6: Diversity of Glut responses to motion in ON and OFF SACs, caption typos?

2. "Left:" without "Right:" to describe the population (I presume) viewed as an image

3. If there should still be A,C and B,D to group the ON and OFF halves, maybe it should be mentioned in the caption

Thank you for bringing this to our attention, the legends were fixed.

References:

Kim, J. S., Greene, M. J., Zlateski, A., Lee, K., Richardson, M., Turaga, S. C., Purcaro, M., Balkam, M., Robinson, A., Behabadi, B. F., Campos, M., Denk, W., Seung, H. S., & EyeWirers (2014). Space-time wiring specificity supports direction selectivity in the retina. Nature, 509(7500), 331-336. https://doi.org/10.1038/nature13240

Gaynes, J. A., Budoff, S. A., Grybko, M. J., Hunt, J. B., & Poleg-Polsky, A. (2022). Classical center-surround receptive fields facilitate novel object detection in retinal bipolar cells. Nature communications, 13(1), 5575. https://doi.org/10.1038/s41467-022-32761-8

Murphy-Baum B. and Awatramani GB (2022). Parallel processing in active dendrites during periods of intense spiking activity, Cell Reports, Volume 38, Issue 8,

Srivastava P, de Rosenroll G., MatsumotoA., Michaels T., Turple Z., Jain V, Sethuramanujam S, Murphy-Baum B, Yonehara K., Awatramani, G.B. (2022) Spatiotemporal properties of glutamate input support direction selectivity in the dendrites of retinal starburst amacrine cells eLife 11:e81533

Strauss, S., Korympidou, M. M., Ran, Y., Franke, K., Schubert, T., Baden, T., Berens, P., Euler, T., & Vlasits, A. L. (2022). Center-surround interactions underlie bipolar cell motion sensitivity in the mouse retina. Nature communications, 13(1), 5574. https://doi.org/10.1038/s41467-022-32762-7

Tukker, J. J., Taylor, W. R., & Smith, R. G. (2004). Direction selectivity in a model of the starburst amacrine cell. Visual neuroscience, 21(4), 611-625. https://doi.org/10.1017/S0952523804214109

Reviewer #2 (Recommendations For The Authors):

Specific comments

1. Line 223. The statement a model trained on only optimal DSI would produce "negligible absolute differences in calcium levels." is unclear. This needs to be better explained.

We have modified and expanded this paragraph to make it more clear

1. Figure 4. The authors use this model to test the hypothesis that space time wiring contribution to SAC process DS requires dendritic isolation. They do this by increasing axial resistance around the soma of their model neuron to isolate each dendrite. They found comparable DS was achieved in both conditions, indicating that the space-time wiring model works in two cases of high and low dendritic isolation. However, to test the claim that "specific details of postsynaptic integration appear to play a lesser role" (line 274) the authors may consider allowing the axial resistance to change as a part of the model rather than testing two extreme states.

Membrane and axial resistances (and active parameters) were allowed to change as part of model evolution in most simulations presented in this manuscript. We have added the information on the final resistance values reached in the evolved models in Figure 9-S1

1. Figure 6: To study glutamatergic input onto SACs, the authors expressed iGLuSnFR in ChAT-Cre mice and grouped similarly responding pixels into ROIs and separated these responses into functional groups based on cluster analysis (Figure 5). The alignment of the responses in Figure 6A was confusing. It appears that average responses for each cluster are aligned based on the peak observed during the stimulus in each direction, but it is unclear how they are aligned relative to each other or what this timing is relative to location of the stimulus (i.e. what is time 0 in 6A?).

The displayed traces represent the average responses to horizontally moving bars (speed = 0.5mm/s), either moving to the left or right. To achieve this alignment, we employed a procedure consistent with our recent publication (Gaynes et al., 2022), which we have now detailed more comprehensively. Here's the step-by-step process we followed:

1. Determination of half-maximum rise times: Initially, we calculated the half-maximum rise times for glutamate signals recorded in response to left and right-moving stimuli.

2. Calculation of mean rise time: We then computed the mean of these rise times, which served as a reference point for alignment.

3. Alignment procedure: To illustrate the alignment process, consider an example. Suppose the 50% rise time for responses to left-moving stimuli occurs at 3 seconds, while responses to right-moving stimuli occur 4 seconds after stimulation onset. This discrepancy suggests that the RF of the cell is shifted to the right from the center of the display (assuming a stimulation speed of 0.5mm/s on the retina, the RF's position would be approximately 250μm from the midline). To align these responses, we shifted both waveforms by 500ms so that their 50% rise times coincided at 3.5 seconds. Importantly, 3.5 seconds would represent the 50% rise time of the ROI if it were precisely centered on the display. This alignment effectively removed any spatial position dependence from the ROIs.

4. Comparative analysis and clustering: With the responses now aligned, we were able to compare their shapes and subsequently cluster the ROIs into distinct functional clusters. For clarity, we opted to highlight the time of response peak for cluster 1. Although this peak closely aligned with the calculated time of stimulus motion over the center of the 'shifted RF' in the adjusted time frame, it provided a more straightforward comparison between response dynamics.

1. The authors need to do a better job explaining how their results differ from Ezra-Tsur et al 2021, which uses the same sort of model to address the same question. The discussion about this study (lines 425-435) are based on how a more constrained version of these models work better but they do not directly address the difference in conclusion with regards to mechanisms that contribute to SAC process direction selectivity.

We have expanded the discussion related to mechanisms that contribute to DS in SACs and discuss the differences between our studies.

Minor point: The authors use the word "probe" to refer to visual stimulus. This is confusing because "probe" is also used to refer to sensors.

In the revised manuscript, we minimized the usage of ‘probe’ to reference visual stimuli

Reviewer #3 (Recommendations For The Authors):

Writing and figure presentations are excellent.

Thank you!

References:

Franke, K., Berens, P., Schubert, T., Bethge, M., Euler, T., & Baden, T. (2017). Inhibition decorrelates visual feature representations in the inner retina. Nature, 542(7642), 439-444. https://doi.org/10.1038/nature21394

Gaynes, J. A., Budoff, S. A., Grybko, M. J., Hunt, J. B., & Poleg-Polsky, A. (2022). Classical Center-Surround Receptive Fields Facilitate Novel Object Detection in Retinal Bipolar Cells. Nat Commun, 13(1), 5575. https://doi.org/https://doi.org/10.1038/s41467-022-32761-8

Kim, J. S., Greene, M. J., Zlateski, A., Lee, K., Richardson, M., Turaga, S. C., Purcaro, M., Balkam, M., Robinson, A., Behabadi, B. F., Campos, M., Denk, W., Seung, H. S., & EyeWirers. (2014). Space-time wiring specificity supports direction selectivity in the retina. Nature, 509(7500), 331-336. https://doi.org/10.1038/nature13240

Matsumoto, A., Agbariah, W., Nolte, S. S., Andrawos, R., Levi, H., Sabbah, S., & Yonehara, K. (2021). Direction selectivity in retinal bipolar cell axon terminals. Neuron. https://doi.org/10.1016/j.neuron.2021.07.008

Matsumoto, A., Briggman, K. L., & Yonehara, K. (2019). Spatiotemporally Asymmetric Excitation Supports Mammalian Retinal Motion Sensitivity. Curr Biol. https://doi.org/10.1016/j.cub.2019.08.048

Srivastava, P., de Rosenroll, G., Matsumoto, A., Michaels, T., Turple, Z., Jain, V., Sethuramanujam, S., Murphy-Baum, B. L., Yonehara, K., & Awatramani, G. B. (2022). Spatiotemporal properties of glutamate input support direction selectivity in the dendrites of retinal starburst amacrine cells. Elife, 11. https://doi.org/10.7554/eLife.81533

Strauss, S., Korympidou, M. M., Ran, Y., Franke, K., Schubert, T., Baden, T., Berens, P., Euler, T., & Vlasits, A. L. (2022). Center-surround interactions underlie bipolar cell motion sensing in the mouse retina. Nat Commun, 13(1), 5574. https://doi.org/https://doi.org/10.1038/s41467-022-32762-7

Tukker, J. J., Taylor, W. R., & Smith, R. G. (2004). Direction selectivity in a model of the starburst amacrine cell. Vis Neurosci, 21(4), 611-625. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=15579224

Wu, J., Kim, Y. J., Dacey, D. M., Troy, J. B., & Smith, R. G. (2023). Two mechanisms for direction selectivity in a model of the primate starburst amacrine cell. Vis Neurosci, 40, E003. https://doi.org/10.1017/S0952523823000019

eLife assessment

This important study uses a combination of computational modeling and glutamate imaging to show how a particular synaptic organization referred to as space-time wiring contributes minimally to a dendritic computation that occurs in the retina. The evidence supporting the claims of the authors is compelling, incorporating new findings regarding dynamic receptive field properties, an improvement over previous modeling and experimental results based on static visual stimuli. The work will be of interest to retinal neurobiologists and neurophysiologists interested in dendritic computations.

Reviewer #1 (Public Review):

Summary:

Direction selectivity (DS) in the visual system is first observed in the radiating dendrites of starburst amacrine cells (SACs). Studies over the last two decades have aimed to understand the mechanisms that underlie these unique properties. Most recently, a 'space-time' model has garnered special attention. This model is based on two fundamental features of the circuit. First, distinct anatomical types of bipolar cells (BCs) are connected to proximal/distal regions of each of the SAC dendritic sectors (Kim et al., 2014). Second, that input across the length of the starburst is kinetically diverse, a hypothesis that has only recently gained some experimental support using iGluSnFR imaging (Srivastava et al., 2022). However, in these prior studies, the sustained/transient distinctions in BC input that are proposed to underlie direction selectivity were shown to be present mainly in responses to stationary stimuli. When BC receptive field properties are probed using white noise stimuli, the kinetic differences between proximal/distal BC input are relatively subtle or nonexistent (Gaynes et al., 2022; Strauss et al., 2022, Srivastava et al., 2022). Thus, if and how BCs contribute to direction selectivity driven by moving spots that are commonly used to probe the circuit remains to be clarified. To address this issue, Gaynes et al., combine evolutionary computational modeling (Ankri et al., 2020) with two-photon iGluSnFR imaging to address to what degree BCs contribute to the generation of direction selectivity in the starburst dendrites.

Strengths:

Combining theoretical models and iGluSnFR imaging is a powerful approach as it first provides a basic intuition on what is required for the generation of robust DS, and then tests the extent to which the experimentally measured BC output meets these requirements.

The conclusion of this study builds on the previous literature and comprehensively considers the diverse BC receptive field properties that may contribute to DS (e.g. size, lag, rise time, decay time).

By 'evolving' bipolar inputs to produce robust DS in a model network, these authors provide a sound framework for understanding which kinetic properties could potentially be important for driving downstream DS. They suggest that response delay/decay kinetics, rather than the center/surround dynamics are likely to be most relevant (albeit the latter could generate asymmetric responses to radiating/looming stimuli).

Weaknesses:

Finally, these authors report that the experimentally measured BC responses are far from optimal for generating DS. Thus, the BC-based DS mechanism does not appear to explain the robust DS observed experimentally (even with mutual inhibition blocked). Nevertheless, I feel the comprehensive description of BC kinetics and the solid assessment of the extent to which they may shape DS in SAC dendrites, is a significant advancement in the field.

Reviewer #2 (Public Review):

Summary:

In this study, the authors sought to understand how the receptive fields of bipolar cells contribute to direction selectivity in starburst amacrine cell (SAC) dendrites, their post synaptic partners. In previous literature, this contribution is primarily conceptualized as the 'space-time wiring model', whereby bipolar cells with slow-release kinetics synapse onto proximal dendrites while bipolar cells with faster kinetics synapse more distally, leading to maximal summation of the slow proximal and fast distal depolarizations in response to motion away from the soma. The space-time wiring contribution to SAC direction selectivity has been extensively tested in previous literature using connectomic, functional, and modeling approaches. However, the authors argue that previous functional studies of bipolar cell kinetics have focused on static stimuli, which may not accurately represent the spatiotemporal properties of the bipolar cell receptive field in response to movement. Moreover, this group and others have recently shown that bipolar cell signal processing can change directionally when visual stimuli starts within the receptive field rather than passing through it, complicating the interpretation of moving stimuli that start within a bipolar cell of interest's receptive field (e.g. stimulating only one branch of a SAC or expanding/contracting rings). Thus, the authors choose to focus on modeling and functionally mapping bipolar cell kinetics in response to moving stimuli across the entire SAC dendritic field.

General Comments:

There have been several studies that have addressed the contribution of space-time wiring to SAC process direction selectivity. This study offers a more complete assessment of potential impact space-time wiring can have on this dendrite computation. The experimental results based on glutamate imaging assess the kinetics of glutamate release under conditions of visual stimulation across a large region of retina largely confirm previous observations. By combining their model with this experiment data, they conclude that even the optimal space-time wiring is not sufficient to explain the SAC process DS. Though there is no conclusion which of the many other proposed cellular and circuit mechanisms could potentially contribute to this computation, the limited role for spacetime wiring is firmly established.

Reviewer #3 (Public Review):

Summary:

Gaynes et al. investigated the presynaptic and postsynaptic mechanisms of starburst amacrine cell (SAC) direction selectivity in the mouse retina by computational modeling and glutamate sensitivity (iGluSnFR) imaging methods. Using the SAC computational simulation, the authors initially tested bipolar cell contributions (space-time wiring model, presynaptic effect) and SAC axial resistance contributions (postsynaptic effect) to the SAC DS. Then, the authors conducted two-photon iGluSnFR imaging from SACs to examine the presynaptic glutamate release and found seven clusters of ON-responding and six clusters of OFF-responding bipolar cells. They were categorized based on their response kinetics: delay, onset phase, decay time, and others. Finally, the authors used cluster data to reconstruct bipolar cell inputs to SACs that generate direction selectivity. They concluded that presynaptic effects through the space-time wiring model only account for a fraction of SAC DS.

The article has many interesting findings, and the data presentation is superb. Strengths and weaknesses are summarized below.

Major Strengths:

The authors utilized solid technology to conduct computational modeling with Neuron software and a machine-learning approach based on evolutionary algorithms. Results are effectively and thoroughly presented.

The space-time wiring model was evaluated by changing bipolar cell response properties in the proximal and distal SAC dendrites. Many response parameters in bipolar cells are compared, and DSI is compared in Figure 3. These parameter comparisons are valuable to the field.

Two-photon microscopy was used to measure the bipolar cell glutamate outputs onto SACs by conducting iGluSnFR imaging. All the data sets, including images and transients, are elegantly presented. The authors analyzed the response based on various parameters, which generated more than several response clusters. The clustering is convincing.

Major Weaknesses:

The computational modeling demonstrates intriguing results: SAC dendritic morphology produces dendritic isolation, and a massive input overcomes the dendritic isolation (Figure 1). This modeling seems to be generated by basic dendritic cable properties. However, it has been reported that SAC dendrites express Kv3 and voltage-gated Ca channels. Are they incorporated into this model? If not, how about comparing these channel contributions?

In Figure 9 the authors generated the bipolar cell cluster alignment based on the space-time wiring model. The space-time wiring model has been proposed based on the EM study that distinct types of bipolar cells synapse on distinct parts of SAC dendrites (Green et al 2016, Kim et al 2014). While this is one of the representative Reicardt models, it is not fully agreed upon in the field (see Stincic et al 2016). Therefore, the authors' approach might be only hypothetical without concrete evidence for geographical cluster distributions. Is there any data suggesting each cluster's location on the SAC dendrites? I assume that the iGluSnFR imaging was conducted on the SAC dendritic network, which does not provide geographical information. How about injecting the iGluSnFR-AAV at a lower titer, which labels only some SACs in a tissue? This method may reveal each cluster's location on SAC dendrites.

The authors found that there are seven ON clusters and six OFF clusters, which are supposed to be bipolar cell terminals. However, bipolar cells reported to provide synaptic inputs are T-7, T-6, and multiple T-5s for ON SACs and T-1, T-2, and T-3s for OFF SACs. The number of types is less than the number of clusters. Is there a possibility of clusters belonging to glutamatergic amacrine cells? Please provide a discussion regarding the relations between clusters and cell types.

In Figure 5B, representative traces are shown responding to moving bars in horizontal directions. These did not show different responses to two directional stimuli. Is there any directional preference from other ROIs? Yonehara's group recently exhibited the bipolar cells' direction selectivity (Matsumoto et al 2021). Did you see any correlations with their results? Please discuss.

eLife assessment

This important study provides convincing evidence of the criticality of estradiol – estrogen receptor-mediated upregulation of kisspeptin within neurons of the preoptic area to generate an ovulation-inducing luteinizing hormone surge. The use of in vivo CRIPSR-Cas9 is novel in this system and provides a road map for future studies in reproductive neuroendocrinology. This paper will be of interest to reproductive neuroscientists and endocrinologists.

Author Response

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

Reviewer #1 (Recommendations For The Authors):

Weaknesses: One minor weakness in this study is the conclusion that the guide RNAs didn't seem to have unique effects on GnRH cFos expression or the reproductive phenotypes. Though the data indicate a 60-70% knockdown for both gRNA2 and gRNA3, 3 of the 4 gRNA2 mice had no cFos expression in GnRH neurons during the time of the LH surge, whereas all mice receiving gRNA3 had at least some cFos/GnRH co-expression. In addition, when mice were re-categorized based on reduction (>75%) in kisspeptin expression, most of the mice in the unilateral or bilateral groups received gRNA2, whereas many of the mice that received gRNA3 were in the "normal" group with no disruption in kisspeptin expression. Thus, additional experiments with increased sample sizes are needed, even if the efficacy of the ESR1 knockdown was comparable before concluding these 2 gRNAs don't result in unique reproductive effects.

Response: A draw back of the CRISPR approach is the substantial mosaicism in gene knockdown that is unavoidable due to the nature of DNA repair in each cell relying on several competing pathways. As such, variable knockdown occurs in each mouse as shown in Fig.1C. In the case of the correlation between RP3V ESR1 knockdown and cFos in GnRH neurons (Fig.4C), three gRNA3 and four 4 gRNA2 mice look to be very similar with two gRNA3 mice having knockdown but normal cFos activation. The reasons for this are not known and it is very likely chance that these two (of nine) mice happened to have received gRNA3. This issue becomes exacerbated when animal group numbers unintentionally become smaller with the re-grouping on the basis of kisspeptin expression. The key point here is that each “kisspeptin grouping” remains mixed in terms of gRNA2 and gRNA3 mice so that gRNA3 mice did contribute to the “bilateral group” even if it was only one of four mice. The practicalities of repeating this work are substantial and we do not think justified. We would note that we have previously used Kiss-Cre mice to undertake CRISPR knockdown of ESR1 in RP3V kisspeptin neurons but this failed to target sufficient cells with Cas9 to be experimentally useful.

In Figure 2B (gRNA2), there appear to be 4 mice (4 lines) that have a normal cycle length and then drop to 0 for the cycle length. However, in the Figure legend, it states that there were 3 gRNA2 mice that had a cycle length of 0. Can the authors clarify if it was 4 mice (as indicated in Figure 2B) or 3 mice (as indicated in the legend) that received gRNA2 and exhibited constant estrus?

Response: We have now clarified in the text that 3 gRNA2 mice went into constant estrus, the other mouse was in constant diestrus, also scored as “0” cycles.

In Figure 3H, there is one green data point that has an LH level of around 0.15 and % VGAT with ESR1 around 10%. However, that data point does not appear in Figures 3I and 3J, when you would expect it to be in a similar place (~10%) on the x-axis in those Figures. Was it excluded? If so, please elaborate on the justification for excluding that data point. Response: This was one of the three mice that exhibited no LH pulses so we were only able to report on mean LH levels.

Similarly, in Figure 3K, there is a blue data point that is almost at 0 for both the x-axis and the y-axis. However, that data point does not show up in Figures 3L and 3M around 0 on the x-axis as you would expect. Can the authors clarify where this data point went in Figures 3L and 3M?

Response: This was one of the three mice that exhibited no LH pulses so we were only able to report on mean LH levels.

Reviewer #2 (Recommendations For The Authors):

Finally, the study leaves unanswered the role of GABA itself. As there was no evident phenotype for the ESR1 knockdown in GABA neurons that do not coexpress kisspeptin, this suggests that GABA neurotransmission in the preoptic area is not involved in the estrogen regulation of LH secretion.

Response: The current evidence for no substantial role of GABA from RP3V neurons in the LH surge agrees with our prior in vivo work showing that low frequency optogenetic stimulation of RP3V kisspeptin neurons (only GABA release) has no impact on LH secretion (doi: 10.1523/JNEUROSCI.0658-18.2018).

1. Title. The present data do not clearly demonstrate the blockade of the LH surge. Thus, the statement that "abolishes the preovulatory surge" is an overinterpretation of the findings.

Response: We agree and now use “suppresses the preovulatory surge”.

1. Fig. 3. The numbers of individual data points per group change for the different LH pulse parameters, but they should not (Fig. 3 E-G).

Response: This occurs because one mouse in each group had no LH pulses so that only a mean value was available for these mice.

1. Fig. 4. (4B) The use of only one terminal blood collection (4B) is insufficient to comprehensively characterize the LH surge. It is not possible to conclude what was the actual effect on the LH surge, whether a blockade or altered amplitude or timing. Serial blood samples at 30- or 60-minute intervals should be used. For comparative purposes, the pulsatile LH secretion, which does not seem to be a major outcome in the study, was fully characterized (Fig. 3). (4C) The linear correlation between c-Fos/GnRH and RP3V/ESR1 appears to be well-fitted for gRNA2 (blue) but not gRNA3 (green). Although this is interpreted as an important result of the study, its description and consistency are not so clear. Authors should perform an Anova/ Kruskal-Wallis analysis of these data as a column graph (as in Fig. 4A, B) and discuss the discrepancies between gRNA2 and gRNA3.

Response: As noted in the manuscript, we agree that a single point LH measurement is a relatively inaccurate assessment of the LH surge and very likely underlies much of the substantial variability between mice. However, the extended duration of cFos expression in GnRH neurons at the time of the surge is a much more accurate “single point” indicator and we feel that these results better reflect the state of surge activation. This was noted in the original manuscript.

The linear correlations for the different preoptic regions are undertaken on the complete data set not on individual gRNA groups due to low N numbers in the sub-divided groups. However, column graphs of the RP3V and MPN look the same as Fig.4A and would not change the current interpretation. Please see comments to Reviewer 1 on discrepancies between gRNA2 and 3.

1. Table. It is unclear why the % VGAT with ESR1 was not statistically reduced in the "bilateral" animals. Would this mean that the ESR1 knockdown was not effective in this subgroup with the more consistent effects?

Response: Yes, this would be a reasonable interpretation suggesting that mice with kisspeptin ablation may have had a slightly different overall impact on ESR1 in VGAT neurons. However, this was not discernable from examining the anatomical distribution of AAV.

1. Discussion 1st paragraph. It is interpreted that mice lacking kisspeptin expression "failed to exhibit an LH surge". This should be revised.

Response: We believe that this is a correct statement. Mice lacking kisspeptin had LH surge values between 0.8 and 2.1 ng/ml that we would not consider consistent with being a surge.

1. Immunohistochemistry. It is not clear in the text how a cross-reaction between goat antirabbit 568 (ERa) and goat antirabbit/streptavidin 647 (mChery) was avoided when used in the same reaction.

Response: We were forced into this option due to the lack of different primary antisera to ESR1 and mCherry. We first stained for rabbit ESR1 detected by biotin anti-rabbit/ strep647 which resulted in confined nuclear staining (pseudo-blue; far red). The subsequent staining for rabbit mCherry was detected by goat anti-rabbit 568 that will indeed cross-react by binding to any free epitopes on the rabbit ESR1 primary antibody. However, this would not compromise interpretation as additional 568 labelling to the nucleus is essentially irrelevant when examining far red 647 nm emission and only mCherry cytoplasmic immunoreactivity was used to define the anatomical locations of the AAV spread. This is now clearly explained in the Methods section.

1. Statistical analysis. It is unclear when repeated measures Wilcoxon tests were used in the manuscript.

Response: Thank you for pointing this out. Only Wilcoxon paired test were used. Amended.

1. Data Availability. Further reference to supplementary information files was not found in the manuscript.

Response: A supplementary file with individual data for each mouse is now attached.

Reviewer #3 (Recommendations For The Authors):

Weaknesses:

One aspect for which I have ambiguous feelings is the minimal level of detail regarding the HPG axis and its regulation by estrogens. This limited amount of detail allows for an easy read with the well-articulated introduction quickly presenting the framework of the study. Although not presenting the axis itself nor mentioning the position of GnRH neurons in this axis or its lack of ERα expression is not detrimental to the understanding of the study, presenting at least the position of GnRH neurons in the axis and their critical role for fertility would likely broaden the impact of this work beyond a rather specialist audience.

Response: We agree that this would provide a more complete picture and have modified the Introduction.

The expression of kisspeptin constitutes a key element for the analysis and conclusion of the present work. However, the quality of the kisspeptin immunostaining seems suboptimal based on the representative images. The staining primarily consists of light punctuated structures and it is very difficult to delineate cytoplasmic immunoreactive material defining the shape of neurons in LacZ animals. For some of the cells marked by an arrow, it is also sometimes difficult to determine whether the staining for ESR1 and Kp are in the same focal plane and thus belong to the same neurons. Although this co-expression is not critical for the conclusions of the study, this begs the question of whether Kp expression was determined directly at the microscope (where the focal plan can be adjusted) or on the picture (without possible focal adjustment). Moreover, in the representative image of Kp loss, several nuclei stained for fos (black) show superimposed brown staining looking like a dense nucleus (but smaller than an actual nucleus). This suggests some sort of condensed accumulation of Kp immunoproduct in the nucleus which is not commented. Given the critical importance of this reported change in Kp expression for the interpretation of the present results, it is important to provide strong evidence of the quality/nature of this staining and its analysis which may help interpret the observed functional phenotype.

Response: The kisspeptin immunoreactivity represents both fiber and cytoplasmic staining that can be difficult to discern in some cases. The reviewer can be assured that all counts were undertaken “live” on the microscope so that the plane of focus was adjusted to establish co-labelling. Please note that the nuclear immunoreactivity is for ESR1 and not cFos. Regardless, we struggle to see condensed brown staining over the black nuclei as suggested by the Reviewer. The kisspeptin staining is light brown and confined to just a few fibers in Fig.5B.

As acknowledged in the introduction, this study is not the first to use in vivo Crisp-Cas editing to demonstrate the role of kisspeptin neurons in the control of positive feedback. Although the present work achieved this indirectly by targeting VGAT neurons, I was surprised that the paper did not include more comparison of their results with those of Wang et al., 2019. In particular, why was the present approach more successful in achieving both lack of surge and complete acyclicity?

Response: Wang et al., reported an ~60% reduction in ESR1 expression in Kiss1-Cre (Elias) driven Cas9-expressing cells in the AVPV. As they did not examine kisspeptin expression itself it is unknown to what degree their editing impacted upon kisspeptin neurons. The other differentiating factor was that Wang focussed on the AVPV that only contains a minority of the preoptic kisspeptin population whereas we targeted the AVPV and PeN together. Thus, we suspect that the Wang phenotype arises from insufficient ESR1 knockdown in just the AVPV sub-population of preoptic kisspeptin neurons. We have added a comment to the Discussion as requested.

Moreover, why is it that targeting ESR1 in a selected fraction of GABAergic neurons can lead to a near-complete absence of Kp expression in this region? This is briefly discussed in the penultimate paragraph but mostly focuses on the non-kisspeptinergic GABA neurons rather than those co-expressing the two markers.

Response: We have modified this section to try and make it clear that it is very likely that all RP3V kisspeptin neurons would have been targeted to express Cas9 in this mouse model. Our very recent unpublished RNA scope data show that >80% of RP3V kisspeptin neurons express Vgat mRNA in adult mice.

• Unless I have missed it, the target sequence of the guide RNAs is not mentioned. For reproducibility purposes and to allow comparison with Wang et al., 2019, this information should be provided.

Response: The target sequences for gRNA2 and gRNA3 were around exon 3 and are provided in the Supplementary files of McQuillan et al., 2022 (https://doi.org/10.1038/s41467-022-35243-z). The Wang et al study used the unusual strategy of designing sense and antisense gRNAs against the same sequence in Exon1.

• The first result section is devoted to the design and validation of the guide RNA reports data that were recently published (McQuillan et al., 2022). It is actually acknowledged that the design was reported previously but as written it is not clear whether the actual validation was already reported. This should be said more clearly.

Response: Clarified as requested.

• What was the rationale for choosing gRNA 2 and 3 and not 3 and 6 like in the McQuillan study?

Response: As all three gRNAs worked equally well, the choice of 2 and 3 was entirely pragmatic and only based upon quantities of packaged AAVs that we had produced and were available at the time.

• Introduction, 4th paragraph: It would be clearer if GABAa receptor dynamics was replaced by GABAa receptors mediated neurotransmission or any other verbiage avoiding possible confusion with receptor mobility.

Response: Clarified as requested.

• The section reporting the location of ESR1 knockdown is really clear about the number of animals included in the functional analyses. This is less clear for the number of mice involved in the evaluation of the extent of ESR1 knockdown in the previous section. Specifically, the text reports that 8 and 9 mice received gRNA3 in PVpo and MPN respectively, but the figure shows 7 and 8. This is likely explained by the mouse that was excluded due to normal ESR1 despite the correct positioning of the injection site. It is thus unclear whether this mouse was included in the calculation of the mean percentage of neurons reported in the previous page. Logically, this mouse should have been removed from this analysis and it is assumed that the sample size reported in the text is incorrect.

Response: thank you for picking this up - you are correct. In reviewing this point we realized that the gRNA-lacZ RP3V N numbers also were incorrect and have re-analyzed the data set completely resulting in even stronger significance levels.

• In the section « CRISPR knockdown ESR1 in RP3V GABA-kisspeptin neurons », the extent of ESR1 knockdown is expressed in a counterintuitive manner as « <20% » which is thought to represent the percentage of cells expressing ESR1 rather than the actual knockdown (>80%). This should be clarified.

Response: Corrected as noted.

• Page 6, 3rd line before the last paragraph, there is a mismatch between the highest p value reported in the text (0.242) and the value reported in the table (0.0242).

Response: Corrected thank you.

• Similar to presenting F values for ANOVAs, H values should also be presented for Kruskal Wallis tests.

Response: Values have been added.

• Immunohistochemistry : Origin and reference numbers of all primary antibodies should be reported as well as citation of studies where they have been validated. Although these protocols are standard, information regarding the duration of incubation is necessary to allow replication or for comparison purposes.

Response: We have included the RRID numbers for each of these antisera and added information on incubation times.

• The section on data availability mentions the existence of supplementary files, but I see none.

Response: These have now been attached.

• There are several typos or redundancies to be corrected. Here are a few examples but the manuscript should be carefully double-checked.

Introduction, 3rd paragraph, line 4: upregulated

Introduction, 4th paragraph, 4th line: « to » or « through » not both.

Page 7, line 11 : Kruskal

Page 7, 6th line to the end: does this indicate 'the' general utility?

Page 8, 2nd paragraph, line 13: Crispr

Response: Thank you for these edits.

Reviewer #3 (Public Review):

Summary: The present study sought to investigate the role ERα expressed in Gabaergic neurons of the rostral periventricular aspect of the third ventricle (RP3V) and medial preoptic nucleus (MPN) in the positive feedback using genetically driven Crispr-Cas9 mediated knockdown of ESR1 in VGAT expressing neurons. ESR1 Knockdown in preoptic gabaergic neurons led to an absence of LH surge and acyclicity when associated with severely reduced kisspeptin (Kp) expression suggesting that a subpopulation of neurons co-expressing Kp and VGAT are key for LH surge since total absence of Kp is associated with an absence of GnRH neuron activation and reduced LH surge. Although the implication of kisspeptin neurons was highly suspected already, the novelty of these results lies in the fact that estrogen signaling is necessary in only a selected fraction of them to maintain both regular cycles and LH surge capacity.

Strengths:<br /> Remarkable aspects of this study are, its dataset which allowed them to segregate animals based on distinct neuronal phenotype matching specific physiological outcomes, the transparency in reporting the results (e.g. all statistical values being reported, all grouping variables being clearly defined, clarity about animals that were excluded and why) and the clarity of the writing. Another remarkable feature of this work lies in the analysis of the dataset. As opposed to the cre-lox approach which theoretically allows for the complete ablation of specific neuronal populations, but may lack specificity regarding timing of action and location, genetically driven in vivo Crispr-Cas9 editing offers both temporal and neuroanatomic selectivity but cannot achieve a complete knock down. This approach based on stereotaxic delivery of the AAV encoded guide RNAs comes with inevitable variability in the location where gene knockdown is achieved. By adjusting their original grouping of the animals based on the evaluation of the extent of kisspeptin expression in the target region, the authors obtained a much clearer and interpretable picture. Although only few animals (n=4) displayed absent kisspeptin expression, the convergence of observations suggesting a central impairment of the reproductive axis is convincing. Finally, the observation that the pulsatile secretion of LH is maintained in the absence of Kp expression in the RP3V lends support to the notion that LH surge and pulsatility are regulated independently by distinct neuronal populations, a model put forward by corresponding author a few years ago.

Reviewer #1 (Public Review):

Summary: The current study examines the necessity of estrogen receptor alpha (ESR1) in GABA neurons located in the anteroventral and preoptic periventricular nuclei and the medial preoptic nucleus of hypothalamus. This brain area is implicated in regulating the pre-ovulatory LH surge in females, but the identity of the estrogen-sensitive neurons that are required remains unknown. The data indicate that approximately 70% knockdown of ESR1 in GABA neurons resulted in variable reproductive phenotypes. However, when the ESR1 knockdown also results in a decrease in kisspeptin expression by these cells, the females had disrupted LH surges, but no alterations in pulsatile LH release. These data support the hypothesis that kisspeptin cells in this region are critical for the pre-ovulatory LH surge in females.

Strengths: The current study examined the efficacy of two guide RNAs to knockdown ESR1 in GABA neurons, resulting in an approximate 70% reduction in ESR1 in GABA neurons. The efficacy of this knockdown was confirmed in the brain via immunohistochemistry and the reproductive outcomes were analyzed several ways to account for differences in guide RNAs or the precise brain region with the ESR1 knockdown. The analysis was taken one step further by grouping mice based on kisspeptin expression following ESR1 knockdown and examining the reproductive phenotypes. Overall, the aims of the study were achieved, the methods were appropriate, and the data were analyzed extensively. This data supports the hypothesis that kisspeptin neurons in the anterior hypothalamus are critical for the preovulatory LH surge.

Weaknesses: One minor weakness in this study is the conclusion that the two different guide RNAs didn't seem to have unique effects on GnRH cFos expression or the reproductive phenotypes. Though the data indicate a 60-70% knockdown for both gRNA2 and gRNA3, 3 of the 4 gRNA2 mice had no cFos expression in GnRH neurons during the time of the LH surge, whereas all mice receiving gRNA3 had at least some cFos/GnRH co-expression. In addition, when mice were re-categorized based on reduction (>75%) in kisspeptin expression, most of the mice in the unilateral or bilateral groups received gRNA2, whereas many of the mice that received gRNA3 were in the "normal" group with no disruption in kisspeptin expression. Whether these results occurred by chance or due to differences in the gRNAs remains unknown. Thus, additional experiments with increased sample sizes would be needed, even if the efficacy of the ESR1 knockdown was comparable, before concluding these 2 gRNAs don't have unique actions.

Reviewer #2 (Public Review):

Clarkson et al investigated the impact of in vivo ESR1 gene disruption selectively in preoptic area GABA neurons on the estrogen regulation of LH secretion. The hypothalamic pathways by which estradiol controls the secretion of gonadotrophins are incompletely understood and relevant to a better understanding of the mechanisms driving fertility and reproduction. Using CRISPR-Cas9 methodology, the authors were able to effectively reduce the expression of estrogen receptor (ER)-alpha in GABA neurons located in the preoptic area of adult female mice. The results obtained were rather variable except in the animals with concomitant suppression of kisspeptin in the rostral periventricular region of the third ventricle (RP3V), which displayed interruption of ovarian cyclicity and an altered estradiol-induced LH surge. The experimental approach used allowed for a cell-selective, temporally-controlled suppression of ER-alpha expression, providing further evidence of the critical role of RP3V kisspeptin neurons in the estrogen positive-feedback effect. The preovulatory LH surge is a variable phenomenon and is better evaluated using serial blood sampling. Although the assessment of the estradiol-induced LH surge was performed in one terminal blood collection, c-Fos expression in GnRH neurons was used as a reliable proxy of the LH surge occurrence. The present findings also suggest that GABA neurotransmission in the preoptic area itself is not involved in the positive-feedback effect of estradiol on LH secretion.

eLife assessment

This is an important paper that revises the canonical model of how olfactory sensory neurons choose which odor receptor to express. The data presented in the paper are convincing and the model proposed is provocative and likely to enable future work.

Reviewer #2 (Public Review):

Summary:

The large-conductance Ca2+ activated K+ channel (BK) has been reported to promote breast cancer progression, but it is not clear how. The present study carried out in breast cancer cell lines, concludes that BK located in mitochondria reprograms cells towards the Warburg phenotype, one of the metabolic hallmarks of cancer.

Strengths:

The use of a wide array of modern complementary techniques, including metabolic imaging, respirometry, metabolomics, and electrophysiology. On the whole, experiments are astute and well-designed and appear carefully done. The use of BK knock-out cells to control for the specificity of the pharmacological tools is a major strength. The manuscript is clearly written. There are many interesting original observations that may give birth to new studies.

Weaknesses:

The main conclusion regarding the role of a BK channel located in mitochondria appears is not sufficiently supported. Other perfectible aspects are the interpretation of co-localization experiments and the calibration of Ca2+ dyes. These points are discussed in more detail in the following paragraphs:

1. May the metabolic effects be ascribed to a BK located in mitochondria? Unfortunately not, at least with the available evidence. While it is clear these cells have a BK in mitochondria (characteristic K+ currents detected in mitoplasts) and it is also well substantiated that the metabolic effects in intact cells are explained by an intracellular BK (paxilline effects absent in the BK KO), it does not follow that both observations are linked. Given that ectopic BK-DEC appeared at the surface, a confounding factor is the likely expression of BK in other intracellular locations such as ER, Golgi, endosomes, etc. To their credit, authors acknowledge this limitation several times throughout the text ("...presumably mitoBK...") but not in other important places, particularly in the title and abstract.

2. MitoBK subcellular location. Pearson correlations of 0.6 and about zero were obtained between the locations of mitoGREEN on one side, and mRFP or RFP-GPI on the other (Figs. 1G and S1E). These are nice positive and negative controls. For BK-DECRFP however, the Pearson correlation was about 0.2. What is the Z resolution of apotome imaging? Assuming an optimum optical section of 600 nm, as obtained by a 1.4 NA objective with a confocal, that mitochondria are typically 100 nm in diameter and that BK-DECRFP appears to stain more structures than mitoGREEN, the positive correlation of 0.2 may not reflect colocalization. For instance, it could be that BK-DECRFP is not just in mitochondria but in a close underlying organelle e.g. the ER. Along the same line, why did BK-RFP also give a positive Pearson? Isn´t that unexpected? Considering that BK-DEC was found by patch clamping at the plasma membrane, the subcellular targeting of the channel is suspect. Could it be that the endogenous BK-DEC does actually reside exclusively in mitochondria (a true mitoBK), but overflows to other membranes upon overexpression? Regarding immunodetection of BK in the mitochondrial Percoll preparation (Fig. S5), the absence of NKA demonstrates the absence of plasma membrane contamination but does not inform about contamination by other intracellular membranes.

3. Calibration of fluorescent probes. The conclusion that BK blockers or BK expression affects resting Ca2+ levels should be better supported. Fluorescent sensors and dyes provide signals or ratios that need to be calibrated if comparisons between different cell types or experimental conditions are to be made. This is implicitly acknowledged here when monitoring ER Ca2+, with an elaborate protocol to deplete the organelle in order to achieve a reading at zero Ca2+.

4. Line 203. "...solely by the expression of BKCa-DECRFP in MCF-7 cells". Granted, the effect of BKCa-DECRFP on the basal FRET ratio appears stronger than that of BK-RFP, but it appears that the latter had some effect. Please provide the statistics of the latter against the control group (after calibration, see above).

eLife assessment

The large-conductance Ca2+ activated K+ channel has been reported to promote breast cancer progression. The present study presents convincing evidence that an intracellular subpopulation of this channel reprograms breast cancer cells towards the Warburg phenotype, one of the metabolic hallmarks of cancer. This important finding advances the field of cancer cell metabolism and has potential therapeutic implications. However, additional experiments are needed to ascribe the metabolic reprogramming to BK channels located in mitochondria.

Reviewer #1 (Public Review):

Bischoff et al present a carefully prepared study on a very interesting and relevant topic: the role of ion channels (here a Ca2+-activated K+ channel BK) in regulating mitochondrial metabolism in breast cancer cells. The potential impact of these and similar observations made in other tumor entities has only begun to be appreciated. That being said, the authors pursue in my view an innovative approach to understanding breast cancer cell metabolism.

Considering the following points would further strengthen the manuscript:

Methods:

1. The authors use an extracellular Ca2+ concentration (2 mM) in their Ringer's solutions that is almost twice as high as the physiologically free Ca2+ concentration (ln 473). Moreover, the free Ca2+ concentration of their pipette solution is not indicated (ln 487).

2. Ca2+I measurements: The authors use ATP to elicit intracellular Ca2+ signals. Is this then a physiological stimulus for Ca2+ signaling in breast cancer? What is the rationale for using ATP? Moreover, it would be nice to see calibrated baseline values of Ca2+i.

3. Membrane potential measurements: It would be nice to see a calibration of the potential measurements; this would allow us to correlate the IV relationship with membrane potential. Without calibration, it is hard to compare unless the identical uptake of the dye is shown.

Does paxilline or IbTx also induce depolarization?

4. Mito-potential measurements: Why did the authors use such a long time course and preincubate cells with channel blockers overnight? Why did they not perform paired experiments and record the immediate effect of the BK channel blockers in the mito potential?

5. MTT assays are also based on mitochondrial function - since modulation of mito function is at the core of this manuscript, an alternative method should be used.

Results:

1. Fig. 5G: The number of BK "positive" mitoplasts is surprisingly low - how does this affect the interpretation? Did the authors attempt to record mitoBK current in the "whole-mitoplast" mode? How does the mitoBK current density compare with that of the plasma membrane? Is it possible to theoretically predict the number of mitoBK channels per mitochondrion to elicit the observed effects? Can these results be correlated with the immuno-localization of mitoBK channels?

2. There are also reports about other mitoK channels (e.g. Kv1.3, KCa3.1, KATP) playing an important role in mitochondrial function. Did the authors observe them, too? Can the authors speculate on the relative importance of the different channels? Is it known whether they are expressed organ-/tumor-specifically?

Reviewer #3 (Public Review):

The original research article, titled "mitoBKCa is functionally expressed in murine and human breast cancer cells and promotes metabolic reprogramming" by Bischof et al, has demonstrated the underlying molecular mechanisms of alterations in the function of Ca2+ activated K+ channel of large conductance (BKCa) in the development and progression of breast cancer. The authors also proposed that targeting mitoBKCa in combination with established anti-cancer approaches, could be considered as a novel treatment strategy in breast cancer treatment.

The paper is clearly written, and the reported results are interesting.

Strengths:

Rigorous biophysical experimental proof in support of the hypothesis.

Weaknesses:

A combinatorial synergistic study is missing.

Reviewer #2 (Public Review):

Singh and colleagues employ a methodic approach to reveal the function of the transcription factors Rela and Stat3 in the regulation of the inflammatory response in the intestine.

Strengths of the manuscript include the focus on the function of these transcription factors in hepatocytes and the discovery of their role in the systemic response to experimental colitis. While the systemic response to induce colitis is appreciated, the cellular and molecular mechanisms that drive such systemic response, especially those involving other organs beyond the intestine are an active area of research. As such, this study contributes to this conceptual advance. Additional strengths are the complementary biochemical and metabolomics approaches to describe the activation of these transcription factors in the liver and their requirement - specifically in hepatocytes - for the production of bile acids in response to colitis.

Some weaknesses are noted in the presentation of the data, including a comprehensive representation of findings in all conditions and genotypes tested.

eLife assessment

This important study reveals the RelA/Stat3-dependent gene program in the liver influences intestinal homeostasis. The evidence supporting the conclusions is compelling, although some additional experiments will strengthen the study. The work will be of interest to scientists in gastrointestinal research fields.

Reviewer #1 (Public Review):

Summary:

In this study, the authors showed that activation of RelA and Stat3 in hepatocytes of DSS-treated mice induced CYPs and thereby produced primary bile acids, particularly CDCA, which exacerbated intestinal inflammation.

Strengths:

This study reveals the RelA/Stat3-dependent gene program in the liver influences intestinal homeostasis.

Weaknesses:

Additional evidence will strengthen the conclusion.

1. In Fig. 1C, photos show that phosphorylation of RelA and Stat3 was induced in only a few hepatocytes. The authors conclude that activation of both RelA and Stat3 induces inflammatory pathways. Therefore, the authors should show that phosphorylation of RelA and Stat3 is induced in the same hepatocytes during DSS treatment.

2. In Fig. 5, the authors treated mice with CDCA intraperitoneally. In this experiment, the concentration of CDCA in the colon of CDCA-treated mice should be shown.

Reviewer #3 (Public Review):

Summary:

The authors try to elucidate the molecular mechanisms underlying the intra-organ crosstalks that perpetuate intestinal permeability and inflammation.

Strengths:

This study identifies a hepatocyte-specific rela/stat3 network as a potential therapeutic target for intestinal diseases via the gut-liver axis using both murine models and human samples.

Weaknesses:

1. The mechanism by which DSS administration induces the activation of the Rela and Stat3 pathways and subsequent modification of the bile acid pathway remains clear. As the authors state, intestinal bacteria are one candidate, and this needs to be clarified. I recommend the authors investigate whether gut sterilization by administration of antibiotics or germ-free condition affects 1. the activation of the Rela and Stat3 pathway in the liver by DSS-treated WT mice and 2. the reduction of colitis in DSS-treated relaΔhepstat3Δhep mice.

2. It has not been shown whether DSS administration causes an increase in primary bile acids, represented by CDCA, in the colon of WT mice following activation of the Rela and Stat3 pathways, as demonstrated in Figure 6.

3. The implications of these results for IBD treatment, especially in what ways they may lead to therapeutic intervention, need to be discussed.

eLife assessment

The authors of this manuscript address the following question in the immunology field: what are the transcriptional regulators that allow macrophages to assume different functional phenotypes in response to immune stimuli? They generate a computational map of the gene regulatory networks involved in determining macrophage phenotypes and experimentally validate the role of putative regulatory factors in a myeloid cell line. This study represents a valuable approach to understanding how gene regulation impacts macrophage polarization but the analyses remain incomplete without further validation in primary cells or by examining the identified genes in the in vivo setting.

Reviewer #2 (Public Review):

Summary:

The authors of this manuscript address an important question regarding how macrophages respond to external stimuli to create different functional phenotypes, also known as macrophage polarization. Although this has been studied extensively, the authors argue that the transcription factors that mediate the change in state in response to a specific trigger remain unknown. They create a "master" human gene regulatory network and then analyze existing gene expression data consisting of PBMC-derived macrophage response to 28 stimuli, which they sort into thirteen different states defined by perturbed gene expression networks. They then identify the top transcription factors involved in each response that have the strongest predicted association with the perturbation patterns they identify. Finally, using S. aureus infection as one example of a stimulus that macrophages respond to, they infect THP-1 cells while perturbing regulatory factors that they have identified and show that these factors have a functional effect on the macrophage response.

Strengths:

- The computational work done to create a "master" hGRN, response networks for each of the 28 stimuli studied, and the clustering of stimuli into 13 macrophage states is useful. The data generated will be a helpful resource for researchers who want to determine the regulatory factors involved in response to a particular stimulus and could serve as a hypothesis generator for future studies.

- The streamlined system used here - macrophages in culture responding to a single stimulus - is useful for removing confounding factors and studying the elements involved in response to each stimulus.

- The use of a functional study with S. aureus infection is helpful to provide proof of principle that the authors' computational analysis generates data that is testable and valid for in vitro analysis.

Weaknesses:

- Although a streamlined system is helpful for interrogating responses to a stimulus without the confounding effects of other factors, the reality is that macrophages respond to these stimuli within a niche and while interacting with other cell types. The functional analysis shown is just the first step in testing a hypothesis generated from this data and should be followed with analysis in primary human cells or in an in vivo model system if possible.

- It would be helpful for the authors to determine whether the effects they see in the THP-1 immortalized cell line are reproduced in another macrophage cell line, or ideally in PBMC-derived macrophages.

- The paper would benefit from an expanded explanation of the network mining approach used, as well as the cluster stability analysis and the Epitracer analysis. Although these approaches may be published elsewhere, readers with a non-computational background would benefit from additional descriptions.

- Although the authors identify 13 different polarization states, they return to the M0/M1/M2 paradigm for their validation and functional assays. It would be useful to comment on the broader applications of a 13-state model.

- The relative contributions of each "switching factor" to the phenotype remain unclear, especially as knocking out each individual factor changes different aspects of the model (Fig. S5).

Reviewer #1 (Public Review):

Summary:

Ravichandran et al investigate the regulatory panels that determine the polarization state of macrophages. They identify regulatory factors involved in M1 and M2 polarization states by using their network analysis pipeline. They demonstrate that a set of three regulatory factors (RFs) i.e., CEBPB, NFE2L2, and BCL3 can change macrophage polarization from the M1 state to the M2 state. They also show that siRNA-mediated knockdown of those 3-RF in THP1-derived M0 cells, in the presence of M1 stimulant increases the expression of M2 markers and showed decreased bactericidal effect. This study provides an elegant computational framework to explore the macrophage heterogeneity upon different external stimuli and adds an interesting approach to understanding the dynamics of macrophage phenotypes after pathogen challenge.

Strengths:

This study identified new regulatory factors involved in M1 to M2 macrophage polarization. The authors used their own network analysis pipeline to analyze the available datasets. The authors showed 13 different clusters of macrophages that encounter different external stimuli, which is interesting and could be translationally relevant as in physiological conditions after pathogen challenge, the body shows dynamic changes in different cytokines/chemokines that could lead to different polarization states of macrophages. The authors validated their primary computational findings with in vitro assays by knocking down the three regulatory factors-NCB.

Weaknesses:

One weakness of the paper is the insufficient analysis performed on all the clusters. They used macrophages treated with 28 distinct stimuli, which included a very interesting combination of pro- and anti-inflammatory cytokines/factors that can be very important in the context of in vivo pathogen challenge, but they did not characterize the full spectrum of clusters. Although they mentioned that their identified regulatory panels could determine the precise polarization state, they restricted their analysis to only the two well-established macrophage polarization states, M1 and M2. Analyzing the other states beyond M1 and M2 could substantially advance the field. They mentioned the regulatory factors involved in individual clusters but did not study the potential pathway involving the target genes of these regulatory factors, which can show the importance of different macrophage polarization states. Importantly, these findings were not validated in primary cells or using in vivo models.

Author Response

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

Public Reviews:

Reviewer #1 (Public Review):

The authors sought to understand the neurocomputational mechanisms of how acute stress impacts human effortful prosocial behavior. Functional neuroimaging during an effort-based decision task and computational modeling were employed. Two major results are reported: 1) Compared to controls, participants who experienced acute stress were less willing to exert effort for others, with a more prominent effect for those who were more selfish; 2) More stressed participants exhibited an increase in activation in the dorsal anterior cingulate cortex and anterior insula that are critical for self-benefiting behaviour. The authors conclude that their findings have important insights into how acute stress affects prosociality and its associated neural mechanisms.

Overall, there are several strengths in this well-written manuscript. The experimental design along with acute stress induction procedures were well controlled, the data analyses were reasonable and informative, and the results from the computational modeling provide important insights (e.g., subjective values). Despite these strengths, there were some weaknesses regarding potential confounding factors in both the experimental design and methodological approach, including selective reporting of only some aspects of this complex dataset, and the interpretation of the observations. These detract from from the overall impact of the manuscript. In particular, the stress manipulation and pro-social task are both effortful, raising the possibility that stressed participants were more fatigued. Other concerns include the opportunity for social dynamics or cues during task administration, the baseline social value orientation (SVO) in each group, and the possibility of a different SVO in individuals with selfish tendencies. Finally, Figure 4 should specify whether the depicted prosocial choices include all five levels of effort.

We thank the reviewer for their comments and suggestions. In our response to the recommendations for the author below, we have dealt with the reviewer’s concerns: - we added additional analysis on the role of fatigue and block effects to the supplementary materials. - we provided further information about the role of social cues and dynamics during task administration. - we showed there were no baseline group differences in SVO angle. - we clarified that Figure 4 refers to the proportion of prosocial choices across all effort levels.

Reviewer #2 (Public Review):

This manuscript describes an interesting study assessing the impact of acute stress on neural activity and helping behavior in young, healthy men. Strengths of the study include a combination of neuroimaging and psychoneuroendocrine measures, as well as computational modeling of prosocial behavior. Weaknesses include complex, difficult to understand 3-way interactions that the sample size may not be large enough to reliably test. Nonetheless, the study and results provide useful information for researchers seeking to better understand the influence of stress on the neural bases of complex behavior.

The stressor was effective at eliciting physiological and psychological stress responses as shown in Figure 2.

Higher perceived stress in more selfish participants (lower social value orientation (SVO) angle) was associated with lower prosocial responding (Figure 4). How can we reconcile this finding with the finding (presented on page 15) that those with a more prosocial SVO showed a significant decline in dACC activation to subjective value at increasing levels of perceived stress? This seems contrary to the behavioral response.

A larger issue with the study is that the power analysis presented on page 23 is based on a 2 (between: stress v. control) by 2 (within: self v. other) design. Most of the reported findings come from analyses of 3-way interactions. How can the readers have confidence in the reliability of results from 3-way interaction analyses, which were not powered to detect such effects?

We thank the reviewer for their comments and suggestions. When considering the influence of dACC activation on the behavioural response (i.e., proportion of prosocial choices), it is important to consider the difference in activation to SVself relative to SVother: - The difference in activation to SVself relative to SVother negatively predicted the proportion of prosocial choices, so more activation to SVself relative to SVother predicted a lower proportion of prosocial choices. - Similarly, SVO angle negatively predicted the difference in activation to SVself relative to SVother, so more activation to SVself relative to SVother was related to a lower (more individualistic) SVO angle (this is shown by the interaction between Recipient and SVO angle in Figure 4; right panel). In both cases, differences in prosociality (i.e. SVO angle or the proportion of prosocial choices) were related to differences in dACC activation to SVself relative to SVother.

Thus, we agree the finding that those participants with a more prosocial SVO showed a significant decline in dACC activation to SV overall (across SVself and SVother) at increasing levels of perceived stress is difficult to interpret. We expected a three-way interaction between Recipient, SVO angle and Perceived Stress to mirror the behavioural results, rather than a two-way interaction between SVO angle and Perceived Stress. We have now acknowledged this in the Discussion, whilst also highlighting the work of Schulreich et al. (2022) who report a related finding.

We have now added the following section to the results:

“When linking activation difference in dACC and AI to behaviour, we found that – independent of the stress manipulation – the difference in activation between SVself and SVother in the dACC predicted the proportion of prosocial choices. Thus, greater activation to SVself relative to SVother predicted a lower proportion of prosocial choices (B=-0.704, SE=0.339, P=0.041). This relationship was not present in the AI (B=-0.423, SE=0.332, P=0.205).”

And we have added the following to the discussion:

“Additionally, participants with a more prosocial SVO showed reduced responses in the dACC to SV (across both self and other trials) at greater levels of perceived stress (Figure 4; middle panel). This suggests that more prosocial individuals may become less sensitive to SV overall following stress, whilst the responses of more individualistic participants to SV do not change under stress. Trying to link these activation differences to changes in effortful prosocial behaviour is difficult given the absence of the three-way interaction between SVO angle, Perceived Stress and Recipient, which would have mirrored the behavioural results. Overall, differences in activation between SVself and SVother in the dACC predicted the proportion of prosocial choices, so greater activation to SVself relative to SVother predicted a lower proportion of prosocial choices. Thus, it remains unclear how activation differences to SV across both self trials and other trials relates to changes in prosocial behaviour under stress. Schulreich et al. (2022) found that a decline in charitable donations following increases in cortisol in high mentalisers was related to a reduced representation of value for donations in the right dlPFC. Whilst there are important differences between the present study and Schulriech et al. (2022), such as the way in which prosocial behaviour was measured, both studies suggest that existing differences in social preferences and abilities (i.e., mentalising, SVO) can have a detrimental effect on the neural representations of value following acute stress. Establishing how these changes in neural representations of value impact behaviour following acute stress is a challenge for future work.”

Concerning the power calculation, we have acknowledged this as a limitation in the discussion.

“Our power calculation was based on a 2 x 2 design (Group x Recipient), however, several of our key findings involved three-way interactions (e.g. between Group, Recipient and Effort). Thus, future studies should aim to replicate our effects with larger sample sizes to ensure the robustness of these effects.

Recommendations for the authors

Reviewer #1 (Recommendations For The Authors):

1. The authors employed an integrative approach on inducing acute stress by combining the strengths of MIST and TSST, as shown by a robust stress response in cortisol. However, some concerns regarding the stress manipulation and the effort-based task need to be addressed. The authors justified the order of deployment as necessary to maintain stress responses throughout the scanning period. It is unclear whether and how potential order effects were controlled, and whether the effort-task performance in the front and back of the line might have different effects in a 90-minute experiment.

Moreover, the stress manipulation itself involved a complex mental arithmetic task, which might have influenced participants' willingness to exert effort for others in the prosocial task. As shown in Figure 3, the proportion of participants working decreases as the effort levels increase for both self and other conditions in the stress and control groups. It is thus possible that participants could consider the prosocial task as an opportunity to take a break from the demanding arithmetic task. It would be helpful to present results from the different runs, particularly for the pre and post three runs.

We thank the reviewer for highlighting this potential issue. We have added several analyses to the supplementary analysis to explore potential block effects and fatigue effects. Here we provide a summary of the key findings.

Firstly, we investigated participants’ ratings of the effort levels, which they experienced immediately before and after the study, to investigate potential fatigue effects. We found that following the experiment compared to the before, participants in the stress group rated squeezing to the required effort levels as more physically demanding compared to the control group (p=.037). There were no group differences in how much more effort they reported exerting (p=.824) or how uncomfortable it was (p=.351) compared to before the experiment. Thus, overall the stress group found it more physically demanding to squeeze to the effort levels following the experiment. Crucially, however, increases in how physically demanding participants found it to squeeze to the required effort levels were not correlated with the number of effortful choices in the Self and Other condition in either group (all Ps >0.4). This suggests that whilst stressed participants rated squeezing to the required effort level as more physically demanding following the task relative to before, this was not related to how often participants exerted effort for self or other rewards.

Secondly, we investigated potential block effects. We repeated the mixed effects logistic regression reported in the manuscript but included the interaction between the factors Group, Recipient and Block (1:6) in the model. Although both groups showed a decline in the number of effortful choices during the experiment, the two-way interaction between Group and Block (p=.188) nor the three-way interaction between Group, Recipient and Block were significant (p=.138). This shows that whilst there was a decline in the number of effortful choices throughout the experiment, this was not more pronounced in the stress group, nor was it more pronounced in the stress group for self relative to other effortful choices compared to the control group. Additionally, the key three-way interaction between Group, Recipient and Block was unaffected when controlling for potential block effects. We now also plot the data by block in the supplementary materials (Figure S3).

Please see the section in the Supplementary Material and a summary of these analyses also appears in the manuscript in the Results section

“We conducted additional analyses to rule out the influence of potential fatigue and block effects (see Fatigue and block effects in the Supplementary Materials). In short, the stress group rated squeezing to the required effort level as more physically demanding immediately after the experiment compared to before, which was not seen in the control group (Figure S2). However, this was not related to the number of effortful choices for self or other rewards (Table S2). Moreover, when we conducted the same mixed effects logistic regression on participants’ choices but also included the interaction between Group, Recipient and Block, there was no significant three-way interaction between these factors, nor a significant two-way interaction between Group and Block (Figure S3). Additionally, the three-way interaction between Group, Recipient and Effort was unaffected when controlling for potential block effects (Type III Wald test χ2[4]=22.06, P<0.001). Thus, whilst the stress group rated squeezing to the required effort level as more physically demanding following the experiment, this was not related to the number of effortful choices (for self or other) and the effects of Block on effortful choices (for self or other) did not differ between the group. Thus, changes in how physically demanding participants rated squeezing to the effort levels did not influence decisions to exert effort.”

1. It would be useful to know whether the authors controlled for factors such as familiarity or gender among participants that might influence their choices on the task. If participants were able to interact or observe each other, it is possible that social dynamics played a role in their behavior, which could confound the interpretation of their results. It would be beneficial if the authors could provide further information on how the task was administered and whether any social cues were present.

For the experimental design, although salivary samples and subjective pressure were measured, did the authors measure participants' subjective ratings of other negative emotions?

Participants did not have the chance to see or interact with the participants in the “other” condition. Participants were told at the start of the experiment that they would be earning money for the next participant in the study, called Thomas. Thus, as all participants were men, the name of the participants was gender matched. Moreover, as they did not see or interact with the next participant, familiarity was controlled across participants.

We have now added a section p. 8 to clarify this:

“As all participants were men, the name of the next participant was gender matched (all participants were told he was called Thomas; see Methods). Moreover, as participants did not see or interact with the next participant, familiarity was controlled across participants.”

We have now added a plot to the supplementary materials (Figure S4) showing the changes in the ratings of the emotions. Apart from the emotions anxious and disgusted, all other emotions (calm, happy, bad, sad, surprised, angry) showed a significant sample timepoint (1:8) by group (stress, control) interaction, thus mirroring the results for the perceived stress ratings. We now refer to this figure in the manuscript on p. 8:

“for changes in other emotions during the experiment please see Figure S4”

1. Regarding the data analysis section, the authors' analysis is careful overall and the results about SVO are interesting. It would be interesting to know if baseline SVO was similar across both stress and control groups, and if there were any differences in SVO among participants with more individualistic or selfish tendencies. Regarding Figure 4, it would be helpful if the authors clarified whether the vertical coordinate "prosocial choices" is a combination of the five levels of effort or if it is specific to one level. Additionally, it would be useful to explore whether there is a correlation between SVO and prosocial choices and whether effort level could be used as a covariate to control for potential confounding effects. These suggestions could improve the clarity and strength of their contributions.

There were no differences in SVO angle between the control group and stress group (p=.956). There was also a significant correlation between SVO angle and the proportion of prosocial choices across the whole sample. This has now been reported in the manuscript on p. 13:

“There were no existing differences in SVO angle between the groups (control group mean = 19.33, SD = 8.67; stress group mean = 19.23, SD=8.14; p=0.956). We found that across the whole sample – independent of the stress manipulation – there was a significant correlation between SVO angle and the proportion of prosocial choices (r=0.225, P=0.032). So, as expected, those with a more prosocial SVO angle showed a higher proportion of prosocial choices in the task.

To clarify, the variable “% prosocial choices” is a combination of all the five effort levels. In other words, we took the total number of prosocial choices (‘work’ for other) across all effort levels relative to the total number of effortful choices. We have now clarified this in the manuscript on p. 13. As this was a combination of all effort levels (and reward levels), it was not possible to include effort level as a covariate.

“This measure combined all reward and effort levels.”

1. It is noteworthy that in the dACC, an effect was observed with regard to the interaction between perceived stress and SVO angle. Considering this observation, another suggestion would be for the authors to include visualization in Figure 4 to present the results of this interaction. This could help readers better comprehend the findings and provide a clearer representation of the results.

We have now updated Figure 4 so that it has three panels showing the behavioural and neural results concerning SVO angle as well as the relationship between SVO angle and activation to SVself and SVother in the dACC.

1. It would be helpful for readers if the authors could label all statistical plots with appropriate statistical values, effect sizes, and their respective significance levels. By doing so, readers would be able to quickly identify major findings of this study and gauge the degree of significance associated with each plot. The authors should consider including such information in their statistical plots to enhance the comprehensibility of the study results.

We have added statistical values (e.g., beta estimates), including indicators of significance to the plots.

1. The authors selected ROIs based on previous work on stress-related and effort-based decision making (i.e., AI and dACC). While other brain regions may also play a role in decision making and social cognition, the authors could choose to focus on these specific ROIs due to their relevance to the experimental question and hypotheses of this study such as prosocial, mentalizing and subjective values.

We agree that several other ROIs may have also been of interest. However, we decided to restrict our analysis to the dACC and the AI as these two ROIs were the focus of a previous study using the same prosocial effort paradigm (Lockwood et al. 2022) and multiple studies suggest these regions are sensitive to stress effects.

1. The authors chose to use one sample t-test with AUC as a covariate to examine brain activations across all participants regardless of their stress or control condition. This approach could identify brain regions that are associated with perceived stress. However, the authors didn't conduct a simple two sample t-test between stress and control groups since their research question and hypotheses focused on the neurocomputational mechanisms underlying prosocial decision-making during stress. Regarding the different stages of decision-making, such as offer, force, and outcome, the authors did not conduct specific analyses for each stage. Instead, they used the computational model to estimate the subjective value of each option at each stage, which allowed them to examine the neural correlates of different value-related parameters across the entire decision-making process. However, it would be interesting to examine the role of different stages as well.

Our design matrix modelled three events during each trial: the offer, force, and outcome phase (as per Lockwood et al. 2022). However, our hypotheses and research question for the effects of acute stress concerned the offer phase, i.e. when participants were deciding whether to exert effort or not (work vs. rest). Therefore, we decided to limit our reporting to this event. We have clarified this on p. 32 in the Methods:

“Our hypotheses and research questions concerning the effects of acute stress concerned the offer phase, i.e., when participants were deciding whether to exert effort or not (work vs. rest). Therefore, we limited our reporting to this event.”

1. The authors' findings pertaining to individual differences are intriguing, particularly for individuals with selfish tendencies to exhibit lower pro-social tendencies under stress. Additionally, group variations in effortful behavior related to benfitting others, relative to oneself, are more evident at lower effort levels rather than higher ones. The authors could dedicate more space in the discussion section to discuss the potential mechanisms involved and address the absence of pertinent theoretical support.

We have now extended the discussion to further outline potential mechanisms. Broadly, we interpret our findings in terms of compromised executive functioning under acute stress: “downregulation of the brain’s ‘executive control network’ (Hermans et al., 2014)”. In the original submission, we focused on changes in inhibition and shifts to habitual/automatic processing. We have now expanded this to include a section on cognitive flexibility (see below). Note that changes in executive functioning have been widely reported following stress (see Shields et al., 2016 for a meta-analyses). However, which specific executive functions influenced our observed changes in prosocial behaviour is an exciting avenue for future work.

We have added this section on p. 20-21 concerning cognitive flexibility:

“The dlPFC has also been implicated in cognitive flexibility under acute stress. For example, Kalia et al. (2018) used functional near infrared spectroscopy to show that reduced cognitive flexibility under stress was related to changes in activation in the dlPFC in men. In our study, participants in the control group were more likely to exert effort for self rewards compared to other rewards at higher, but not at lower, levels of effort. Whilst participants in the stress group favoured exerting effort for self rewards at every effort level (Figure 3). This consistent preference for self rewards compared to other rewards at all effort level suggests that stressed participants did not adapt their social behaviour in response to changing contextual information. This supports multiple studies showing reduced cognitive flexibility under stress (Goldfarb et al., 2017; Kalia et al., 2018; Raio et al., 2017; Shields et al., 2016). An exciting avenue for future work is to test whether individual differences in executive functions, such as inhibition and cognitive flexibility, predict changes in social behaviour following acute stress. This would be analogous to the finding in non-social domains, where greater working memory capacity protects against stress-induced changes in learning (Otto et al., 2013).

Reviewer #2 (Recommendations For The Authors):

The manuscript suggests that the stress group made more selfish responses than the control group at lower, but not higher, levels of effort (as shown in Figure 3). I recommend that Figure 3, showing these data, be modified for clarity. Currently, data for the between-subjects comparison (Control and Stress groups) are linked by a dashed line. This linkage (at least in my mind) connotes that these data points are from the same people at different times. In fact, the within-subjects data are not linked by a line, but are noted by different colored symbols. Please reconsider how these data are presented.

We have redrawn Figure 3. For each effort level, the self vs. other manipulation is shown on the x axis and the two groups (Control vs. Stress) are shown by black and grey lines. For each group, the lines are connected to show that the Self vs. Other manipulation is a within-subject manipulation.

eLife assessment

This study reports useful findings on the influence of acute stress on prosocial behavior and its neural correlates. The approach is solid, combining neuroimaging and neuroendocrine measures with computational cognitive modeling. The results will be of interest to researchers seeking to better characterize the influence of stress on neural computations mediating complex social behavior.

Joint Public Review:

This study sought to characterize the influence of acute stress on prosocial behavior, combining an effort-based task with neuroimaging, neuroendocrinological measures, and computational cognitive modeling. Two major results are reported: 1) Compared to controls, participants who experienced acute stress were less willing to exert effort for others, with more prominent effects for those who were more selfish; 2) More stressed participants exhibited an increase in activation in the dorsal anterior cingulate cortex and anterior insula, which are implicated in self-benefiting behavior. The approach is sophisticated and the findings are informative. Concerns regarding potential confounds and data reporting were addressed in a revised submission.

Author Response

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

Thank you for the response and reviews of our manuscript eLife-RP-RA-2023-86638 “Energetics of the Microsporidian Polar Tube Invasion Machinery”. We are grateful for the comments and constructive criticism from all three reviewers, which have helped us to improve our manuscript.

As a summary to the editor, we here provide a list of the major revisions we have implemented to address all the comments provided by the referees.

1. We added Supplementary Section A.9 and Figure S4 to explain the details of calculation and have magnified sketches of flow fields.

2. We clarified the term "required pressure" to "required pressure differences", and explained that the same pressure differences can be achieved by either positive or negative pressure. We invoke the fact that the spore wall buckled inward to deduce that germination is a negative pressure process.

3. We only rank the hypotheses based on calculation of total energy requirement. The peak pressure and peak power requirement calculations are now just for quantitative reference. The ranking of hypotheses does not change.

4. We clarified the definition of topological connections in Section "Systematic evaluation of possible topological configurations of a spore," making it explicit that the topological questions listed only involved the "original PT content" (not PT space at all time).

Thank you again for the opportunity to revise our work. We attach a point-by-point response to the referees below.

Public Reviews:

Reviewer #1 (Public Review):

1. The authors used mathematical models to explore the mechanism(s) underlying the process of polar tube extrusion and the transport of the sporoplasm and nucleus through this structure. They combined this with experimental observations of the structure of the tube during extrusion using serial block face EM providing 3 dimensional data on this process. They also examined the effect of hyperosmolar media on this process to evaluate which model fit the predicted observed behavior of the polar tube in these various media solutions.

We thank the reviewer for their accurate summary of our work. One subtle point, however, is that we examine the effect of hyperviscous media on the polar tube extrusion process, rather than hyperosmolar media. In Supplementary Section A.6 of our updated manuscript, we have shown that the changes in osmolarity due to methylcellulose is negligible.

1. Overall, this work resulted in the authors arriving at a model of this process that fit the data (model 5, E-OE-PTPV-ExP). This model is consistent with other data in the literature and provides support for the concept that the polar tube functions by eversion (unfolding like a finger of a glove) and that the expanding polar vacuole is part of this process. Finally, the authors provide important new insights into the buckling of the spore wall (and possible cavitation) as providing force for the nucleus to be transported via the polar tube. This is an important observation that has not been in previous models of this process.

We thank the reviewer for acknowledging the novelty and importance of our study.

Reviewer #2 (Public Review):

1. Microsporidia has a special invasion mechanism, which the polar tube (PT) ejects from mature spores at ultra-fast speeds, to penetrate the host and transfer the cargo to host. This work generated models for the physical basis of polar tube firing and cargo transport through the polar tube. They also use a combination of experiments and theory to elucidate possible biophysical mechanisms of microsporidia. Moreover, their approach also provided the potential applications of such biophysical approaches to other cellular architecture.

We thank the reviewer for their accurate summary and acknowledging the potential applications on other organisms.

1. The conclusions of this paper are mostly well supported by data, but some analyses need to be clarified. According to the model 5 (E-OE-PTPV-ExP) in P42 Fig. 6, is the posterior vacuole connected with the polar tube? If yes, how does the nucleus unconnected with the posterior vacuole enter the polar tube?

As we mentioned in our glossary and detailed in Section "Systematic evaluation of possible topological configurations of a spore", Model 5 requires the "original PT content" (any material inside the PT prior to cargo entering the tube) to permit fluid flow to posterior vacuole and external environment post anchoring disc rupture, but cannot permit fluid flow to the sporoplasm that is transported through the tube. As the the germination process progresses, our model does not require the connection between PT and posterior vacuole to be maintained afterwards, and that creates space allowing sporoplasm (including nucleus) sporoplasm (including nucleus) to enter PT space through fluid entrainment. We have clarified the definitions in Section "Systematic evaluation of possible topological configurations of a spore" and have additional clarification in the caption of Fig. 6 in the updated manuscript.

1. In Fig. 6, would the posterior vacuole become two parts after spore germination? One part is transported via the polar tube, and the other is still in the spore. I recommend this process requires more experiments to prove.

According to our Model 5, the membrane connection between PT and posterior vacuole must be broken for the infectious cargo to extrude. However, our current data does not allow us to prove nor disprove the membrane fission event. In theory, the membrane content in PT can potentially be severed into multiple parts by Plateau-Rayleigh instability, an interfacial-tension-driven fluid thread breakup mechanism. Note that it is possible to have membrane fission at the time scale of germination process, as when the time scale of shearing is faster than the viscoelastic time of lipid membranes (roughly 10 msec), membrane fission can happen (Morlot & Roux 2013). For time scale longer than viscoelastic time of lipid membrane, protein complexes like dynamin would be required for membrane fission. Future cryo-EM study of the vacuole-PT connection at the anterior tip (and in the spore as a whole) is needed to clarify the physical process. We added this discussion in Section "Predictions and proposed future experiments".

Reviewer #3 (Public Review):

Abstract:

The paper follows a recent study by the same team (Jaroenlak et al Plos Pathogens 2020), which documented the dramatic ejection dynamics of the polar tube (PT) in microsporidia using live-imaging and scanning electron microscopy. Although several key observations were reported in this paper (the 3D architecture of the PT within the spore, the speed and extent of the ejection process, the translocation dynamics of the nucleus during germination), the precise geometry of the PT during ejection remain inaccessible to imaging, making it difficult to physically understand the phenomenon.

This paper aims to fill this gap with an indirect "data-driven" approach. By modeling the hydrodynamic dissipation for different unfolding mechanisms identified in the literature and by comparing the predictions with experiments of ejection in media of various viscosities, authors shows that data are compatible with an eversion (caterpillar-like) mechanism but not compatible with a "jack-in-the-box" scenario. In addition, the authors observe that most germinated spores exhibit an inward bulge, which they attribute to buckling due to internal negative pressure and which they suggest may be a mean of pushing the nucleus out of the PT during the final stage of ejection.

We thank the reviewer for their accurate summary of our work.

Major strengths:

Probably the most impressive aspect of the study is the experimental analysis of the ejection dynamics (velocity, ejection length) in medium of various viscosities over 3 orders of magnitudes, which, combined with a modeling of the viscous drag of the PT tube, provides very convincing evidence that the unfolding mechanism is not a global displacement of the tube but rather an apical extension mechanism, where the motion is localized at the end of the tube. The systematic classification of the different unfolding scenarios, consistent with the previous literature, and their confrontation with data in terms of energy, pressure and velocity also constitute an original approach in microbiology where in-situ and real time geometry is often difficult to access.

We thank the reviewer for acknowledging the novelty and importance of our study.

Major weaknesses:

1a. While the experimental part of the paper is clear, I had (and still have) a hard time understanding the modeling part. Overall, the different unfolding mechanisms should be much better explained, with much more informative sketches to justify the dissipation and pressure terms, magnifying the different areas where dissipation occurs, showing the velocity field and pressure field, etc.

We thank the reviewer for their comments and suggestions. In the Figure S4 and SI Section A.9 of the updated manuscript, we have magnified the sketches with flow field, and added a detailed explanation of the derivations of dissipation terms.

1b. In particular, a key parameter of eversion models is the geometry of the lubrication layers inside and outside the spore (h_sheath, h_slip). Where do the values of h_sheath and h_slip come from? What is the physical process that selects these parameters?

As we described in SI Section A.9, h_sheath was set to be 25 nm based on the observed translucent space around PT in activated spores (Lom 1972), and h_slip was set to be 6 nm based on the observed gap thickness between PT and cargo (Takovarian et al. 2020). Although we don't expect these numbers to be the same for each spore, the uncertainty in these two parameters are much less than the uncertainty in cytoplasmic viscosity (which varies several orders of magnitude) and boundary slip length. Our sensitivity testing on cytoplasmic viscosity and boundary slip length thus covers any uncertainty in h_sheath or h_slip already.

1c. For clarity, the figures showing the unfolding mechanics in the different scenarios should be in the main text, not in the supplemental materials.

We have added Figure S4 and SI Section A.9 to explain the details of our sketches. We believe, however, putting all the details of the mechanics and how each term is derived in the main text may detract from the flow of the manuscript, and result in it being less accessible to readers who are not as familiar with the physics. We therefore decided to keep this information in supplemental materials.

2a. The authors compute and discuss in several places "the pressure" required for ejection, but no pressure is indicated in the various sketches and no general "ejection mechanism" involving this pressure is mentioned in the paper.

In the updated manuscript, we have changed the term “pressure” to “pressure difference” or “required pressure difference”. We did not calculate the detailed pressure field around each structure, but only estimated the required pressure difference to overcome the drag force and drive fluid flow in various spaces. We also clarified this point in Section "Developing a mathematical model for PT energetics".

Also, as we mentioned in Section “Posterior vacuole expansion and the role of osmotic pressure”, we made no assumptions on how the pressure difference is generated in this paper. The unfolding mechanism of polar tube, how eversion is sustained, and the driving mechanism are ongoing research projects, and we decided not to make premature comments on that without strong support from experiments or simulation results.

2b. What is this "required pressure" and to what element does it apply?

The “required pressure” in the manuscript indicates the required pressure difference between the spore and the tip of the polar tube for it to push the tip forward and sustain the fluid flow within the polar tube. In the updated manuscript, we thus changed the term “required pressure” to “required pressure difference”. We also added this clarification to Section "Developing a mathematical model for PT energetics".

2c. I understand that the article focuses on the dissipation required to the deployment of the PT but I find it difficult to discuss the unfolding mechanism without having any idea on the driving mechanism of the movement. How could eversion be initiated and sustained?

As we mentioned in Section “Posterior vacuole expansion and the role of osmotic pressure”, we made no assumptions on how the energy, pressure or power is generated in this paper. We agree that the unfolding mechanism of the polar tube, how eversion is sustained, and the driving mechanism are important questions, and these are ongoing research projects. As no assumptions about this are required for our models, we decided not to comment on these aspects without strong support from experiments or simulation results. We have clarified this in Section “Posterior vacuole expansion and the role of osmotic pressure” of the updated manuscript.

1. Finally, the authors do not explain how pressure, which appears to be a positive, driving quantity at the beginning of the process, can become negative to induce buckling at the end of ejection. Although the hypothesis of rapid translocation induced by buckling is interesting, a much better mechanistic description of the process is needed to support it.

As discussed in Point 2-b above, the “required pressure” actually means “required pressure difference”. The same pressure difference can possibly be achieved by either positive pressure (the spore has a higher pressure than the ambient, pushing the fluid into PT) or negative pressure (the PT tip has a lower pressure than the ambient, sucking the fluid from the spore). Hydrodynamic dissipation analysis alone cannot tell the differences between positive or negative pressure, as it only tells you the required pressure differences between the spore and the polar tube tip. It will have to be inferred from the implied mechanisms or other evidence. We added these discussions in the 4th paragraph of Section "Developing a mathematical model for PT energetics" in the updated manuscript.

That being said, from our observations of buckled spore walls, it is still sufficient to deduce that the polar tube ejection process is a negative pressure driven process. For the spore wall to buckle inwards, the ambient pressure has to be higher than the pressure within the spore, but that would contradict with the positive pressure hypothesis as elaborated above. We added these clarifications in the 2nd paragraph of Section "Models for the driving force behind cargo expulsion".

References:

Lom, J. (1972). On the structure of the extruded microsporidian polar filament. Zeitschrift Für Parasitenkunde, 38(3), 200–213.

Takvorian, P. M., Han, B., Cali, A., Rice, W. J., Gunther, L., Macaluso, F., & Weiss, L. M. (2020). An Ultrastructural Study of the Extruded Polar Tube of Anncaliia algerae (Microsporidia). The Journal of Eukaryotic Microbiology, 67(1), 28–44.

Morlot, S., & Roux, A. (2013). Mechanics of dynamin-mediated membrane fission. Annual Review of Biophysics, 42, 629–649.

Reviewer #1 (Recommendations For The Authors):

The work is solid and supported by the experimental data presented, the literature and the biophysical modeling.

1. The model (Model 5) indicates that the polar tube is connected to the posterior vacuole and that the contents of this vacuole may be transported by the polar tube before the sporoplasm. This needs experimental validation in the future, which will require the identification of posterior vacuole markers (i.e. proteins specific to this structure). I find the topology of this idea difficult to understand. If the polar tube is outside of the sporoplasm membrane then how does it connect to the posterior vacuole? If the expanded posterior vacuole is still in the spore at the end of germination then how does the sporoplasm get out?

Model 5 requires the "original PT content" (any material inside the PT prior to cargo entering the tube) to permit fluid flow to posterior vacuole and external environment post anchoring disc rupture, but cannot permit fluid flow to sporoplasm. As the germination process progresses, our model does not require the connection between PT and posterior vacuole to be maintained afterwards, and that creates space allowing sporoplasm (including nucleus) to enter PT space through fluid entrainment.

We agree with the reviewer that the specific predictions from Model 5 need to be experimentally validated in the future, and identification of posterior vacuole markers is a good direction. We have mentioned this in Section "Predictions and proposed future experiments".

1. I have always thought that the polaroplast was the initial cargo in the polar tube and that this formed the limiting membrane of the sporoplasm and nucleus after passage through the polar tube (i.e., the limiting membrane of the sporont).

In this manuscript, we only analyze the possible topology of the organelles that are relevant for energy dissipation calculations. Our final hypothesis (E-OE-PTPV-ExP) indicates that there is a limiting membrane of the infectious cargo as they pass through PT, but the energy calculation cannot tell you where this membrane comes from. That being said, our final hypothesis is consistent with the common belief that polaroplast provides the limiting membrane of the sporoplasm, even though our analysis neither proved nor disproved it.

1. I understand that the model indicates that during eversion the end of the PT moves away from the posterior vacuole allowing the sporoplasm access to the PT lumen, however, I am not clear how this process occurs (although I understand the reason that this model was the best fit for the available data). Does the model distinguish between connected (as in the PV is in the polar tube lumen) to the idea of it being in proximity (i.e. the PT is at the PV at the start of eversion)?

As we mentioned in our reply to Point 1 of the same reviewer above, "connectivity" simply means whether fluid flow is permitted across the end connections among organelles and sub-spaces within the spores. For Model 5, the content of posterior vacuole can pass to the original PT content and to the external environment post anchoring disc disruption through fluid flow, but not to sporoplasm. However, as the germination progresses, the PT does not have to maintain its spatial proximity or membrane connection to posterior vacuole, as the topological connectivity questions are pertaining to the "original PT content". We clarified this point in Section "Systematic evaluation of possible topological configurations of a spore" in the updated manuscript.

Reviewer #2 (Recommendations For The Authors):

1. The connection of polar tube and posterior vacuole need to be analyzed by Cryo -EM.

We thank the reviewer for their comments. This work is underway.

Reviewer #3 (Recommendations For The Authors):

1a. As stated in the public review, the explanation and description of the unfolding mechanism should be much better described and associated with clear sketches, magnifying all the areas where the flow shear rate is concentrated (surrounding zone, lubrication inside and outside the spore, etc) and drawing the velocity field, the boundary solid motion and pressure distribution in order to clearly understand, for each model, the dissipation and pressure terms given in figs. S2 and S3.

In the updated manuscript, we added Figure S4 to enlarge all the regions where fluid shear is considered, with sketches of velocity fields.

1b. This is particularly important for explaining the eversion models (see comment in the Public Review) but even the "jack-in-the-box" model sketched in Fig. S2 is confusing: Why does the blue tube disappear outside the spore? What happens to the tube in this case?

The blue tube in the sketch of Model 1 in Fig. S2 is the fluid between the two outermost layers of PT, not the PT itself. We have clarified that in the newly added Fig. S4.

1. Many ejection mechanisms based on the deployment of invaginated appendages have been described in the literature (e.g. Zuckerkandl Biol. Bull. 1950, Karabulut et al Nat. Com. 2022) and also mimicked for robotic applications (e.g. Hawkes et al Science Robotics 2017). Although this is not the main topic of the paper, it would be very useful if the authors could discuss in the introduction the most acceptable theory for motion generation (eversion driven by an overpressure in the spore?). In the current version, this comes too late in the discussion.

As we discussed in Section “Lack of biophysical models explaining the microsporidian infection process”, PT eversion is the most widely accepted hypothesis because of experimental evidence (e.g. microscopic observations of PT extrusions, and pulse-labeling of half-ejected tubes). However, whether or not it is driven by an overpressure in the spore remains controversial. In fact, our observations of inwardly buckled spores indicates that the ejection process likely involves negative pressure.

In our work, we thus take a data-driven approach to generate models for the physical basis of PT extrusion process, without immediately assuming that eversion is the correct hypothesis. It would therefore not make sense to have elaborated discussion on other eversion mechanisms in Introduction.

1. About the physical constraints, I understand that the stored energy must be the same when the viscosity is changed (by conservation of energy), but what physical basis do you have for requiring that the power and pressure also be the same (lines 295-298)? For e.g. when a spring is stretched and released in a very viscous fluid without inertia, the total energy dissipated is the same whatever the viscosity but the power is not the same. The formulation of the chosen physical constraints should be better justified.

We thank the reviewer for their feedback. In our updated manuscript, we only use total energy requirement for the ranking, and the peak pressure difference requirement and peak power requirements are calculated just for quantitative reference. The ranking of the 5 hypotheses does not change.

1. About the mechanism for cargo translocation, authors should explain the physical origin of the hypothetical negative pressure. How could the initial positive pressure become negative?

As we mentioned in our reply to Point 3 of the same reviewer in the public review, the “required pressure” actually means “required pressure difference”. The same pressure difference can possibly be achieved by either positive pressure (the spore has a higher pressure than the ambient, pushing the fluid into PT) or negative pressure (the PT tip has a lower pressure than the ambient, sucking the fluid from the spore). Hydrodynamic dissipation analysis alone cannot tell the differences between positive or negative pressure, as it only tells you the required pressure differences between the spore and the polar tube tip. It will have to be inferred from the implied mechanisms or other evidence. We added these discussions in the 4th paragraph of Section "Developing a mathematical model for PT energetics" in the updated manuscript.

That being said, from our observations of buckled spore walls, it is still sufficient to deduce that the polar tube ejection process is a negative pressure driven process. For the spore wall to buckle inwards, the ambient pressure has to be higher than the pressure within the spore, but that would contradict with the positive pressure hypothesis as elaborated above. We added these clarifications in the 2nd paragraph of Section "Models for the driving force behind cargo expulsion".

More minor comments:

1. The videos are amazing but it is not clear if the PT is ejected through a bulk fluid or if the spores (and ejected PT) are in contact with a solid.

As described in Supplementary Section A.6, purified spores were spotted on a coverslip and let water evaporate. 2.0 μL of germination buffer (10 mM Glycine-NaOH buffer pH 9.0 and 100 mM KCl) with different concentration (0%, 0.5%, 1%, 2%, 3%, 4%) of methylcellulose was added to the slide and place the coverslip on top. So the spore is attached to the coverslip and ejected through a bulk liquid of germination buffer.

1. S2 caption: please be precise that H is the Heaviside step function.

We have updated the captions for both Figure S2 and S3 to make it explicit.

1. Line 233 a pi is missing, no?

We thank the reviewer for their careful read. We have corrected that.

1. The notations are quite unfortunate and confusing. In fluid mechanics capital D usually refers to the dissipation, capital C to the drag coefficient. It would be much clearer to call D the dissipation power (in Watt) and P the pressure requirement (in Pa), whatever the mechanism and put the different contribution (drag, lubrication, cytoplasm flow) in subscript.

We thank the reviewer for their feedback. The notation of this paper is challenging as there are many symbols while keeping everything relatively intuitive to both people with biology background and physics background. We will keep these feedback in mind in our future work.

1. Fig S2: what is D (in the formula of the total dissipation power)? Why not use R instead?

D is the PT diameter, as we mentioned in the caption. We keep that as it is used in the definition of the shape factor.

1. Fig S3 why the pressure requirement for the "jack-in-the-box" hypothesis is 2\mu (vLf(epsilon)/R^2)?

We have now elaborated the calculation in SI Section A.9.

1. Lines 486-497: Although shear thinning fluids have their viscosity that decreases with the shear rate, in most cases the resistance (stress) still increases with speed with these fluids. Is mucin a "velocity-weakening" fluid, i.e. a fluid in which stress decreases when shear rate increases.

We agree that stress still increases with speed for most shear thinning fluids. The mechanical properties of mucin solution strongly depend on its compositions and buffers. In our discussion, we thus simply mention this possibility without claiming whether mucin (or other biopolymer environment that microsporidia species actually experience in vivo) is a velocity-weakening fluid or not.

eLife assessment

This important study combines experiments and fluid mechanics modeling to determine the mechanism of the ultrafast ejection of the polar tube of the Microsporidia parasite and of transport through this tube. The methods and the analysis, based on the variation of the viscosity of the external medium, are compelling and allow for the first time to discriminate among proposed ejection mechanisms. This approach where simple physical principles are used for distinguishing between mechanisms when the precise geometry is inaccessible through imaging is potentially applicable to other systems in microbiology.

Reviewer #1 (Public Review):

The authors used mathematical models to explore the mechanism(s) underlying the process of polar tube extrusion and the transport of the sporoplasm and nucleus through this structure. They combined this with experimental observations of the structure of the tube during extrusion using serial block face EM providing 3 dimensional data on this process. They also examined the effect of hyperosmolar media on this process to evaluate which model fit the predicted observed behavior of the polar tube in these various media solutions. Overall, this work resulted in the authors arriving at a model of this process that fit the data (model 5, E-OE-PTPV-ExP). This model is consistent with other data in the literature and provides support for the concept that the polar tube functions by eversion (unfolding like a finger of a glove) and that the expanding polar vacuole is part of this process. Finally, the authors provide important new insights into the bucking of the spore wall (and possible cavitation) as providing force for the nucleus to be transported via the polar tube. This is an important observation that has not been in previous models of this process.

Reviewer #2 (Public Review):

The paper follows a recent study by the same team (Jaroenlak et al Plos Pathogens 2020), which documented the dramatic ejection dynamics of the polar tube (PT) in microsporidia using live-imaging and scanning electron microscopy. Although several key observations were reported in this paper (the 3D architecture of the PT within the spore, the speed and extent of the ejection process, the translocation dynamics of the nucleus during germination), the precise geometry of the PT during ejection remain inaccessible to imaging, making it difficult to physically understand the phenomenon.

This paper aims to fill this gap with an indirect "data-driven" approach. By modeling the hydrodynamic dissipation for different unfolding mechanisms identified in the literature and by comparing the predictions with experiments of ejection in media of various viscosities, authors shows that data are compatible with an eversion (caterpillar-like) mechanism but not compatible with a "jack-in-the-box" scenario. In addition, the authors observe that most germinated spores exhibit an inward bulge, which they attribute to buckling due to negative pressure difference. They suggest that this buckling may be a mean of pushing the nucleus out of the PT during the final stage of ejection.

Major strengths:

The most compelling aspect of the study is the experimental analysis of the ejection dynamics (velocity, ejection length) in medium of various viscosities over 3 orders of magnitudes, which, combined with a modeling of the viscous drag of the PT tube, provides very convincing evidence that the unfolding geometry is not a global displacement of the tube but rather an apical extension, where the motion is localized at the end of the tube.

The systematic classification of the different unfolding scenarios, consistent with the previous literature, and their confrontation with data in terms of energy, pressure and velocity also constitute an original approach in microbiology, where in-situ and real time geometry is often difficult to access.

Major weaknesses:

The revised version has clarified some details of the model, adding a paragraph and a figure in the Sup Mat. However, in my opinion, it remains difficult to understand the precise topology and ejection mechanism from the various sketches presented in the article.

The article does not address the mechanical driver (force) of ejection, and the role of pressure is unclear. The revised version replaced the term "negative pressure" with "negative pressure difference", arguing that a positive or negative pressure difference could not be differentiated. However, it is not clear how a lower pressure in the spore than in the bath could eject the tube outside.

Author Response

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

Reviewer #1 (Public Review):

In this study, the authors investigated the role of MAM and the Notch signaling pathway in the onset of the atrophic phenotype in both in vivo and in vitro models. The rationale used to obtain the data is one of the main strengths of the study. Already from the reading, the reasoning scheme used by the authors in setting up the study and evaluating the data obtained is clear. Using both cellular and mouse models in vivo consolidates the data obtained. The authors also methodologically described all the choices made in the supplementary section. A weakness, on the other hand, is the failure to include averages and statistical data in the results that would give a quantifiable idea of the data obtained. To complete the picture, the authors could also investigate the possible involvement of the intrinsic apoptosis pathway as well as describe probable metabolic shifts to muscle cells in atrophic conditions. The rationale used by the authors to obtain the result is linear. The data obtained are useful for understanding the onset and characterization of the atrophic phenotype under disuse and microgravity conditions. The methods used are in line with those used in the field and can be a starting point for other studies. The cellular models are well described in the Materials and methods section. The selected mouse models followed a logical rationale and were in line with the intended aim.

We thank this reviewer for comments that have led us to clarify several points.

Reviewer #1 (Recommendations For The Authors):

• In order to reinforce and justify the results obtained, I would suggest that the authors include numerical and statistical data in the results obtained.

Answer) As the reviewer suggested, we have incorporated actual numerical and statistical data into each graph in all figures.

• With the aim of better framing the picture of an atrophic muscle phenotype caused by microgravity or disuse, I would advise the authors to also focus on the possible involvement of the intrinsic apoptosis pathway. To this end, it would be interesting to assess a possible relationship between MAM and apoptosis. It would be useful to integrate this part into the discussion.

Answer) It has been shown that suppression of Mfn2 expression attenuates calcium influx into mitochondria during apoptosis-inducing stimuli, thereby inhibiting apoptosis (Martins de Brito & Scorrano, Nature 2008), however, in our study, we found that apoptotic pathways, including Caspase3 or p-AKT were not significantly altered in differentiated human myocytes by microgravity for 7 days in culture, suggesting that microgravity-induced apoptosis is not an initial pathway to MAM. We have added these data in the new supplementary file 3 and mentioned it in the results.

• In addition to TA, did the authors investigate what was seen in other muscles impacted by microgravity? If so, I would recommend supplementing what is available or, on the contrary, justifying the exclusivity of the choice of TA.

Answer) It has been reported that the soleus, a slow-type muscle is more susceptible than the fast-type tibialis anterior muscle during gravity changes, so it makes more sense for the content of this study to analyze the soleus muscle. However, we chose the tibialis anterior muscle as our target because it provides the most stable results as a site for stem cell transplantation to observe muscle regeneration.

• The authors affirm that there is an altered distribution and morphology of mitochondria under microgravity conditions. To corroborate this assertion, I would recommend including a morphological image that confirms it.

Answer) The morphology of mitochondria in cultured myotubes, as observed by mitotracker staining in Figure 4G, varied widely, from finely divided to fused even within a single fiber compared to MFN2-mutated human iPS cells, making it difficult to conclude whether these changes were brought about by microgravity. Therefore, in this study, we have shown that they are reduced in microgravity by the difference in fluorescence intensity of mitotracker, which is directly proportional to mitochondrial activity.

• It would be interesting if the authors would show whether there are changes in myosin expression or metabolic changes in cells subjected to microgravity and in the cell model with Mnf2 deletion. It would also be interesting to evaluate this in the presence of DAPT.

Answer) As the reviewer’s suggestion, we have checked MYH1, MYH3, and MYH7 transcripts in differentiated myotubes under microgravity, with or without DAPT in the new supplementary file 12. We have added the data showing that not MYH1 but MYH7 transcript was partially recovered in the Results.

A detailed description of the metabolic analyses with myogenic cells cultured in microgravity conditions will be published elsewhere (Sugiura et al., “Mitochondria aconitase is a main target for unloading-mediated mitochondria dysfunction toward muscle atrophy”, in preparation). We have described it in the Materials and methods of the manuscript.  

Reviewer #2 (Public Review):

In this study, the authors examined how the maintenance of mitochondrial-associated endoplasmic reticulum membranes (MAM) is critical for the prevention of muscle atrophy under microgravity conditions. They observed, a reduction in MAM in myotubes placed in a microgravity condition; in addition, MFN2-deficient human iPS cells showed a decrease in the number of MAM, similar to in myotubes differentiated under microgravity conditions, in addition to the activation of the Notch signaling pathway. The authors, moreover, observed that treatment with the gamma-secretase inhibitor with DAPT preserved the atrophic phenotype of differentiated myotubes in microgravity and improve the regenerative capacity of Mfn2-deficient muscle stem cells in dystrophic mice. The entire study was well conducted, bringing an interesting analysis in vitro and in vivo of aging conditions. In my opinion, it is necessary to improve the analysis of both genes and proteins to better support the conclusions

The study can contribute to a better understanding of one of the major problems of aging, such as muscle atrophy and inhibition of muscle regeneration, emphasizing the importance of the NOTCH pathway in these pathological situations. The work will be of interest to all scientists working on aging

We thank this reviewer for the positive comments and remarks that we have attempted to address.

Reviewer #2 (Recommendations For The Authors):

Results:

In Figure 1b authors observed an increase in the transcripts of MuRF1 and FBXO32 after 7 days of microgravity condition. I suggest to investigate the protein expression of these genes to give more validation to this data.

Answer) As the reviewer’s suggestion, we have investigated the western blotting with atrophic markers in microgravity samples. These data have been added in Figure 1D.

Moreover, I suggest investigating not only Myogenin as an earlier gene of myotubes formation but also MRF4.

Methods:

I suggest when doing real-time PCR not to use a single gene as housekeeping but the average of three genes, to avoid the influence of a single housekeeping gene affecting the results.

Answer) As the reviewer’s suggestion, we have investigated MRF4 expression by qPCR experiments with 3 different housekeeping genes (RPL13a, GAPDH, and ACTB). Our experiments showed no significant differences among these three housekeeping genes. We have added these data to Figure 1C and Methods in the manuscript.

eLife assessment

This interesting and important manuscript combines in vitro and in vivo experiments to investigate the reciprocal regulation between mitochondria-associated membranes and Notch signaling in skeletal muscle atrophy, with implications beyond the single subfield of muscle atrophy. The methods, data, and analyses are solid and broadly support the claims.

Reviewer #1 (Public Review):

In this study, the authors investigated the role of MAM and the Notch signalling pathway in the onset of the atrophic phenotype in both in vivo and in vitro models. The rationale used to obtain the data is one of the main strengths of the study. Already from the reading, the reasoning scheme used by the authors in setting up the study and evaluating the data obtained is clear. Using both cellular and mouse models in vivo consolidates the data obtained. The authors also methodologically described all the choices made in the supplementary section.

Reviewer #2 (Public Review):

In this study, the authors examined how maintenance of mitochondrial-associated endoplasmic reticulum membranes (MAM) are critical for the prevention of muscle atrophy under microgravity conditions. They observed, a reduction in MAM in myotubes placed in a microgravity condition; in addition, MFN2-deficient human iPS cells showed a decrease in the number of MAM, similar to in myotubes differentiated under microgravity conditions, in addition to the activation of the Notch signaling pathway. The authors, morover, obsreved that by treatment with the gamma-secretase inhibitor with DAPT preserved from the atrophic phenotype of differentiated myotubes in microgravity and improve the regenerative capacity of Mfn2-deficient muscle stem cells in dystrophic mice.

The entire study was well conducted, bringing an interesting analysis in vitro and in vivo of aging condition. In my opinion it is necessary to implement the analysis of both genes and proteins for better supporting the conclusions

The study can contribute to better understand one of the major problems of aging, such as muscle atrophy and inhibition of muscle regeneration, emphasizing the importance of NOTCH patway in these pathological situations. The work will be of interest to all scientist working on aging.

Author Response

We thank the reviewers and editor for their careful evaluation of our manuscript, and we appreciate their favorable assessment of our work. Below, we clarify a few points concerning the relationship between our study and previous studies evaluating ligand docking to protein models.

As reviewer 2 correctly notes, several previous assessments of AF2 models have simply excluded templates above a sequence identity cutoff when using AF2 to predict structures. Such AF2 predictions are still informed by all structures in the PDB before April 30, 2018, because these structures were used to train AF2—that is, to determine the tens of millions of parameters (“weights”) in the AF2 neural network. Machine learning methods nearly always perform better when evaluated on the data used to train them than when evaluated on other data. For this reason, we consider AF2 models only for proteins whose structures were not used to train AF2—that is, for proteins whose structures were not available in the PDB before April 30, 2018.

Previous papers (including Beuming and Sherman, 2012, https://doi.org/10.1021/ci300411b) have shown a clear correlation between the binding pocket RMSD of a protein model and pose prediction accuracy based on that model. Our main findings are unexpected in light of these previous reports: we find that AF2 models yield pose prediction accuracy similar to that of traditional homology models despite having much better binding pocket RMSDs, and that AF2 models yield substantially worse pose prediction accuracy than experimentally determined structures with different ligands bound despite having similar binding pocket RMSDs.

Reviewer 2 also correctly notes that previous papers have described AF2 models as “apo models,” because these models do not include coordinates for bound ligands. As noted by the AF2 developers (e.g., https://alphafold.ebi.ac.uk/faq), however, AF2 is designed to predict coordinates of protein atoms as they might appear in the PDB, and AF2 models are thus frequently consistent with structures in the presence of ligands even though those ligands are not included in the models. GPCR structures in the PDB, including those used to train AF2, nearly always contain a ligand in the orthosteric binding pocket. An AF2 model of a GPCR should thus not be viewed as an attempt to predict the GPCR’s structure in the unliganded (apo) state.

Finally, we did not apply flexible docking in this study because previous work has found that standard flexible docking protocols typically improve pose prediction performance only when given prior information on which amino acid residues to treat as flexible. For example, previous studies that performed successful flexible docking to AF2 models generally used prior knowledge of the ligand’s experimentally determined binding pose to identify the residues to treat as flexible.

Author Response

Reviewer #3 (Public Review):

Summary:

The manuscript from Tariq and Maurici et al. presents important biochemical and biophysical data linking protein phosphorylation to phase separation behavior in the repressive arm of the Neurospora circadian clock. This is an important topic that contributes to what is likely a conceptual shift in the field. While I find the connection to the in vivo physiology of the clock to be still unclear, this can be a topic handled in future studies.

Strengths: The ability to prepare purified versions of unphosphorylated FRQ and P-FRQ phosphorylated by CK-1 is a major advance that allowed the authors to characterize the role of phosphorylation in structural changes in FRQ and its impact on phase separation in vitro.

Weaknesses: The major question that remains unanswered from my perspective is whether phase separation plays a key role in the feedback loop that sustains oscillation (for example by creating a nonlinear dependence on overall FRQ phosphorylation) or whether it has a distinct physiological role that is not required for sustained oscillation.

The reviewer raises the key question regarding data suggesting LLPS and phase separated regions in circadian systems. To date condensates have been seen in cyanobacteria (Cohen et al, 2014, Pattanayak et al, 2020) where there are foci containing KaiA/C during the night, in Drosophila (Xiao et al, 2021) where PER and dCLK colocalize in nuclear foci near the periphery during the repressive phase, and in Neurospora (Bartholomai et al, 2022) where the RNA binding protein PRD-2 sequesters frq and ck1a transcripts in perinuclear phase separated regions. Because the proteins responsible for the phase separation in cyanobacteria and Drosophila are not known, it is not possible to seamlessly disrupt the separation to test its biological significance (Yuan et al, 2022), so only in Neurospora has it been possible to associate loss of phase separation with clock effects. There, loss of PRD-2, or mutation of its RNA-binding domains, results in a ~3 hr period lengthening as well as loss of perinuclear localization of frq transcripts. A very recent manuscript (Xie et al., 2024) calls into question both the importance and very existence of LLPS of clock proteins at least as regards to mammalian cells, noting that it may be an artefact of overexpression in some places where it is seen, and that at normal levels of expression there is no evidence for elevated levels at the nuclear periphery. Artefacts resulting from overexpression plainly cannot be a problem for our study nor for Xiao et al. 2021 as in both cases the relevant clock protein, FRQ or PER, was labeled at the endogenous locus and expressed under its native promoter. Also, it may be worth noting that although we called attention to enrichment of FRQ[NeonGreen] at the nuclear periphery, there remained abundant FRQ within the core of the nucleus in our live-cell imaging.

Cohen SE, et al.: Dynamic localization of the cyanobacterial circadian clock proteins. Curr Biol 2014, 24:1836–1844, https://doi.org/10.1016/j.cub.2014.07.036.

Pattanayak GK, et al.: Daily cycles of reversible protein condensation in cyanobacteria. Cell Rep 2020, 32:108032, https://doi.org/10.1016/j.celrep.2020.108032.

Xiao Y, Yuan Y, Jimenez M, Soni N, Yadlapalli S: Clock proteins regulate spatiotemporal organization of clock genes to control circadian rhythms. Proc Natl Acad Sci U S A 2021, 118, https://doi.org/10.1073/pnas.2019756118.

Bartholomai BM, Gladfelter AS, Loros JJ, Dunlap JC. 2022 PRD-2 mediates clock-regulated perinuclear localization of clock gene RNAs within the circadian cycle of Neurospora. Proc Natl Acad Sci U S A. 119(31):e2203078119. doi: 10.1073/pnas.2203078119.

Yuan et al., Curr Biol 78: 102129, 2022. https://doi.org/10.1016/j.ceb.2022.102129

Pancheng Xie, Xiaowen Xie, Congrong Ye, Kevin M. Dean, Isara Laothamatas , S K Tahajjul Taufique, Joseph Takahashi, Shin Yamazaki, Ying Xu, and Yi Liu (2024). Mammalian circadian clock proteins form dynamic interacting microbodies distinct from phase separation. Proc. Nat. Acad. Sci. USA. In press.

Author Response

We appreciate the reviewers’ and editors’ advice on further improving this manuscript. We have provided point by point responses to the reviewers’ comments mentioned below. A revised version of this manuscript will be uploaded within a few weeks.

Authors’ response to Reviewer 1 comments:

• We appreciate the reviewer’s time in highlighting the strengths and weaknesses of this manuscript.

• Per the reviewer’s advice, we will provide further description of the methods in a revised version of this manuscript.

• The interpretation about the biological threat in response to elevated glycosuria in renal Glut2 KO mice is based on our observation that these mice exhibit changes in acute phase proteins measured using plasma proteomics. We will further discuss this in a revised version of this manuscript.

• We acknowledge that this manuscript provides a resource for future mechanistic studies. Because multiple secreted proteins are changed between the control and experimental groups, some of them could be causal and others corelative in the context of enhancing compensatory glucose production in response to elevated glycosuria. Through future studies we will determine the causal proteins that trigger the increase in glucose production and identify the tissues that secrete these proteins.

• We have shown previously (Cordeiro et al., Diabetologia 2022) that renal Glut2 deficiency doesn’t change insulin sensitivity (i.e. renal Glut2 KO mice don’t exhibit insulin resistance despite the activation of the HPA axis). It is likely that the massive glycosuria in renal Glut2 KO mice may overcome or mask the phenotype of insulin resistance potentially induced by an increase in the stress hormones.

• In this manuscript, our major goal was to determine how elevated glycosuria leads to an increase in compensatory glucose production. We are not suggesting renal Glut2 as a therapeutic in this manuscript (that was already demonstrated in our previously published manuscript, Cordeiro et al., Diabetologia 2022).

Authors’ response to Reviewer 2 comments:

1) Renal Glut2 KO mice didn’t exhibit sex differences for the variables reported in our previous manuscript (Cordeiro et al., Diabetologia 2022). Therefore, in the present manuscript we decided to use male or female mice depending on their availability for each reported experiment. Per the reviewer’s advice, we will describe these details including age and sexes in each figure legend.

2) For the method description, we have cited previous publications and mentioned ‘as described previously’. Based on the reviewer’s suggestion we will further describe the methods in detail to clarify the reviewer’s concerns. In addition, we will include age and sexes in the legends of each figure.

3) For littermate controls, we had used Glut2loxp/loxp mice (which are like WT controls as described in Cordeiro et al., Diabetologia 2022) that were injected with tamoxifen exactly in the same way as the experimental mice. Het mice for Cre were not used as controls because they would have confounded the results as pointed out by the reviewer.

4) Because elevated HPA activity is known to increase blood glucose levels, we suggested ‘the HPA axis may…..’. Given the nature of this manuscript, we agree the secreted proteins identified using plasma proteomics could contribute to enhanced glucose production directly or through secondary mechanisms. Afferent renal denervation using capsaicin reduced blood glucose levels concomitant with the suppression of the HPA axis in renal Glut2 KO mice. Based on these findings we speculated that the HPA axis may be partly responsible for increasing glucose production in renal Glut2 KO mice.

We had considered using CRF antagonist and glucocorticoid receptor antagonists to determine the causal role of the HPA axis in contributing to the increase in glucose production in renal Glut2 KO mice. However, these drugs activate compensatory mechanisms including changes in insulin sensitivity. Therefore, use of these drugs would further confound the results instead of providing a clarity on the causal role of the HPA axis in enhancing glucose production in renal Glut2 KO mice.

5) We understand the reviewer’s concerns whether the results reported here are translatable to humans. Please note that expression of SGLT2 is not kidney-specific; therefore, pleiotropic effects of SGLT2 inhibition in tissues other than the kidney cannot be excluded in animal models and humans. In contrast, the mouse model reported in this manuscript is kidney-specific Glut2 KO mice. Therefore, phenotype produced in renal Glut2 KO mice cannot be directly compared with that produced after SGLT2 inhibition. It may be too early to speculate whether the results reported in this manuscript are translatable to humans.

In the referred research papers by the reviewer, the authors have used either models of different types of diabetes or included individuals with diabetes in their study. Notedly, diabetes itself affects the HPA axis independently of SGLT2 or GLUT2 inhibition. Therefore, it may not be appropriate to compare results obtained from animals or individuals with diabetes with that reported in this manuscript from renal Glut2 KO mice.

6) Yes, we are currently performing mechanistic studies including assessment of mitochondrial function in renal Glut2 KO mice to determine whether and how the kidneys sense loss of glucose in urine.

7) We apologize for the lack of methods description. We will provide additional method details in a revised version of this manuscript. All the assays were performed as per manufacturer’s instructions. Aliquots of the same samples were used for analyses of the hormones and for consistency across different assays.

eLife assessment

This study presents a useful characterization of mechanisms underlying glycosuria-mediated increase in compensatory glucose production in Glut2 knockout mice. The strength of support is incomplete but the data represent a starting point for further studies regarding the role of the HPA axis and acute phase proteins in regulating blood glucose during glycosuria.

Reviewer #1 (Public Review):

Summary:<br /> In this study, Faniyan and colleagues build on their recent finding that renal Glut2 knockout mice display normal fasting blood glucose levels despite massive glucosuria. Renal Glut2 knockout mice were found to exhibit increased endogenous glucose production along with decreased hepatic metabolites associated with glucose metabolism. Crh mRNA levels were higher in the hypothalamus while circulating ACTH and corticosterone were elevated in this model. While these mice were able to maintain normal fasting glucose levels, ablating afferent renal signals to the brain resulted in substantially lower blood glucose levels compared to wildtype mice. In addition, the higher CRH and higher corticosterone levels of the knockout mice were lost following this denervation. Finally, acute phase proteins were altered, plasma Gpx3 was lower, and major urinary protein MUP18 and its gene expression were higher in renal Glut2 knockout mice. Overall, the main conclusion that afferent signaling from the kidney is required for renal glut2 dependent increases in endogenous glucose production is well supported by these findings.

Strengths:<br /> An important strength of the paper is the novelty of the identification of kidney-to-brain communication as being important for glucose homeostasis. Previous studies had focused on other functions of the kidney modulated by or modulating brain activity. This work is likely to promote interest in CNS pathways that respond to afferent renal signals and the response of the HPA axis to glucosuria. Additional strengths of this paper stem from the use of incisive techniques. Specifically, the authors use isotope-enabled measurement of endogenous glucose production by GC-MS/MS, capsaicin ablation of afferent renal nerves, and multifiber recording from the renal nerve. The authors also paid excellent attention to rigor in the design and performance of these studies. For example, they used appropriate surgical controls, confirmed denervation through renal pelvic CGRP measurement, and avoided the confounding effects of nerve regrowth over time. These factors strengthen confidence in their results. Finally, humans with glucose transporter mutations and those being treated with SGLT2 inhibitors show a compensatory increase in endogenous glucose production. Therefore, this study strengthens the case for using renal Glut2 knockout mice as a model for understanding the physiology of these patients.

Weaknesses:<br /> A few weaknesses exist. Clarification of some aspects of the experimental design would improve the manuscript. However, most concerns relate to the interpretation of this study's findings. The authors state that loss of glucose in urine is sensed as a biological threat based on the HPA axis activation seen in this mouse model. This interpretation is understandable but speculative. Importantly, whether stress hormones mediate the increase in endogenous glucose production in this model and in humans with altered glucose transporter function remains to be demonstrated conclusively. For example, the paper found several other circulating and local factors that could be causal. In addition, the approach used in these studies cannot definitively determine whether renal glucose production or only hepatic glucose production was altered. This model is also unable to shed light on how elevated stress hormones might interact with insulin resistance, which is known to increase endogenous glucose production. That issue is of substantial clinical relevance for patients with T2D and metabolic disease. Finally, while findings from the Glut2 knockout mice are of scientific interest, it should be noted that the Glut2 receptor is critical to the function of pancreatic islets and as such is not a good candidate for pharmacological targeting.

Reviewer #2 (Public Review):

Summary:<br /> The authors previously generated renal Glut2 knockout mice, which have high levels of glycosuria but normal fasting glucose. They use this as an opportunity to investigate how compensatory mechanisms are engaged in response to glycosuria. They show that renal and hepatic glucose production, but not metabolism, is elevated in renal Glut2 male mice. They show that renal Glut2 male mice have elevated Crh mRNA in the hypothalamus and elevated plasma levels of ACTH and corticosterone. They also show that temporary denervation of renal nerves leads to a decrease in fasting and fed blood glucose levels in female renal Glut2 mice, but not control mice. Finally, they perform plasma proteomics in male mice to identify plasma proteins with a greater than 25% (up or down) between the knockouts and controls.

Strengths:<br /> The question that is trying to be addressed is clinically important: enhancing glycosuria is a current treatment for diabetes, but is limited in efficacy because of compensatory increases in glucose production.<br /> Also, the mouse line used is an inducible knockout, thus minimizing the impact of compensatory mechanisms engaged in early development.

Weaknesses:<br /> 1) Though the Methods specify that both male and female mice were used, it appears each experiment was performed only on one sex, rather than each experiment being performed on both sexes. For example, renal denervation was performed only on females, whereas all other experiments were performed exclusively on males. This makes it impossible to examine whether there are sex differences in any measures.

2) This study appears to use an inducible Glut2 knockout with tamoxifen, yet nothing describes when the tamoxifen was delivered relative to the experimental manipulations. Was the knockout performed in young animals? In adult animals? This is important both for the ability of readers to repeat the experiment, but also to interpret the results in light of potential compensatory changes (if the knockout was performed at an early age, for example).

3) In Methods, please clarify whether littermate controls were WT, het, or both. If het mice were used as controls, this is potentially problematic.

4) Conclusions like "the HPA axis may contribute to the compensatory increase in glucose production in renal Glut2 knockout mice" (line 215) are premature. All that is shown is that renal Glut2 male mice have elevated HPA activity. There are no experiments establishing causation. For example, the authors could administer a CRF antagonist or a glucocorticoid receptor antagonist in this mouse line, and examine whether this impacts blood glucose. This was not done.

5) If elevated glycosuria drives HPA activity, one would expect to see elevated HPA activity in humans who take SGLT2 inhibitors. Yet, this does not seem to be the case (Higashikawa et al, 2021; see also Perry et al, 2021 for rodent example). This raises the question of whether the glycosuria observed in the mouse line here is relevant to any human conditions. The relevance of the mechanisms proposed here would be much more convincing if a second model of glycosuria was used here (for example, inducing diabetes in mice and treating with SGLT2 inhibitors). Without these types of experiments, any relevance to human conditions is highly speculative and should be reserved for the Discussion. What the authors are studying here is one mechanism for maintaining blood glucose when glycosuria is induced by a genetic knockout.

6) The experiment examining the impact of renal denervation is nice but incomplete. For example, what is the relevance to the hepatic glucose production that was reported? It is interesting that the renal denervation normalized the elevated HPA activity in Glut2 female mice, but it is not clear how this signaling would alter HPA activity.

7) The Methods need to describe the plasma collection procedure for both ELISA and plasma proteomic experiments. What time of day were samples collected? Were samples collected when animals were euthanized from other experiments after experimental manipulations, or in animals without other experimentation?

8) In general, the links between the disparate mechanisms (signals in the plasma, changes in renal activity, changes in HPA activity) are weak. There are more experiments needed to establish a direct kidney-hypothalamus axis. If renal activity elevates blood glucose in the face of glycosuria, why are there no differences in renal activity between control and Glut2 knockout mice? If the blood glucose levels are regulated by renal activity, it must be the sensitivity to the renal activity that differs between control and knockout mice - perhaps this should be investigated. If one stimulates afferent renal nerves, can one drive HPA activation and elevate blood glucose? How are these measures related to the plasma proteins identified? Without these links, this study is descriptive and correlational.

Author Response

We highly appreciate the constructive feedback provided by the reviewers, which we believe will greatly improve the quality of our work. We were encouraged to see that our manuscript was considered to be “important”, of “great interest” as well as to “yield valuable results”.

We also highly appreciate the overall positive eLife assessment. However, we were surprised to read that our “results range from solid from inadequate”. This especially applies given the positive and engaging nature of the reviews which seem to mainly concern the results interpretation being “inadequate” rather than the results themselves. Hence, we kindly request a reconsideration of this aspect of the assessment.

Moreover, there is one Reviewer comment we would like to address directly. Reviewer #3 pointed out that “this study did not conduct a direct association analysis between MetS and cognitive levels without considering subgroup comparisons.” and that “After a thor-ough review of the methods and results sections” she/he “found no direct or strong evidence supporting the authors' claim that the identified latent variables were related to more severe MetS to worse cognitive performance. While a sub-group comparison was conducted, it did not adequately account for confounding factors such as educational level.”.

We appreciate the observations of Reviewer #3 regarding the absence of a direct association analysis between Metabolic Syndrome (MetS) and cognitive levels without subgroup comparisons, and the lack of evidence linking latent variables to MetS severity and cognitive performance. Our apologies for any confusion caused by unclear presentation. Our study incorporated association analyses between MetS, brain structure, and cognition using MetS components, regional cortical thickness, and cognitive performance data in a PLS. These analyses were separately performed on the UK Biobank and HCHS datasets, due to their distinct cognitive assessments. We adjusted for age, sex, and education in the subgroup analyses by removing their effects from the input variables. The primary latent variables demonstrated significant associations with MetS components, cortical thickness, and cognitive scores, indicating that higher obesity, blood pressure, lipidemia, and glycemia levels correlate with lower cognitive performance. These relationships are detailed in supplementary figures S15b and S16b, with negligible loadings for age, sex, and education, confirming effective deconfounding. We acknowledge the reviewer's constructive feedback and will enhance the clarity of the Methods and Results sections, including conducting a mediation analysis.

Furthermore, we strive to incorporate the Reviewers’ other suggestions into the analysis. The revision will include major changes to the manuscript.

In response to Reviewer #1:

• We will revise considering non-fasting plasma glucose as a surrogate marker of insuline resistance.

• We will report Field IDs of the used UK Biobank variables.

• We aim to moderate causal interpretations and reword the indicated passages for clarity.

In response to Reviewer #2:

• We will reconsider claims of binarizing vascular dementia and Alzheimer’s dementia pathophysiology.

• We will further explore the cell type associations of the other latent variables.

• We will expand the discussion regarding conclusions from our results and the future outlook.

In response to Reviewer #3.

• We will add an additional flowchart to detail the virtual histology analysis.

• We will add a discussion of the second latent variable.

• We will conduct a mediation analysis to statistically assess the mediation effect of brain structure on the relationship between MetS and cognitive performance.

We are convinced that with these revisions, our manuscript will align even more closely with the high standards of eLife and make a strong contribution to its distinguished portfolio. We thank you for your consideration.

eLife assessment

This important study enhances our understanding of the relationship between metabolic syndrome (MetS) and brain health from two large-scale datasets and crosses different scales of investigation. The results range from solid to inadequate, with the overall effects of MetS on the brain well supported, but the claimed inference of non-fasting blood glucose reflecting insulin resistance and suggestions of causative link to cognitive function need to be revised or tempered. Overall this study will be of great interest to researchers and clinicians seeking to understand metabolic syndrome.

Reviewer #1 (Public Review):

Summary:<br /> In their study, Petersen et al. investigated the relationship between parameters of metabolic syndrome (MetS) and cortical thickness using partial least-squares correlation analysis (PLS) and performed subsequently a group comparison (sensitivity analysis). To do this, they utilized data from two large-scale population-based cohorts: the UK BioBank (UKB) and the Hamburg City Health Study (HCHS). They identified a latent variable that explained 77% of the shared variance, driven by several measures related to MetS, with obesity-related measures having the strongest contribution. Their results highlighted that higher cortical thickness in the orbitofrontal, lateral prefrontal, insular, anterior cingulate, and temporal areas is associated with lower MetS metric severity. Conversely, the opposite pattern was observed in the superior frontal, parietal, and occipital regions. A similar pattern was then observed in the sensitivity analysis when comparing two groups (MetS vs. matched controls) separately. They then mapped local cellular and network topological attributes to the observed cortical changes associated with MetS. This was achieved using cell-type-specific gene expressions from the Allen Human Brain Atlas and the group consensus functional and structural connectomes of the Human Connectome Project (HCP), respectively. This contextualization analysis allowed them to identify potential cellular contributions in these structures driven by endothelial cells, microglial cells, and excitatory neurons. It also indicated functional and structural interconnectedness of areas experiencing similar MetS effects.

Strengths:<br /> The effects of metabolic syndrome on the brain are still incompletely understood, and such multi-scale analyses are important for the field. Despite the study's sole 'correlation-based' nature, it yields valuable results, including several scales of brain parameters (cortical thickness, cellular, and network-based). The results are robust and benefit from two 'large-scale' datasets, resulting in highly powered statistics.

Weaknesses:<br /> However, some concerns arise regarding certain interpretations and claims made by the authors. In particular, it is not entirely convincing that the authors' results are relevant for studying insulin resistance as a clinical measure of MetS. This is due to the use of non-fasting glycemia as a metric, which the authors claim represents insulin resistance. While non-fasting blood glucose is a potential, albeit poor, indicator of insulin resistance, claiming a direct correlation between insulin resistance and cortical thickness does not seem entirely convincing. By doing so, the authors suggest that insulin resistance might have a weak contribution to cortical thickness abnormalities, with a rather low 'loading' of glycemia compared to the other MetS metrics, although this cannot be conclusively determined from these results.

Reviewer #2 (Public Review):

Summary:<br /> In this manuscript, Petersen et al. aimed for a comprehensive assessment of the relationship between cardiometabolic risk factors and cortical thickness. They found that a latent variable reflecting higher obesity, hypertension, LDL cholesterol, triglyerides, glucose, and lower HDL cholesterol was associated with lower cortical thickness in orbitofrontal, lateral prefrontal, insular, anterior cingulate, and temporal areas. In sensitivity analyses, they showed that this pattern replicated across cohorts and was also consistent with a clinical definition of the metabolic syndrome.

Further, when including cognition in the multivariate analysis, the pattern remained unchanged and indicated that cardiometabolic risk factors were associated with worse cognitive performance across different tests. The authors investigated the cell types implicated in the regions associated with cardiometabolic risk using the Allen brain atlas and found that the density of excitatory neurons type 8, endothelial cells, and microglia reliably co-located with the pattern of cortical thickness. Furthermore, they showed that cortical regions more strongly associated with MetS were more closely structurally & functionally connected than others.

Strengths:<br /> This study performed a comprehensive assessment of the combined association of cardiometabolic risk factors and brain structure and investigated micro- and macroscopic underpinnings. A major strength of the study is the methodological approach of Partial Least Squares which allows the authors to not single out risk factors but to take them into account simultaneously. The large sample size from two cohorts allowed for different sensitivity analyses and convincing evidence for the stability of the first latent variable. The authors demonstrated that the component was also reliably related to cognitive performance, replicating multiple previous studies that evidenced associations of different components of the MetS with worse cognitive performance.

The novel contribution of the study lies in the virtual histology and brain topology investigation of the cortical pattern related to MetS. The virtual histology provided clear evidence of the co-localization of endothelial, glial, and excitatory neuronal cells with the regions of MetS-associated cortical thinning while the brain topology analysis highlighted the disproportionate structural and functional connectivity between associated regions. This analysis provides insights into the role of inflammatory processes and the intricate link between gray matter morphology and microvasculature, both locally and in relation to long-range connectivity. This information is valuable to inform future mechanistic studies.

Weaknesses:<br /> The study is exclusively cross-sectional which does not allow to the authors to disentangle causes from consequences. While studies indicate that most of the differences seen in middle age are probably consequences of the MetS on the vasculature, blood-brain barrier, or inflammatory processes, differences in cortical morphology might also represent a risk factor for weight gain.

Another limitation is the omission of subcortical structures and the cerebellum which might have provided additional information on the pattern of GM differences associated with MetS.

The study is exploratory in nature and for the contextualization analyses it is difficult to judge whether those were selected from a larger pool of analyses. The analysis approach taken to relate the cardiometabolic risk, brain structure, and cognition does not allow the reader to determine whether brain regions most strongly related to the MetS are the ones also most strongly associated with cognitive performance. The cortical pattern arising from the models including cognition is not thoroughly compared to the MetS-only pattern and therefore, it is difficult to estimate to which extent the MetS-related cortical patterns explain variance in cognitive performance.

Reviewer #3 (Public Review):

Summary:<br /> This study investigates the continuous effect of MetS components - namely, obesity, arterial hypertension, dyslipidemia, and insulin resistance - on cortical thickness. It also examines the spatial correlations between MetS effects on cortical thickness with brain cellular and network topological attributes. Additionally, the authors attempt to explore the complex interplay among MetS, cognitive function, and cortical thickness.

The results reveal a latent relationship between MetS and cortical thickness based on a clinical-anatomical dimension. Furthermore, the effect of MetS on cortical thickness is linked to local cell types and network topological attributes. These findings suggest that the authors achieved most, though not all, of their research objectives.

The conclusions are mostly well supported by data and results. However, the use of "was governed by" in the conclusion section suggests a causal relationship. This phrasing is inappropriate given that the study primarily employs correlational analyses.

Strengths<br /> The study presents several strengths:

This study undertakes a comprehensive assessment encompassing the full range of MetS components, such as obesity or arterial hypertension, rather than adopting a case-control study approach (categorizing participants into MetS or non-MetS groups) as seen in some previous research. Utilizing Partial Least Squares (PLS) for correlational analysis effectively addresses issues of multicollinearity (or high covariance among MetS components) and explores the relationship between MetS and brain morphology.

The study leverages two datasets, examining a large sample size of 40,087 individuals. This substantial sample potentially aids in identifying nuanced and underexplored brain anomalies. By incorporating high-quality MRI images, standardized data, and statistical analysis procedures, as well as sensitivity analyses, the results gain robustness, which addresses the limitations of small samples and low reproducibility.

In the context of MetS, this research uniquely employs the concept of imaging transcriptomics, i.e. virtual histology analysis. This approach allows the study to explore intricate relationships between cellular types and cortical thickness anomalies.

Weaknesses<br /> While this work has foundational strengths, the analyses and data seem inadequate to fully support the key claim and analysis. In particular:

After a thorough review of the methods and results sections, I found no direct or strong evidence supporting the authors' claim that the identified latent variables were related to more severe MetS to worse cognitive performance. While a sub-group comparison was conducted, it did not adequately account for confounding factors such as educational level. Additionally, the strength of evidence from such a sub-group comparison is substantially weaker than that from randomized controlled trials or longitudinal cohort studies. Therefore, it is inaccurate for the authors to assert a direct relationship between MetS and cognitive function based on the presented data. A more appropriate research design or data analysis approach, such as mediation analysis, can be employed to address this issue.

The use of the imaging transcriptomics pipeline (virtual histology analysis) to explore the microscale associations with MetS effects on the brain is commendable and has shown promising results. Nevertheless, variations in gene sets may introduce a degree of heterogeneity in the results (Seidlitz, et al., 2020; Martins et al., 2021). Consequently, further validation or exploratory analyses utilizing different gene sets can yield more compelling results and conclusions.

Author Response

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

We are grateful to the reviewers for their remarks, which significantly improved the paper. We repeated the biochemical assay concerning SIRT6 activity on H3-K27Ac and quantified the results as requested. Please find our detailed answers bellow each recommendation of the reviewers.

Major recommendations:

1. Grammatical errors are still common; the authors may need to consider an external editing service if they intend to fix the problems as they indicate that they believe the errors have been removed. The Results section is relatively clean, but parts of the Abstract, Introduction, and Discussion are more difficult to understand, and errors are especially common in the Methods section and those parts of the manuscript that are new in this revision.

We corrected the grammatical errors.

1. The introduction doesn't mention the other structures published; this is considered to be a serious deficiency as it prevents the reader from understanding the context for the contributions described here. Withholding the comparison with (or mention of) the previously published work to the last sentence of the Discussion seems misleading and does not give the reader adequate ability to judge the novelty of the results presented in this manuscript.

A paragraph comparing our paper to the other structures published appear at the end of the discussion. We feel this is still the right place for such a paragraph.

1. The addition of the assay for deacetylation is a significant improvement over the initial submission. This is important both for validating the importance of the acidic patch contacts and for helping to resolve the conflicting reports regarding activity on H3-K27Ac. Given the importance of this assay for the impact of the manuscript, it is not clear why the authors chose to 1) put the data in the supplement instead of in the main manuscript, and 2) provide only single samples without quantitation. These both seem to be significant limitations.

We repeated the experiment and provided quantification of the results. We placed the figure in the main manuscript.

1. The authors should add text or a table to the Methods section explaining which maps were used for each figure. By our count, there are 8 maps and 5 models (plus MD models) based on two datasets, but the relationships among them are not clearly stated, and the names of the maps (such as "Zn-finger focused" and "Rossman-Fold-Focused") might be changed to be more helpful to the reader (for example, the latter includes more than the Rossman fold and might be renamed "Sirt6-focused"). The authors should also explain how the maps were validated, which data were deposited in public repositories, and why some data were not deposited. For example, no statistics or methods regarding how particles were separated into integrated vs. non-integrated motion are provided for the CryoDRGN models. Further, the "two principle movements" described are depicted in 4 maps from two CryoDRGN runs using two separate sets of particles, but the relationships among them are not defined clearly. Finally, the connectivity of densities in Fig 8 are not obvious in the submitted maps. Until these points are addressed, the work is considered incomplete.

AND

1. The PDB model provided for review and submitted to the PDB database shows loosely bound DNA at the nucleosomal entry/exit points near the binding site of SIRT6, but the maps provided for review and submitted to the EMDB show stronger density for the canonical location of the DNA expected at these sites. The CryoDRGN maps support a more extended conformation, but these maps were not deposited or provided for review so their validity cannot be assessed.

We added a section to the methods listing the different maps used for the figures. We deposited the map we used to trance the H2A N-terminal tail (EMD-18497). Unfortunately, we couldn’t deposit the cryoDRGN maps as the deposition system either accepts composite maps, where the consensus should be deposited too or experimental maps, where the deposition of half maps are mandatory. Nevertheless, the cryoDRGN maps are available upon request. We also added a supplementary figure (Supplementary Fig 6) to show how the cryoDRGN analyses were performed.

1. The orientation, angle and threshold used in Fig 1 make it difficult to see the multiple DNA orientations that are visible in the deposited consensus map. Examination of the map suggests that the DNA model submitted to PDB corresponds to a weaker DNA conformation than is present in the map where both DNA conformations are visible. The authors should consider modeling both conformations in their deposited model to provide a more complete, accurate representation of the data. It is concerning that a key conclusion of the manuscript is that the DNA conformation changes upon SIRT6 binding, but density for the canonical position is observable in Fig 8a.

Figure 1 is showing the overall representation of the SIRT6 bound nucleosome structure. We show the DNA linker orientations in the subsequent figure. Figure 8 (now Figure 9) shows the rearrangement of the SIRT6 Rossmann fold domain not the DNA linker.

1. Figure 4 needs a more complete legend, indicating that it is a hybrid of the consensus structure (one color) and the MD simulations (another color). In general, the colors used in the figure should be changed to make the main points more accessible.

As there is a color code for the histones, changing colors might be confusing. The figure legend mentions that panels c, d and e are from MD simulations.

Minor recommendations:

1. Figures 2c, e, and f are not referenced in the text.

We now referenced all figure panels in the text.

1. Consider moving Supp. 5C to Fig. 2 as the models in that figure come from the CryoDRGN maps and not the consensus map.

Supplemental Figure 5c show the DNA linker deviation upon SIRT6 binding from another angle. We prefer to keep it there.

1.) Supp Fig 3 is labeled "ZnF-nucleosome" refinement, but this appears to come from Data Set #2 processing. The map might be labeled ZnF-nucleosome but then a mask should be shown that excludes the Rossman Fold. It is not clear if this is a focused refinement or just a 2.9 A map that was merged with the "Rossman-fold" map.

We changed both supplemental figures accordingly.

1. The orientation of Fig 2 b and e do not show the differences in these models as well as panels c and f. Panels b and e could be replaced with the 4 CryoDRGN maps.

The models reflect the cryoDRGN maps and panels c and f were added to clarify the movement.

1. The MD description should emphasize that the H3 tails are moving with respect to the active site, as it currently suggests the active site is moving.

In the results and in the discussion section we mention that we observe new conformations of the H3 tail, not of the active site.

1. The authors refer to the "flexibility of the Rossmann fold domain," but the Rossman Fold domain isn't flexible, the linkage to the ZnF is flexible. Perhaps "observed conformational space" or "dynamic Rossman-fold domain position" are meant.

The text was changed accordingly.

1. The H2A C-terminal tail present in Fig 1 (bottom right) and Figure 3e is not present in the model in Fig 4a,b.

The H2A tails conformation was not resolved in the cryoDRGN maps so we didn’t model it.

1. The crosslinking agent used is not specified.

The crosslinking agent used is specified more clearly in the methods.

1. Supp Table 1 and EM methods do not agree on the magnification for Dataset #1. Verify nominal versus binned magnification and reported pixel size.<br /> The magnification in the methods was changed.

2. Fig 3F showing the difference between affinity for H2A and H2A.Z-containing nucleosomes would be more convincing with a titration rather than the current comparison of a single concentration.

We agree with this remark however, we find single concentration comparison is convincing enough for the purposes of this paper as it is not a central finding.

1. Fig S1 legend; both the Zn-finger and helix bundle are stated to be shown in green.

Figure S1 legend was changed.

eLife assessment

This manuscript provides a useful reconstruction of the structure of the sirtuin-class histone deacetylase Sirt6 bound to a nucleosome based on cryo-EM observations, and additional characterization of the flexibility of the histone tails in the complex based on molecular dynamics simulations. While similar structures have recently been published, this solid study supports the conclusions of those papers and also includes new insights into the potential dynamics of Sirt6 bound to a nucleosome, insights that help explain its substrate specificity. Unfortunately, the authors do not mention the other recent publications until the end of their Discussion, and therefore provide little opportunity for comparison or context for the results presented.

Reviewer #1 (Public Review):

Smirnova et al. present a cryo-EM structure of a nucleosome-SIRT6 complex to understand how the histone deacetylase SIRT6 deacetylates the N-terminal tail of histone H3. The authors obtained the structure at sub-4 Å resolution and can visualize how interactions between the nucleosome and SIRT6 position SIRT6 to allow for H3 tail deacetylation. Through additional conformational analysis of their cryo-EM data, they reveal that SIRT6 positioning is flexible on the nucleosome surface, and this could accommodate the targeting of certain H3 tail residues. This work is significant as it represents the visualization of a histone deacetylase on its native nucleosomal target and reveals how substrate specificity is achieved. Importantly, it should be noted that recently two additional structures of the nucleosome-SIRT6 complex were already published. Therefore, Smirnova et al. confirm and complement these previous findings. Additionally, Smirnova et al. expand our understanding of the structural flexibility of SIRT6 on the nucleosome and clarify that SIRT6 also shows histone deacetylase activity on H3K27Ac.

Reviewer #2 (Public Review):

Smirnova et al. present a cryo-EM structure of human SIRT6 bound to a nucleosome as well as the results from molecular dynamics simulations. The results show that the combined conformational flexibilities of SIRT6 and the N-terminal tail of histone H3 limit the residues with access to the active site, partially explaining the substrate specificity of this sirtuin-class histone deacetylase. Two other groups have recently published cryo-EM structures of SIRT6:nucleosome complexes; this manuscript confirms and complements these previous findings, with the addition of some novel insights into the role of structural flexibility in substrate selection.

eLife assessment

In this convincing study, the authors examine the interactions between stellate cells and PV+ interneurons in the medial entorhinal cortex, shedding light on the circuit mechanisms that underlie grid cell activity. Huang et al., focus on the spatial distribution of synaptic inputs and report that closely located neuron pairs receive common inputs, suggesting a structured functional organization in the entorhinal cortex. Advanced dual whole-cell patch recordings further reveal patterns of postsynaptic activation, indicating intensive interactions within clusters of these neurons, with weaker interactions between clusters. These important findings offer significant insights into the functional dynamics of the entorhinal cortex.

Reviewer #2 (Public Review):

Summary:<br /> In this study, Huang et al. employed optogenetic stimulation alongside paired whole-cell recordings in genetically defined neuron populations of the medial entorhinal cortex to examine the spatial distribution of synaptic inputs and the functional-anatomical structure of the MEC. They specifically studied the spatial distribution of synaptic inputs from parvalbumin-expressing interneurons to pairs of excitatory stellate cells. Additionally, they explored the spatial distribution of synaptic inputs to pairs of PV INs. Their results indicate that both pairs of SCs and PV INs generally receive common input when their relative somata are within 200-300 ums of each other. The research is intriguing, with controlled and systematic methodologies. There are interesting takeaways based on the implications of this work to grid cell network organization in MEC.

Major concerns<br /> 1) Results indicate that in brain slices, nearby cells typically share a higher degree of common input. However, some proximate cells lack this shared input. The authors interpret these findings as: "Many cells in close proximity don't seem to share common input, as illustrated in Figures 3, 5, and 7. This implies that these cells might belong to separate networks or exist in distinct regions of the connectivity space within the same network.".

Every slice orientation could have potentially shared inputs from an orthogonal direction that are unavoidably eliminated. For instance, in a horizontal section, shared inputs to two SCs might be situated either dorsally or ventrally from the horizontal cut, and thus removed during slicing. Given the synaptic connection distributions observed within each intact orientation, and considering these distributions appear symmetrically in both horizontal and sagittal sections, the authors should be equipped to estimate the potential number of inputs absent due to sectioning in the orthogonal direction. How might this estimate influence the findings, especially those indicating that many close neurons don't have shared inputs?

2) The study examines correlations during various light-intensity phases of the ramp stimuli. One wonders if the spatial distribution of shared (or correlated) versus independent inputs differs when juxtaposing the initial light stimulation phase, which begins to trigger spiking, against subsequent phases. This differentiation might be particularly pertinent to the PV to SC measurements. Here, the initial phase of stimulation, as depicted in Figure 7, reveals a relatively sparse temporal frequency of IPSCs. This might not represent the physiological conditions under which high-firing INs function.

While the authors seem to have addressed parts of this concern in their focal stim experiments by examining correlations during both high and low light intensities, they could potentially extract this metric from data acquired in their ramp conditions. This would be especially valuable for PV to SC measurements, given the absence of corresponding focal stimulation experiments.

3) Re results from Figure 2: Please fully describe the model in the methods section. Generally, I like using a modeling approach to explore the impact of convergent synaptic input to PVs from SCs that could effectively validate the experimental approach and enhance the interpretability of the experimental stim/recording outcomes. However, as currently detailed in the manuscript, the model description is inadequate for assessing the robustness of the simulation outcomes. If the IN model is simply integrate-and-fire with minimal biophysical attributes, then the findings in Fig 2F results shown in Fig 2F might be trivial. Conversely, if the model offers a more biophysically accurate representation (e.g., with conductance-based synaptic inputs, synapses appropriately dispersed across the model IN dendritic tree, and standard PV IN voltage-gated membrane conductances), then the model's results could serve as a meaningful method to both validate and interpret the experiments.

Reviewer #3 (Public Review):

Summary:<br /> This paper presents convincing data from technically demanding dual whole-cell patch recordings of stellate cells in medial entorhinal cortex slice preparations during optogenetic stimulation of PV+ interneurons. The authors show that the patterns of postsynaptic activation are consistent with dual recorded cells close to each other receiving shared inhibitory input and sending excitatory connections back to the same PV neurons, supporting a circuitry in which clusters of stellate cells and PV+IN interact with each other with much weaker interactions between clusters. These data are important to our understanding of the dynamics of functional cell responses in the entorhinal cortex. The experiments and analysis are quite complex and would benefit from some revisions to enhance clarity.

Strengths:<br /> These are technically demanding experiments, but the authors show quite convincing differences in the correlated response of cell pairs that are close to each other in contrast to an absence of correlation in other cell pairs at a range of relative distances. This supports their main point of demonstrating anatomical clusters of cells receiving shared inhibitory input.

Weaknesses:<br /> The overall technique is complex and the presentation could be more clear about the techniques and analysis. In addition, due to this being a slice preparation they cannot directly relate the inhibitory interactions to the functional properties of grid cells which was possible in the 2-photon in vivo imaging experiment by Heys and Dombeck, 2014.

Reviewer #1 (Public Review):

Summary:<br /> The circuit mechanism underlying the formation of grid cell activity and the organization of grid cells in the medial entorhinal cortex (MEC) is still unclear. To understand the mechanism, the current study investigated synaptic interactions between stellate cells (SC) and PV+ interneurons (IN) in layer 2 of the MEC by combing optogenetic activations and paired patch-clamp recordings. The results convincingly demonstrated highly structured interactions between these neurons: specific and direct excitatory-inhibitory interactions existed at the scale of grid cell phase clusters, and indirect interactions occurred at the scale of grid modules.

Strengths:<br /> Overall, the manuscript is very well written, the approaches used are clever, and the data were thoroughly analyzed. The study conveyed important information for understanding the circuit mechanism that shapes grid cell activity. It is important not only for the field of MEC and grid cells, but also for broader fields of continuous attractor networks and neural circuits.

Weaknesses:<br /> (1) The study largely relies on the fact that ramp-like wide-field optogenetic stimulation and focal optogenetic activation both drove asynchronous action potentials in SCs, and therefore, if a pair of PV+ INs exhibited correlated activity, they should receive common inputs. However, it is unclear what criteria/thresholds were used to determine the level of activity asynchronization, and under these criteria, what percentage of cells actually showed synchronized or less asynchronized activity. A notable percentage of synchronized or less asynchronized SCs could complicate the results, i.e., PV+ INs with correlated activity could receive inputs from different SCs (different inputs), which had synchronized activity. More detailed information/statistics about the asynchronization of SC activity is necessary for interpreting the results.

(2) The hypothesis about the "direct excitatory-inhibitory" synaptic interactions is made based on the GABAzine experiments in Figure 4. In the Figure 8 diagram, the direct interaction is illustrated between PV+ INs and SCs. However, the evidence supporting this "direct interaction" between these two cell types is missing. Is it possible that pyramidal cells are also involved in this interaction? Some pieces of evidence or discussions are necessary to further support the "direction interaction".

eLife assessment

This study provides important new insights into how multisensory information is processed in the lateral cortex of the inferior colliculus, a poorly understood part of the auditory midbrain. By developing new imaging techniques that provide the first optical access to the lateral cortex in a living animal, the authors provide convincing in vivo evidence that this region contains separate subregions that can be distinguished by their sensory inputs and neurochemical profiles, as suggested by previous anatomical and in vitro studies. Additional information and analyses are needed, however, to allow readers to fully appreciate what was done, and the comparison of multisensory interactions between awake and anesthetized mice would benefit from being explored in more detail.

Reviewer #1 (Public Review):

Summary:<br /> In this paper, the authors provide a characterisation of auditory responses (tones, noise, and amplitude-modulated sounds) and bimodal (somatosensory-auditory) responses and interactions in the higher-order lateral cortex (LC) of the inferior colliculus (IC) and compare these characteristics with the higher order dorsal cortex (DC) of the IC - in awake and anaesthetised mice. Dan Llano's group has previously identified gaba'ergic patches (modules) in the LC distinctly receiving inputs from somatosensory structures, surrounded by matrix regions receiving inputs from the auditory cortex. They here use 2P calcium imaging combined with an implanted prism to - for the first time - get functional optical access to these subregions (modules and matrix) in the lateral cortex of IC in vivo, in order to also characterise the functional difference in these subparts of LC. They find that both DC and LC of both awake and anaesthetised mice appear to be more responsive to more complex sounds (amplitude-modulated noise) compared to pure tones and that under anesthesia the matrix of LC is more modulated by specific frequency and temporal content compared to the gabaergic modules in LC. However, while both LC and DC appear to have low-frequency preferences, this preference for low frequencies is more pronounced in DC. Furthermore, in both awake and anesthetized mice, somatosensory inputs are capable of driving responses on their own in the modules of LC, but very little (possibly not at all) in the matrix. However, bimodal interactions may be different under awake and anesthesia in LC, which warrants deeper investigation by the authors: They find, under anesthesia, more bimodal enhancement in modules of LC compared to the matrix of LC and bimodal suppression dominating the matrix of LC. In contrast, under awake conditions bimodal enhancement is almost exclusively found in the matrix of LC, and bimodal suppression dominates both matrix and modules of LC.

The paper provides new information about how subregions with different inputs and neurochemical profiles in the higher-order auditory midbrain process auditory and multisensory information, and is useful for the auditory and multisensory circuits neuroscience community.

Strengths:<br /> The major strength of this study is undoubtedly the fact that the authors for the first time provide optical access to a subcortical region (the lateral cortex of the inferior colliculus (i.e. higher order auditory midbrain)) which we know (from previous work by the same group) have optically identifiable subdivisions with unique inputs and neurotransmitter release, and plays a central role in auditory and multisensory processing. A description of basic auditory and multisensory properties of this structure is therefore very useful for understanding auditory processing and multisensory interactions in subcortical circuits.

Weaknesses:<br /> I have divided my comments about weaknesses and improvements into major and minor comments. All of which I believe are addressable by the reviewers to provide a more clear picture of their characterisation of the higher-order auditory midbrain.

Major comment:<br /> 1. The differences between multisensory interactions in LC in anaesthetised and awake preparations appear to be qualitatively different, though the authors claim they are similar (see also minor comment related to figure 10H for further explanation of what I mean). However, the findings in awake and anaesthetised conditions are summarised differently, and plotting of similar findings in the awake figures and anaesthetised figures are different - and different statistics are used for the same comparisons. This makes it very difficult to assess how multisensory integration in LC is different under awake and anaesthetised conditions. I suggest that the authors plot (and test with similar statistics) the summary plots in Figure 8 (i.e. Figure 8H-K) for awake data in Figure 10, and also make similar plots to Figures 10G-H for anaesthetised data. This will help the readers understand the differences between bimodal stimulation effects on awake and anaesthetised preparations - which in its current form, looks very distinct. In general, it is unclear to me why the awake data related to Figures 9 and 10 is presented in a different way for similar comparisons. Please streamline the presentation of results for anaesthetised and awake results to aid the comparison of results in different states, and explicitly state and discuss differences under awake and anaesthetised conditions.

2. The claim about the degree of tonotopy in LC and DC should be aided by summary statistics to understand the degree to which tonotopy is actually present. For example, the authors could demonstrate that it is not possible/or is possible to predict above chance a cell's BF based on the group of other cells in the area. This will help understand to what degree the tonotopy is topographic vs salt and pepper. Also, it would be good to know if the gaba'ergic modules have a higher propensity of particular BFs or tonotopic structure compared to matrix regions in LC, and also if general tuning properties (e.g. tuning width) are different from the matrix cells and the ones in DC.

3. Throughout the paper more information needs to be given about the number of cells, sessions, and animals used in each panel, and what level was used as n in the statistical tests. For example, in Figure 4 I can't tell if the 4 mice shown for LC imaging are the only 4 mice imaged, and used in the Figure 4E summary or if these are just examples. In general, throughout the paper, it is currently not possible to assess how many cells, sessions, and animals the data shown comes from.

4. Throughout the paper, to better understand the summary maps and plots, it would be helpful to see example responses of the different components investigated. For example, given that module cells appear to have more auditory offset responses, it would be helpful to see what the bimodal, sound-only, and somatosensory responses look like in example cells in LC modules. This also goes for just general examples of what the responses to auditory and somatosensory inputs look like in DC vs LC. In general example plots of what the responses actually look like are needed to better understand what is being summarised.

Reviewer #2 (Public Review):

Summary:<br /> The study describes differences in responses to sounds and whisker deflections as well as combinations of these stimuli in different neurochemically defined subsections of the lateral and dorsal cortex of the inferior colliculus in anesthetised and awake mice.

Strengths:<br /> The main achievement of the work lies in obtaining the data in the first place as this required establishing and refining a challenging surgical procedure to insert a prism that enabled the authors to visualise the lateral surface of the inferior colliculus. Using this approach, the authors were then able to provide the first functional comparison of neural responses inside and outside of the GABA-rich modules of the lateral cortex. The strongest and most interesting aspects of the results, in my opinion, concern the interactions of auditory and somatosensory stimulation. For instance, the authors find that a) somatosensory-responses are strongest inside the modules and b) somatosensory-auditory suppression is stronger in the matrix than in the modules. This suggests that, while somatosensory inputs preferentially target the GABA-rich modules, they do not exclusively target GABAergic neurons within the modules (given that the authors record exclusively from excitatory neurons we wouldn't expect to see somatosensory responses if they targeted exclusively GABAergic neurons), and that the GABAergic neurons of the modules (consistent with previous work) preferentially impact neurons outside the modules, i.e. via long-range connections.

Weaknesses:<br /> While the findings are of interest to the subfield they have only rather limited implications beyond it. The writing is not as precise as it could be. Consequently, the manuscript is unclear in some places. For instance, the text is somewhat confusing as to whether there is a difference in the pattern (modules vs matrix) of somatosensory-auditory suppression between anesthetized and awake animals. Furthermore, there are aspects of the results which are potentially very interesting but have not been explored. For example, there is a remarkable degree of clustering of response properties evident in many of the maps included in the paper. Taking Figure 7 for instance, rather than a salt and pepper organization we can see auditory responsive neurons clumped together and non-responsive neurons clumped together and in the panels below we can see off-responsive neurons forming clusters (although it is not easy to make out the magenta dots against the black background). This degree of clustering seems much stronger than expected and deserves further attention.

Reviewer #3 (Public Review):

The lateral cortex of the inferior colliculus (LC) is a region of the auditory midbrain noted for receiving both auditory and somatosensory input. Anatomical studies have established that somatosensory input primarily impinges on "modular" regions of the LC, which are characterized by high densities of GABAergic neurons, while auditory input is more prominent in the "matrix" regions that surround the modules. However, how auditory and somatosensory stimuli shape activity, both individually and when combined, in the modular and matrix regions of the LC has remained unknown.

The major obstacle to progress has been the location of the LC on the lateral edge of the inferior colliculus where it cannot be accessed in vivo using conventional imaging approaches. The authors overcame this obstacle by developing methods to implant a microprism adjacent to the LC. By redirecting light from the lateral surface of the LC to the dorsal surface of the microprism, the microprism enabled two-photon imaging of the LC via a dorsal approach in anesthetized and awake mice. Then, by crossing GAD-67-GFP mice with Thy1-jRGECO1a mice, the authors showed that they could identify LC modules in vivo using GFP fluorescence while assessing neural responses to auditory, somatosensory, and multimodal stimuli using Ca2+ imaging. Critically, the authors also validated the accuracy of the microprism technique by directly comparing results obtained with a microprism to data collected using conventional imaging of the dorsal-most LC modules, which are directly visible on the dorsal IC surface, finding good correlations between the approaches.

Through this innovative combination of techniques, the authors found that matrix neurons were more sensitive to auditory stimuli than modular neurons, modular neurons were more sensitive to somatosensory stimuli than matrix neurons, and bimodal, auditory-somatosensory stimuli were more likely to suppress activity in matrix neurons and enhance activity in modular neurons. Interestingly, despite their higher sensitivity to somatosensory stimuli than matrix neurons, modular neurons in the anesthetized prep were far more responsive to auditory stimuli than somatosensory stimuli (albeit with a tendency to have offset responses to sounds). This suggests that modular neurons should not be thought of as primarily representing somatosensory input, but rather as being more prone to having their auditory responses modified by somatosensory input. However, this trend was reversed in the awake prep, where modular neurons became more responsive to somatosensory stimuli than auditory stimuli. Thus, to this reviewer, the most intriguing result of the present study is the dramatic extent to which neural responses in the LC changed in the awake preparation. While this is not entirely unexpected, the magnitude and stimulus specificity of the changes caused by anesthesia highlight the extent to which higher-level sensory processing is affected by anesthesia and strongly suggest that future studies of LC function should be conducted in awake animals.

Together, the results of this study expand our understanding of the functional roles of matrix and module neurons by showing that responses in LC subregions are more complicated than might have been expected based on anatomy alone. The development of the microprism technique for imaging the LC will be a boon to the field, finally enabling much-needed studies of LC function in vivo. The experiments were well-designed and well-controlled, and the limitations of two-photon imaging for tracking neural activity are acknowledged. Appropriate statistical tests were used. There are three main issues the authors should address, but otherwise, this study represents an important advance in the field.

1) Please address whether the Thy1 mouse evenly expresses jRGECO1a in all LC neurons. It is known that these mice express jRGECO1a in subsets of neurons in the cerebral cortex, and similar biases in the LC could have biased the results here.

2) I suggest adding a paragraph or two to the discussion to address the large differences observed between the anesthetized and awake preparations. For example, somatosensory responses in the modules increased dramatically from 14.4% in the anesthetized prep to 63.6% in the awake prep. At the same time, auditory responses decreased from 52.1% to 22%. (Numbers for anesthetized prep include auditory responses and somatosensory + auditory responses.). In addition, the tonotopy of the DC shifted in the awake condition. These are intriguing changes that are not entirely expected from the switch to an awake prep and therefore warrant discussion.

3) For somatosensory stimuli, the authors used whisker deflection, but based on the anatomy, this is presumably not the only somatosensory stimulus that affects LC. The authors could help readers place the present results in a broader context by discussing how other somatosensory stimuli might come into play. For example, might a larger percentage of modular neurons be activated by somatosensory stimuli if more diverse stimuli were used?

Reviewer #1 (Public Review):

This paper studies the effects of tACS on detection of silence gaps in an FM modulated noise stimulus. Both FM modulation of the sound and the tACS are at 2Hz, and the phase of the two is varied to determine possible interactions between the auditory and electric stimulation. Additionally, two different electrode montages are used to determine if variation in electric field distribution across the brain may be related to the effects of tACS on behavioral performance in individual subjects.

Major strengths and weaknesses of the methods and results.

The study appears to be well powered to detect modulation of behavioral performance with N=42 subjects. There is a clear and reproducible modulation of behavioral effects with the phase of the FM sound modulation. The study was also well designed and executed in terms of fMRI, current flow modeling, montage optimization targeting, and behavioral analysis. A particular merit of this study is to have repeated the sessions for most subjects in order to test repeat-reliability, which is so often missing in human experiments. The results and methods are generally well described and well conceived. The portion of the analysis related to behavior alone is excellent. The analysis of the tACS results are also generally well described, candidly highlighting how variable results are across subjects and sessions. The figures are all of high quality and clear. One weakness of the experimental design is that no effort was made to control for sensation effects. tACS at 2Hz causes prominent skin sensations which could have interacted with auditory perception and thus, detection performance.

The central claim is that tACS modulates behavioral detection performance across the 0.5s  cycle of stimulation. Statistical analysis with randomize relative phase (between audio and tACS) show that detection performance is modulated by tACS. Neither the relative phase or the strength of this effect reproduces across subjects or sessions, which makes the interpretation of these results difficult. These result could be of interest to investigators in the field of tACS.

The claim that the variation in the strength of the effect can be explained by variation of electric fields is not compelling.

The following are more detailed comments to specific sections of the paper, including details on the concerns with the statistical analysis of the tACS effects.<br /> The introduction is well balanced, discussing the promise and limitations of previous results with tACS. The objectives are well defined.

The analysis surrounding behavioral performance and its dependence on phase of the FM modulation (Figure 3) is masterfully executed and explained. It appears that it reproduces previous studies and points to a very robust behavioral task that may be of use in other studies.

The definition of tACS(+) vs tACS(-) phase is adjusted to each subject/session, which seems unconventional.  For argument sake, let's assume the curves in Fig. 3E are random fluctuations. Then aligning them to best-fitting cosine will trivially generate a FM-amplitude fluctuation with cosine shape as shown in Fig. 4a. Selecting the positive and negative phase of that will trivially be larger and smaller than sham, respectively, as shown in Fig 4b.

"Data from the optimal tACS lag and its opposite lag (corresponding trough) were excluded to avoid any artificial bias in estimating tACS effects induced by the alignment procedure (33)." The delay was found by fitting a cosine, so removing just the peaks of that cosine does little to avoid this problem.

To demonstrate that this is not a trivial result of the definition, the analysis compares this to the same analysis but with a randomize alignment to the two stimuli (audio and tACS) in Figure 4d. Assuming this shuffle was done correctly, this shows that the modulation observed in 4b is not just a result of the analysis procedure.

The authors are to be commended for analyzing the robustness of their observation across subjects and across sessions in Fig. 5. The lack of consistency in the optimal time delay between the two stimuli is hard to reconcile with the common theory that tACS entrains brain function.

"To better understand what factors might be influencing inter-session variability in tACS effects, we estimated multiple linear models ..." "Inter-individual variability in the simulated E-field predicts tACS effects" Authors here are attempting to predict a property of the subjects that was just shown to not be a reliable property of the subject. Authors are picking 9 possible features for this, testing 33 possible models with N=34 data points. With these circumstances it is not hard to find something that correlates by chance. And some of the models tested had interaction terms, possibly further increasing the number of comparisons. In the absence of multiple comparison correction, what is happening here is that multiple models are fit to the data, and a statistical test is performed for the best model on the same (training) data. The corresponding claim that variations are explained by variations in electric field is not persuasive.

"Can we reduce inter-individual variability in tACS effects ..." This section seems even more speculative and with mixed results.

Given the concerns with the statistical analysis above, there are concerns about the following statements in the summary of the Discussion:

"4) individual variability in tACS effect size was partially explained by two interactions: between the normal component of the E-field and the field focality, and between the normal component of the E-field and the distance between the peak of the electric field and the functional target ROIs."

The complexity of this statement alone may be a good indication that this could be the result of false discovery due to multiple comparisons.

For the same reason as stated above, the following statements in the Abstract do not appear to have adequate support in the data:

"Inter-individual variability of tACS effects was best explained by the strength of the inward electric field, depending on the field focality and proximity to the target brain region. Although additional evidence is necessary, our results<br /> 42 also provided suggestive insights that spatially optimizing the electrode montage could be a promising tool to reduce inter-individual variability of tACS effects."

Reviewer #2 (Public Review):

I thank the authors for considering my comments and think the manuscript has been significantly improved with revision. However while I considered that the analysis employed for predicting tACS effects with linear models was convincing, I am still concerned by a multiple comparison issue for this analysis. An alternative option would be to report the results of a Partial Least Squares (PLS) analysis, with the stimulation properties as predictor variables and tACS effects as response variables. The authors could use PLS instead of multiple linear regression models to take into account the multicollinearity in the predictor variables, and also this can be done with only one PLS model. They could then extract the fitted responses values and estimate if the model can significantly fit the tACS effects.

Then, to determine which variables contribute more to the prediction, they can calculate the variable importance in projection (VIP) scores for the PLS regression model.<br /> An alternative option for the authors would be to temper their conclusions regarding how well field modeling/montage explains the variance observed across subjects.

Joint Public Review:

This manuscript tackles an important question, namely how K+ affects substrate transport in the SLC6 family. K+ effects have previously been reported for DAT and SERT, but the prototypical SLC6-fold transporter LeuT was not known to be sensitive to the K+ concentration. In this manuscript, the authors demonstrate convincingly that K+ inhibits Na+ binding, and Na+-dependent amino acid binding at high concentrations, and that K+ inside of vesicles containing LeuT increases the transport rate. However, outside K+ apparently had very little effect. Uptake data are supplemented with binding data, using the scintillation proximity assay, and transition metal FRET, allowing the observation of the distribution of distinct conformational states of the transporter.

Overall, the data are of high quality. I was initially concerned about the use of solutions of very high ionic strength (the Km for K+ is in the 200 mM range), however, the authors performed good controls with lower ionic strength solutions, suggesting that the K+ effect are specific and not caused by artifacts from the high salt concentrations.

eLife assessment

This study by Nandy and colleagues examined relationships between behavioral state, neural activity, and trial-by-trial variability in the ability to detect weak visual stimuli. They present useful findings indicating that certain changes in arousal and eye-position stability, along with patterns of synchrony in the activity of neurons in different layers of cortical area V4, can show modest correspondences to changes in the ability to correctly detect a stimulus. At present, however, the findings are based on data and analyses that are somewhat incomplete but could be improved with further revisions.

Reviewer #1 (Public Review):

Summary:<br /> In this study, Nandy and colleagues examine neural and behavioral correlates of perceptual variability in monkeys performing a visual change detection task. They used a laminar probe to record from area V4 while two macaque monkeys detected a small change in stimulus orientation that occurred at a random time in one of two locations, focusing their analysis on stimulus conditions where the animal was equally likely to detect (hit) or not-detect (miss) a briefly presented orientation change (target). They discovered two behavioral measures that are significantly different between hit and miss trials - pupil size tends to be slightly larger on hits vs. misses, and monkeys are more likely to miss the target on trials in which they made a microsaccade shortly before target onset. They also examined multiple measures of neural activity across the cortical layers and found some measures that are significantly different between hits and misses.

Strengths:<br /> Overall the study is well executed and the analyses are appropriate (though multiple issues do need to be addressed).

Weaknesses:<br /> My main concern with this study is that with the exception of the pre-target microsaccades, the physiological and behavioral correlates of perceptual variability (differences between hits and misses) appear to be very weak and disconnected. Some of these measures rely on complex analyses that are not hypothesis-driven and where statistical significance is difficult to assess. The more intuitive analysis of the predictive power of trial outcomes based on the behavioral and neural measures is only discussed at the end of the paper. This analysis shows that some of the significant measures have no predictive power, while others cannot be examined using the predictive power analysis because these measures cannot be estimated in single trials. Given these weak and disconnected effects, my overall sense is that the current results do not significantly advance our understanding of the neural basis of perceptual variability.

Reviewer #2 (Public Review):

In this manuscript, the authors conducted a study in which they measured eye movements, pupil diameter, and neural activity in V4 in monkeys engaged in a visual attention task. The task required the monkeys to report changes in the orientation of Gabors' visual stimuli. The authors manipulated the difficulty of the trials by varying the degree of orientation change and focused their analysis on trials of intermediate difficulty where the monkeys' hit rate was approximately 50%. Their key findings include the following: 1) Hit trials were preceded by larger pupil diameter, reflecting higher arousal, and by more stable eye positions; 2) V4 neurons exhibit larger visual responses in hit trials; 3) Superficial and deep layers exhibited greater coherence in hit trials during both the pre-target stimulus period and the non-target stimulus presentation period. These findings have useful implications for the field, and the experiments and analyses presented in this manuscript validly support the authors' claims.

Strengths:<br /> The experiments were well-designed and executed with meticulous control. The analyses of both behavioural and electrophysiological data align with the standards in the field.

Weaknesses:<br /> Many of the findings appear to be incremental compared to previous literature, including the authors' own work. While incremental findings are not necessarily a problem, the manuscript lacks clear statements about the extent to which the dataset, analysis, and findings overlap with the authors' prior research. For example, one of the main findings, which suggests that V4 neurons exhibit larger visual responses in hit trials (as shown in Fig. 3), appears to have been previously reported in their 2017 paper. Additionally, it seems that the entire Fig1-S1 may have been reused from the 2017 paper. These overlaps should have been explicitly acknowledged and correctly referenced.

Previous studies have demonstrated that attention leads to decorrelation in V4 population activity. The authors should have discussed how and why the high coherence across layers observed in the current study can coexist with this decorrelation.

Furthermore, the manuscript does not explore potentially interesting aspects of the dataset. For instance, the authors could have investigated instances where monkeys made 'false' reports, such as executing saccades towards visual stimuli when no orientation change occurred. It would be valuable to provide the fraction of the monkeys' responses in a session, including false reports and correct rejections in catch trials, to allow for a broader analysis that considers the perceptual component of neural activity over pure sensory responses.

eLife assessment

The study by Zhu et al. provides important insights into cell-specific genome-wide histone modifications in the frontal cortex of individuals with schizophrenia, as well as shedding light on the role of age and antipsychotic treatment in these associations. The evidence supporting the conclusions is solid, although more details regarding methodology would be helpful, and the integration of additional data could further enhance the novelty of the study.

Reviewer #1 (Public Review):

Zhu, et al present a genome-wide histone modification analysis comparing patients with schizophrenia (on or off antipsychotics) to non-psychiatric controls. The authors performed analyses across the dorsolateral prefrontal cortex and tested for enrichment of nearby genes and pathways. The authors performed an analysis measuring the effect of age on the epigenomic landscape as well. While this paper provides a unique resource around SCZ and its epigenetic correlates, and some potentially intriguing findings in the antipsychotic response dataset there were some potential missed opportunities - related to the integration of outside datasets and genotypes that could have strengthened the results and novelty of the paper.

Major Comments

1. Is there genotype data available for this cohort of donors or can it be generated? This would open several novel avenues of investigation for the authors. First the authors can test for enrichment of heritability for SCZ or even highly comorbid disorders such as bipolar. Second, it would allow the authors to directly measure the genetic regulation of histone markers by calculating QTLs (in this case histone hQTLs). The authors assert that although interesting, ATAC-seq approach does not provide the same chromatin state information as histone mods mapped by ChiP. Why do the authors not test this? There are several ATAC-seq datasets available for SCZ [https://pubmed.ncbi.nlm.nih.gov/30087329/]and an additional genomic overlap could help tease apart genetic regulation of the changes observed.

2. Can the authors theorize why their analysis found significant effects for H3K27Ac for antipsychotic use when a recent epigenomic study of SCZ using a larger cohort of samples and including the same histone modifications did not [https://pubmed.ncbi.nlm.nih.gov/30038276/]? Given the lower n and lower number of cells in this group, it would be helpful if the authors could speculate on why they see this. Do the authors know if there is any overlap with the Girdhar study donors or if there are other phenotypic differences that could account for this?

3. The reviewer is concerned about the low concordance between bulk nuclei RNA-seq and single-cell RNA-seq for SCZ (236 of 802 DEGs in NeuN+ and 63 of 1043 NEuN-). While it is not surprising for different cohorts to have different sets of DEGs these seem to be vastly different. Was there a particular cell type(s) that enriched for the authors' DEGs in the single-cell dataset? Do the authors know if any donors overlapped between these cohorts?

4. Functional enrichment analyses: details are not provided by the authors and should be added. The authors need to consider a) providing a gene universe, ie only considering the sets of genes with nearby H3K4me3/ H3K27ac levels, to such pathway tools, and b) should take into account the fact that some genes have many more peaks with data. There are known biases in seemingly just using the best p-value per gene in other epigenetic analysis (ie. DNA methylation data) and software is available to run correct analyses: https://pubmed.ncbi.nlm.nih.gov/23732277.

Reviewer #2 (Public Review):

The manuscript by Zhu has generated ChIP-seq and RNA-seq data from sizeable cohorts of SCZ patient samples and controls. The samples include 15 AF-SCZ samples and 15 controls, as well as 14 AT-SCZ samples and 14 controls. The genomics data was generated using techniques optimized for low-input samples: MOWChIP-seq and SMART-seq2 for histone profiles and transcriptome, respectively. The study has generated a significant data resource for the investigation of epigenomic alterations in SCZ. I am not convinced that the hierarchical pairwise design - first comparing AF-SCZ and AT-SCZ with their corresponding controls and secondarily contrasting the two comparisons is fully justified. The authors should repeat the statistical analysis by modeling all three groups simultaneously with an interaction effect for treatment or directly compare AF-SCZ to AT-SCZ groups and evaluate if the main conclusions remain supported.

Major comments

1. The manuscript did not discuss (mention) the quality control of RNA-seq data shown in Fig. 1B. The color scheme choice for the heatmap visualization did not provide a quantitative presentation of the specificity of the RNA-seq data. I would recommend using bar plots to present the results more quantitatively.

2. How does the specificity of this RNA-seq dataset compare to previous studies using a similar NeuN sorting strategy?

3. I appreciate the effort to assess the ChIP-seq data quality using phantompeakqualtools. However, prior knowledge/experience with this tool is required to fully understand the QC results. The authors should additionally provide browser shots at different scales for key neuronal/glial genes, so readers can have a more direct assessment of data quality, such as the enrichment of H3K4me3 at promoters (but not elsewhere), and H3K27ac at promoters and enhancers. Existing browser views, such as Fig. 2B are too zoomed out for assessing the data quality.

4. The pairwise regression model should be explicitly reported in methods.

5. The statistical strategy to compare AF-SCZ and AT-SCZ to their corresponding control groups was unjustified. Why not model all three groups simultaneously with an interaction effect for treatment or directly compare AF-SCZ to AT-SCZ groups? If the manuscript argues that the antipsychotic effect is the main novelty, why not directly compare AF-SCZ and AT-SCZ?

6. The method of pairwise comparison to corresponding control groups, then further comparing the pairwise results opens the study to a number of statistical vulnerabilities. For example, on page 12, the studies identified 166 DEGs between AF and control, and 1273 DEGs between AT and control. Instead of implicating a greater amount of difference between AT and control, such a result can often be driven by differences in between-group variance, rather than between-group means, that is, are the SCZ-AF and SCZ-treated effect size magnitudes and directionalities similar (but the treated group has lower variance) or are the two groups truly different in terms of means? The result in Fig. 5A suggests effect sizes for the two comparisons (AF-Ctrl and AT-Ctrl) are similar but have lower variability in the treated group.

7. The pairwise comparison further raised the possibility the results were driven by the difference in the two control cohorts rather than the two SCZ cohorts.

eLife assessment

This important study reports the identification of a new amino acid sequence motif (i.e., "internal beta-signal") on outer membrane proteins, which is recognized by beta-assembly machinery in gram-negative bacteria. The authors carried out rigorous experiments, providing compelling evidence in support of their conclusions. This work significantly advances our understanding of the biogenesis of outer membrane proteins.

Joint Public Review:

The biogenesis of outer membrane proteins (OMPs) into the outer membranes of Gram-negative bacteria is not fully understood, particularly client recognition and insertion by the conserved beta-assembly machinery (BAM), which is itself integrated in the outer membranes. So far, the last strand of an OMP, referred to as the beta-signal, has been known as a primary recognition motif by BAM. Here, authors have identified additional sequence motifs that are located in the upstream of the last strand.

Here, authors carried out rigorous biochemical, biophysical, and genetic approaches to prove that the newly identified internal motifs are critical to the assembly of outer membrane proteins as well as to the interaction with the BAM complex. The identification of important regions on the substrates and Bam proteins during biogenesis is an important contribution that gives clues to the path substrates take en route to the membrane. Assessing the effect of the internal motifs in the assembly of model OMPs in the absence (in vitro) and presence (in vitro and in vivo) of BAM machinery aids a precise definition of the role of the motifs, solidifying the conclusions.

The initial reviews raised several concerns:

1. Strengthening the claim that the recognition of the internal signal by BAM is mediated by BamA and BamD via specific interactions.

2. Justification of the rationale of the peptide inhibition assays as a primary tool to identify significant recognition motifs.

3. More careful interpretation of the mutational effects on OMP assembly - that is, discerning the impairment of BAM-nascent polypeptide chain interaction from the impairment of intrinsic folding.

4. Providing further clarification of the discrepancy between in vitro assay and in vivo assay regarding the effect of the mutation Y286A on OMP assembly.

5. More elaboration on the rationale, interpretation, and representation of neutron refractory data.

6. An explanation is lacking why the strain with BamD R197A does not display VCN sensitivity in contrast to the strain with BamD Y62A.

Those concerns were well addressed in the revised manuscript in a rigorous manner.

Overall, this study comprehensively addresses an important question in the field. The notion that additional signals assist in biogenesis is a novel concept that has been tested and verified at least for a subset of model OMPs in this study. The generalization of the conclusion awaits a further proof of the concept.

eLife assessment

This important study examines the effects of prenatal alcohol exposure and maternal diet on offspring DNA methylation, revealing that alcohol can alter epigenetic patterns and impact brain and organ development in the fetus, with some changes preventable by a diet rich in folate and choline. The work identifies several differentially methylated regions linked to adverse health outcomes from alcohol exposure, but the evidence is somewhat incomplete, as the paper currently lacks comprehensive methodological details and sensitivity analyses. Further analysis of the functional relevance of these DNA methylation changes, particularly addressing the current technical and statistical shortcomings, would increase the study's novelty and significance.

Reviewer #1 (Public Review):

Summary:<br /> This manuscript examined the impact of prenatal alcohol exposure on genome-wide DNA methylation in the brain and liver, comparing ethanol-exposed mice to unexposed controls. They also investigated whether a high-methyl diet (HMD) could prevent the DNA methylation alterations caused by alcohol. Using bisulfite sequencing (n=4 per group), they identified 78 alcohol-associated differentially methylated regions (DMRs) in the brain and 759 DMRs in the liver, of which 85% and 84% were mitigated by the HMD group, respectively. The authors further validated 7 DMRs in humans using previously published data from a Canadian cohort of children with FASD.

Overall, the findings from this study provide new insight into the impact of prenatal alcohol exposure, while also showing evidence for methyl-rich diets as an intervention to prevent the effects of alcohol on the epigenome. However, several methodological concerns limit the robustness of these results and should be addressed to further strengthen the conclusions of this study and its applicability to broader settings.

Strengths:<br /> - The use of whole genome bisulfite sequencing allowed for the interrogation of the entire DNA methylome and DMR analysis, rather than a subset of CpGs.<br /> - The combination of data from animal models and humans allowed the authors to make stronger inferences regarding their findings.<br /> - The authors investigated a potential mechanism (high methyl diet) to buffer against the effects of prenatal alcohol exposure, which increases the relevance and applicability of this research.

Weaknesses:<br /> - Mistakes and discontinuities in the reporting of results and methods made the manuscript difficult to follow. There was also some overuse of causal language and overinterpretation of differences.<br /> - The authors provide insufficient details to replicate their analyses, particularly for data quality control steps and statistical analyses.<br /> - The sample size was very small for the epigenetic analyses, which limits the robustness of the findings. This limitation is further emphasized by the cutoffs used to identify DMRs, which did not include multiple test corrections and used a delta cutoff that was not supported by the sequencing depth.<br /> - The authors do not account for potential confounders in their analyses, including birthweight, alcohol levels, and sex. This is a particular problem for the high-methyl diet analyses, in which the alcohol-exposed mice seemed to consume less alcohol than their non-diet counterparts.

Reviewer #2 (Public Review):

Summary:<br /> Bestry et al. investigated the effects of prenatal alcohol exposure (PAE) and high methyl donor diet (HMD) on offspring DNA methylation and behavioral outcomes using a mouse model that mimics common patterns of alcohol consumption in pregnancy in humans. The researchers employed whole-genome bisulfite sequencing (WGBS) for unbiased assessment of the epigenome in the newborn brain and liver, two organs affected by ethanol, to explore tissue-specific effects and to determine any "tissue-agnostic" effects that may have arisen prior to the germ-layer commitment during early gastrulation. The authors found that PAE induces measurable changes in offspring DNA methylation. DNA methylation changes induced by PAE coincide with non-coding regions, including enhancers and promoters, with the potential to regulate gene expression. Though the majority of the alcohol-sensitive differentially methylated regions (DMRs) were not conserved in humans, the ones that were conserved were associated with clinically relevant traits such as facial morphology, educational attainment, intelligence, autism, and schizophrenia. Finally, the study provides evidence that maternal dietary support with methyl donors alleviates the effects of PAE on DNA methylation, suggesting a potential prenatal care option.

Strengths:<br /> The strengths of the study include the use of a mouse model where confounding factors such as genetic background and diet can be well controlled. The study performed whole-genome bisulfite sequencing which allows a comprehensive analysis of the effects of PAE on DNA methylation. However, some weaknesses and limitations of the study are detected.

Weaknesses:<br /> 1. The low generalizability between mouse and human data alerts the validity of the mouse model designed in the study. On the same note, the authors failed to detect any significant effect on PAE-induced behavioral outcomes. I recognize that it is difficult to model all possible conditions of PAE in mice because the amount, frequency, and duration of alcohol consumption in humans vary significantly. Therefore, if the authors only focus on moderate PAE, it should be emphasized in the title and throughout the paper to avoid misinterpretation. In addition, is it possible to stratify the human data based on the level of PAE and compare it to the mouse data?<br /> 2. A major finding of the study is that PAE affects non-coding genomic regions in mice including enhancers and promoters. To improve the significance of the study, the authors need to back up this finding with transcriptome analysis and determine if these DMRs indeed affect gene expression.<br /> 3. The low generalizability between mouse and human data suggests that the regions affected by PAE may be species-specific. It is critical to analyze if PAE-induced DMRs in humans are also enriched in non-coding genomic regions. Considering the huge difference between mouse and human development, particularly in the brain, it is not surprising that different genomic loci are affected, but the affected loci may share similar features.<br /> 4. The specific brain regions and the lobes of the liver where the samples were taken should be clearly indicated.<br /> 5. I don't fully agree with the authors' interpretation that the two shared genomic regions affected in the brain and the liver "must have arisen before the germ layers separated". To claim so, the authors need to exclude the possibility that the two regions are just a coincidence due to the stochastic effect of PAE on DNA methylation.

eLife assessment

This useful study seeks to address the importance of physical interaction between proteins in higher-order complexes for covariation of evolutionary rates at different sites in these interacting proteins. Following up on a previous analysis with a smaller dataset, the authors provide solid evidence that the exact contribution of physical interactions, if any, remains difficult to quantify. A weakness of the study is that alternative hypotheses, specifically the importance of similar expression levels and patterns of genes that encode interacting proteins -- for which there is already substantial evidence in the literature -- are not sufficiently considered. The work will be of relevance to anyone interested in protein evolution.

Reviewer #1 (Public Review):

Summary:<br /> The manuscript titled "Coevolution due to physical interactions is not a major driving force behind evolutionary rate covariation" by Little et al., explores the potential contribution of physical interaction between correlated evolutionary rates among gene pairs. The authors find that physical interaction is not the main driving of evolutionary rate covariation (ECR). This finding is similar to a previous report by Clark et al. (2012), Genome Research, wherein the authors stated that "direct physical interaction is not required to produce ERC." The previous study used 18 Saccharomycotina yeast species, whereas the present study used 332 Saccharomycotina yeast species and 11 outgroup taxa. As a result, the present study is better positioned to evaluate the interplay between physical interaction and ECR more robustly.

Strengths & Weaknesses:<br /> Various analyses nicely support the authors' claims. Accordingly, I have only one significant comment and several minor comments that focus on wordsmithing - e.g., clarifying the interpretation of statistical results and requesting additional citations to support claims in the introduction.

Reviewer #2 (Public Review):

Summary:<br /> The authors address an important outstanding question: what forces are the primary drivers of evolutionary rate covariation? Exploration of this topic is important because it is currently difficult to interpret the functional/mechanistic implications of evolutionary covariation. These analyses also speak to the predictive power (and limits) of evolutionary rate covariation. This study reinforces the existing paradigm that covariation is driven by a varied/mixed set of interaction types that all fall under the umbrella explanation of 'co-functional interactions'.

Strengths:<br /> Very smart experimental design that leverages individual protein domains for increased resolution.

Weaknesses:<br /> Nuanced and sometimes inconclusive results that are difficult to capture in a short title/abstract statement.

Reviewer #3 (Public Review):

Summary:<br /> The paper makes a convincing argument that physical interactions of proteins do not cause substantial evolutionary co-variation.

Strengths:<br /> The presented analyses are reasonable and look correct and the conclusions make sense.

Weaknesses:<br /> The overall problem of the analysis is that nobody who has followed the literature on evolutionary rate variation over the last 20 years would think that physical interactions are a major cause of evolutionary rate variation. First, there have been probably hundreds of studies showing that gene expression level is the primary driver of evolutionary rate variation (see, for example, [1]). The present study doesn't mention this once. People can argue the causes or the strength of the effect, but entirely ignoring this body of literature is a serious lack of scholarship. Second, interacting proteins will likely be co-expressed, so the obvious null hypothesis would be to ask whether their observed rates are higher or lower than expected given their respective gene expression levels. Third, protein-protein interfaces exert a relatively weak selection pressure so I wouldn't expect them to play much role in the overall evolutionary rate of a protein.

On point 3, the authors seem confused though, as they claim a co-evolving interface would evolve *faster* than the rest of the protein (Figure 1, caption). Instead, the observation is they evolve slower (see, for example, [2]). This makes sense: A binding interface adds additional constraint that reduces the rate at which mutations accumulate. However, the effect is rather weak.

All in all, I'm fine with the analysis the authors perform, and I think the conclusions make sense, but the authors have to put some serious effort into reading the relevant literature and then reassess whether they are actually asking a meaningful question and, if so, whether they're doing the best analysis they could do or whether alternative hypotheses or analyses would make more sense.

[1] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4523088/<br /> [2] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4854464/

eLife assessment

This important study identifies the mitotic localization mechanism for Aurora B and INCENP (parts of the chromosomal passenger complex, CPC) in Trypanosoma brucei. The mechanism is different from that in the more commonly studied opisthokonts and there is solid support from RNAi and imaging experiments, targeted mutations, immunoprecipitations with crosslinking/mass spec, and AlphaFold interaction predictions. The results could be strengthened by biochemically testing proposed direct interactions and demonstrating that the targeting protein KIN-A is a motor. The findings will be of interest to parasitology researchers as well as cell biologists working on mitosis and cell division, and those interested in the evolution of the CPC.

Reviewer #1 (Public Review):

Summary:<br /> The CPC plays multiple essential roles in mitosis such as kinetochore-microtubule attachment regulation, kinetochore assembly, spindle assembly checkpoint activation, anaphase spindle stabilization, cytokinesis, and nuclear envelope formation, as it dynamically changes its mitotic localization: it is enriched at inner centromeres from prophase to metaphase but it is relocalized at the spindle midzone in anaphase. The business end of the CPC is Aurora B and its allosteric activation module IN-box, which is located at the C-terminal part of INCENP. In most well-studied eukaryotic species, Aurora B activity is locally controlled by the localization module of the CPC, Survivin, Borealin, and the N-terminal portion of INCENP. Survivin and Borealin, which bind the N terminus of INCENP, recognize histone residues that are specifically phosphorylated in mitosis, while anaphase spindle midzone localization is supported by the direct microtubule-binding capacity of the SAH (single alpha helix) domain of INCENP and other microtubule-binding proteins that specifically interact with INCENP during anaphase, which are under the regulation of CDK activity. One of these examples includes the kinesin-like protein MKLP2 in vertebrates.

Trypanosoma is an evolutionarily interesting species to study mitosis since its kinetochore and centromere proteins do not show any similarity to other major branches of eukaryotes, while orthologs of Aurora B and INCENP have been identified. Combining molecular genetics, imaging, biochemistry, cross-linking IP-MS (IP-CLMS), and structural modeling, this manuscript reveals that two orphan kinesin-like proteins KIN-A and KIN-B act as localization modules of the CPC in Trypanosoma brucei. The IP-CLMS, AlphaFold2 structural predictions, and domain deletion analysis support the idea that (1) KIN-A and KIN-B form a heterodimer via their coiled-coil domain, (2) Two alpha helices of INCENP interact with the coiled-coil of the KIN-A-KIN-B heterodimer, (3) the conserved KIN-A C-terminal CD1 interacts with the heterodimeric KKT9-KKT11 complex, which is a submodule of the KKT7-KKT8 kinetochore complex unique to Trypanosoma, (4) KIN-A and KIN-B coiled-coil domains and the KKT7-KKT8 complex are required for CPC localization at the centromere, (5) CD1 and CD2 domains of KIN-A support its centromere localization. The authors further show that the ATPase activity of KIN-A is critical for spindle midzone enrichment of the CPC. The imaging data of the KIN-A rigor mutant suggest that dynamic KIN-A-microtubule interaction is required for metaphase alignment of the kinetochores and proliferation. Overall, the study reveals novel pathways of CPC localization regulation via KIN-A and KIN-B by multiple complementary approaches.

Strengths:<br /> The major conclusion is collectively supported by multiple approaches, combining site-specific genome engineering, epistasis analysis of cellular localization, AlphaFold2 structure prediction of protein complexes, IP-CLMS, and biochemical reconstitution (the complex of KKT8, KKT9, KKT11, and KKT12).

Weaknesses:<br /> - The predictions of direct interactions (e.g. INCENP with KIN-A/KIN-B, or KIN-A with KKT9-KKT11) have not yet been confirmed experimentally, e.g. by domain mutagenesis and interaction studies.

- The criteria used to judge a failure of localization are not clearly explained (e.g., Figure 5F, G).

- It remains to be shown that KIN-A has motor activity.

- The authors imply that KIN-A, but not KIN-B, interacts with microtubules based on microtubule pelleting assay (Fig. S6), but the substantial insoluble fractions of 6HIS-KINA and 6HIS-KIN-B make it difficult to conclusively interpret the data. It is possible that these two proteins are not stable unless they form a heterodimer.

- For broader context, some prior findings should be introduced, e.g. on the importance of the microtubule-binding capacity of the INCENP SAH domain and its regulation by mitotic phosphorylation (PMID 8408220, 26175154, 26166576, 28314740, 28314741, 21727193), since KIN-A and KIN-B may substitute for the function of the SAH domain.

Reviewer #2 (Public Review):

How the chromosomal passenger complex (CPC) and its subunit Aurora B kinase regulate kinetochore-microtubule attachment, and how the CPC relocates from kinetochores to the spindle midzone as a cell transitions from metaphase to anaphase are questions of great interest. In this study, Ballmer and Akiyoshi take a deep dive into the CPC in T. brucei, a kinetoplastid parasite with a kinetochore composition that varies greatly from other organisms.

Using a combination of approaches, most importantly in silico protein predictions using alphafold multimer and light microscopy in dividing T. brucei, the authors convincingly present and analyse the composition of the T. brucei CPC. This includes the identification of KIN-A and KIN-B, proteins of the kinesin family, as targeting subunits of the CPC. This is a clear advancement over earlier work, for example by Li and colleagues in 2008. The involvement of KIN-A and KIN-B is of particular interest, as it provides a clue for the (re)localization of the CPC during the cell cycle. The evolutionary perspective makes the paper potentially interesting for a wide audience of cell biologists, a point that the authors bring across properly in the title, the abstract, and their discussion.

The evolutionary twist of the paper would be strengthened 'experimentally' by predictions of the structure of the CPC beyond T. brucei. Depending on how far the authors can extend their in-silico analysis, it would be of interest to discuss a) available/predicted CPC structures in well-studied organisms and b) structural predictions in other euglenozoa. What are the general structural properties of the CPC (e.g. flexible linkers, overall dimensions, structural differences when subunits are missing etc.)? How common is the involvement of kinesin-like proteins? In line with this, it would be good to display the figure currently shown as S1D (or similar) as a main panel.

Reviewer #3 (Public Review):

Summary:<br /> The protein kinase, Aurora B, is a critical regulator of mitosis and cytokinesis in eukaryotes, exhibiting a dynamic localisation. As part of the Chromosomal Passenger Complex (CPC), along with the Aurora B activator, INCENP, and the CPC localisation module comprised of Borealin and Survivin, Aurora B travels from the kinetochores at metaphase to the spindle midzone at anaphase, which ensures its substrates are phosphorylated in a time- and space-dependent manner. In the kinetoplastid parasite, T. brucei, the Aurora B orthologue (AUK1), along with an INCENP orthologue known as CPC1, and a kinetoplastid-specific protein CPC2, also displays a dynamic localisation, moving from the kinetochores at metaphase to the spindle midzone at anaphase, to the anterior end of the newly synthesised flagellum attachment zone (FAZ) at cytokinesis. However, the trypanosome CPC lacks orthologues of Borealin and Survivin, and T. brucei kinetochores also have a unique composition, being comprised of dozens of kinetoplastid-specific proteins (KKTs). Of particular importance for this study are KKT7 and the KKT8 complex (comprising KKT8, KKT9, KKT11, and KKT12). Here, Ballmer and Akiyoshi seek to understand how the CPC assembles and is targeted to its different locations during the cell cycle in T. brucei.

Strengths & Weaknesses:<br /> Using immunoprecipitation and mass-spectrometry approaches, Ballmer and Akiyoshi show that AUK1, CPC1, and CPC2 associate with two orphan kinesins, KIN-A and KIN-B, and with the use of endogenously expressed fluorescent fusion proteins, demonstrate for the first time that KIN-A and KIN-B display a dynamic localisation pattern similar to other components of the CPC. Most of these data provide convincing evidence for KIN-A and KIN-B being bona fide CPC proteins, although the evidence that KIN-A and KIN-B translocate to the anterior end of the new FAZ at cytokinesis is weak - the KIN-A/B signals are very faint and difficult to see, and cell outlines/brightfield images are not presented to allow the reader to determine the cellular location of these faint signals (Fig S1B).

They then demonstrate, by using RNAi to deplete individual components, that the CPC proteins have hierarchical interdependencies for their localisation to the kinetochores at metaphase. These experiments appear to have been well performed, although only images of cell nuclei were shown (Fig 2A), meaning that the reader cannot properly assess whether CPC components have localised elsewhere in the cell, or if their abundance changes in response to depletion of another CPC protein.

Ballmer and Akiyoshi then go on to determine the kinetochore localisation domains of KIN-A and KIN-B. Using ectopically expressed GFP-tagged truncations, they show that coiled-coil domains within KIN-A and KIN-B, as well as a disordered C-terminal tail present only in KIN-A, but not the N-terminal motor domains of KIN-A or KIN-B, are required for kinetochore localisation. These data are strengthened by immunoprecipitating CPC complexes and crosslinking them prior to mass spectrometry analysis (IP-CLMS), a state-of-the-art approach, to determine the contacts between the CPC components. Structural predictions of the CPC structure are also made using AlphaFold2, suggesting that coiled coils form between KIN-A and KIN-B, and that KIN-A/B interact with the N termini of CPC1 and CPC2. Experimental results show that CPC1 and CPC2 are unable to localise to kinetochores if they lack their N-terminal domains consistent with these predictions. Altogether these data provide convincing evidence of the protein domains required for CPC kinetochore localisation and CPC protein interactions. However, the authors also conclude that KIN-B plays a minor role in localising the CPC to kinetochores compared to KIN-A. This conclusion is not particularly compelling as it stems from the observation that ectopically expressed GFP-NLS-KIN-A (full length or coiled-coil domain + tail) is also present at kinetochores during anaphase unlike endogenously expressed YFP-KIN-A. Not only is this localisation probably an artifact of the ectopic expression, but the KIN-B coiled-coil domain localises to kinetochores from S to metaphase and Fig S2G appears to show a portion of the expressed KIN-B coiled-coil domain colocalising with KKT2 at anaphase. It is unclear why KIN-B has been discounted here.

Next, using a mixture of RNAi depletion and LacI-LacO recruitment experiments, the authors show that kinetochore proteins KKT7 and KKT9 are required for AUK1 to localise to kinetochores (other KKT8 complex components were not tested here) and that all components of the KKT8 complex are required for KIN-A kinetochore localisation. Further, both KKT7 and KKT8 were able to recruit AUK1 to an ectopic locus in the S phase, and KKT7 recruited KKT8 complex proteins, which the authors suggest indicates it is upstream of KKT8. However, while these experiments have been performed well, the reciprocal experiment to show that KKT8 complex proteins cannot recruit KKT7, which could have confirmed this hierarchy, does not appear to have been performed. Further, since the LacI fusion proteins used in these experiments were ectopically expressed, they were retained (artificially) at kinetochores into anaphase; KKT8 and KIN-A were both able to recruit AUK1 to LacO foci in anaphase, while KKT7 was not. The authors conclude that this suggests the KKT8 complex is the main kinetochore receptor of the CPC - while very plausible, this conclusion is based on a likely artifact of ectopic expression, and for that reason, should be interpreted with a degree of caution.

Further IP-CLMS experiments, in combination with recombinant protein pull-down assays and structural predictions, suggested that within the KKT8 complex, there are two subcomplexes of KKT8:KKT12 and KKT9:KKT11, and that KKT7 interacts with KKT9:KKT11 to recruit the remainder of the KKT8 complex. The authors also assess the interdependencies between KKT8 complex components for localisation and expression, showing that all four subunits are required for the assembly of a stable KKT8 complex and present AlphaFold2 structural modelling data to support the two subcomplex models. In general, these data are of high quality and convincing with a few exceptions. The recombinant pulldown assay (Fig. 4H) is not particularly convincing as the 3rd eluate gel appears to show a band at the size of KKT11 (despite the labelling indicating no KKT11 was present in the input) but no pulldown of KKT9, which was present in the input according to the figure legend (although this may be mislabeled since not consistent with the text). The text also states that 6HIS-KKT8 was insoluble in the absence of KKT12, but this is not possible to assess from the data presented. It is also surprising that data showing the effects of KKT8, KKT9, and KKT12 depletion on KKT11 localisation and abundance are not presented alongside the reciprocal experiments in Fig S4G-J.

The authors also convincingly show that AlphaFold2 predictions of interactions between KKT9:KKT11 and a conserved domain (CD1) in the C-terminal tail of KIN-A are likely correct, with CD1 and a second conserved domain, CD2, identified through sequence analysis, acting synergistically to promote KIN-A kinetochore localisation at metaphase, but not being required for KIN-A to move to the central spindle at anaphase. They then hypothesise that the kinesin motor domain of KIN-A (but not KIN-B which is predicted to be inactive based on non-conservation of residues key for activity) determines its central spindle localisation at anaphase through binding to microtubules. In support of this hypothesis, the authors show that KIN-A, but not KIN-B can bind microtubules in vitro and in vivo. However, ectopically expressed GFP-NLS fusions of full-length KIN-A or KIN-A motor domain did not localise to the central spindle at anaphase. The authors suggest this is due to the GPF fusion disrupting the ATPase activity of the motor domain, but they provide no evidence that this is the case. Instead, they replace endogenous KIN-A with a predicted ATPase-defective mutant (G209A), showing that while this still localises to kinetochores, the kinetochores were frequently misaligned at metaphase, and that it no longer concentrates at the central spindle (with concomitant mis-localisation of AUK1), causing cells to accumulate at anaphase. From these data, the authors conclude that KIN-A ATPase activity is required for chromosome congression to the metaphase plate and its central spindle localisation at anaphase. While potentially very interesting, these data are incomplete in the absence of any experimental data to show that KIN-A possesses ATPase activity or that this activity is abrogated by the G209A mutation, and the conclusions of this section are rather speculative.

Impact:<br /> Overall, this work uses a wide range of cutting-edge molecular and structural predictive tools to provide a significant amount of new and detailed molecular data that shed light on the composition of the unusual trypanosome CPC and how it is assembled and targeted to different cellular locations during cell division. Given the fundamental nature of this research, it will be of interest to many parasitology researchers as well as cell biologists more generally, especially those working on aspects of mitosis and cell division, and those interested in the evolution of the CPC.

Author Response

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

Response to Reviewers:

Thank you for taking the time to review our manuscript and provide us with helpful comments. Your comments enabled us to improve the clarity of the manuscript, in particular:

1. We improved the organization of the figures by associating each supplemental figure with a main-text figure using the eLife “figure supplements” format.

2. We reduced the length of figure captions where possible.

3. We improved organizational clarity by adding a brief organizational summary statement at the beginning of the results section which outlines the contents of the results subsections in the context of the introduction. We took particular care to use the same language, so the parallelism is clearer.

4. In addition, we made various modifications to the main text to improve clarity for the reader. For this we asked specific help of our biologist co-authors to indicate which aspects would benefit from further clarification to enable the broad biology readership of eLife to comprehend our research better.

Reviewer #1 (Public Review):

The authors sought to resolve the coordinated functions of the two muscles that primarily power flight in birds (supracoracoideus and pectoralis), with particular focus on the pectoralis. Technology has limited the ability to resolve some details of pectoralis function, so the authors developed a model that can make accurate predictions about this muscle's function during flight. The authors first measured aerodynamic forces, wing shape changes, and pectoralis muscle activity in flying doves. They used cutting-edge techniques for the aerodynamic and wing shape measurements and they used well-established methods to measure activity and length of the pectoralis muscle. The authors then developed two mathematical models to estimate the instantaneous force vector produced by the pectoralis throughout the wing stroke. Finally, the authors applied their mathematical models to other-sized birds in order to compare muscle physiology across species.

The strength of the methods is that they smoothly incorporate techniques from many complementary fields to generate a comprehensive model of pectoralis muscle function during flight. The high-speed structured-light technique for quantifying surface area during flight is novel and cutting-edge, as is the aerodynamic force platform used. These methods push the boundaries of what has historically been used to quantify their respective aspects of bird flight and their use here is exciting. The methods used for measuring muscle activation and length are standard in the field. Together, these provide both a strong conceptual foundation for the model and highlight its novelty. This model allows for estimations of muscle function that are not feasible to measure in live birds during flight at present. The weakness of this approach is that it relies heavily on a series of assumptions. While the research presented in this paper makes use of powerful methods from multiple fields, those methods each have assumptions inherent to them that simplify the biological system of study. This reduction in the complexity of phenomena allows the specific measurements to be made. In joining the techniques of multiple fields to study the greater complexity of the phenomenon of interest, the assumptions are all incorporated also. Furthermore, assumptions are inherent to mathematical modeling of biological phenomena. That being said, the authors acknowledge and justify their assumptions at each step and their model seems to be quite good at predicting muscle function.

Indeed, the authors achieve their aims. They effectively integrate methods from multiple disciplines to explore the coordination and function of the pectoralis and supracoracoideus muscles during flight. The conclusions that the authors derive from their model address the intended research aim.

The authors demonstrate the value of such interdisciplinary research, especially in studying complex behaviors that are difficult or infeasible to measure in living animals. Additionally, this work provides predictions for muscle function that can be tested empirically. These methods are certainly valuable for understanding flight but also have implications for biologists studying movement and muscle function more generally.

Thank you for your thorough and positive review. We appreciate that you read our manuscript carefully and gave detailed feedback.

Recommendations For The Authors:

I thought that your manuscript was very interesting and your integration of techniques from multiple fields was effective. You address the weaknesses I highlighted in the public review well throughout the manuscript.

Thank you for your well-measured feedback on this weakness and how we addressed it.

I sometimes found that the manuscript was difficult to follow. With the interdisciplinary nature of your work, your manuscript has a lot of complexity. Your introduction is clear and I think that the last paragraph outlines your study very well. In the subsequent sections, the sub-headings are helpful, but I think your manuscript could be improved by indicating where those subsections fit into the phases you outline in your introduction (namely, muscle function, kinematics and aerodynamics, and mathematical modeling).

Complied: throughout the manuscript we made modifications to improve the clarity. We also added a brief organizational summary statement at the beginning of the results section which outlines the contents of the results section in the context of the language introduced in the introduction. Finally, we reorganized the supplemental figures into eLife’s favored format of “figure supplements”, so that each extra figure is now associated with a figure in the main text. This should help the reader access information in an easier, hierarchical manner.

Reviewer #2 (Public Review):

In this work, the authors investigated the pectoralis work loop and the function of the supracoracoideus muscle in the down stroke during slow flight in doves. The aim of this study was to determine how aerodynamic force is generated, using simultaneous high-speed measurements of the wings' kinematics, aerodynamics, and activation and strain of pectoralis muscles during slow flight. The measurements show a reduction in the angle of attack during mid-downstroke, which induces a peak power factor and facilitates the tensioning of the supracoracoideus tendon with pectoralis power, which then can be released in the up-stroke. By combining the data with a muscle mechanics model, the timely tuning of elastic storage in the supracoracoideus tendon was examined and showed an improvement of the pectoralis work loop shape factor. Finally, other bird species were integrated into the model for a comparative investigation.

The major strength of the methods is the simultaneous application of four high-speed techniques - to quantify kinematics, aerodynamics and muscle activation and strain - as well as the implementation of the time-resolved data into a muscle mechanics model. With a thorough analysis which supports the conclusions convincingly, the authors achieved their goal of reaching an improved understanding of the interplay of the pectoralis and supracoracoideus muscles during slow flight and the resulting energetic benefits.

Thank you for your helpful and positive review. We appreciate that you summarized our manuscript accurately in a way that can help the reader.

Recommendations For The Authors:

The manuscript is very detailed and appears a bit long, including all the supplementary materials. It seems that the manuscript could easily have been separated into several publications, especially the comparative investigation including other extant bird species into the new model could have been a separate publication. This would have reduced the length of the supplements.

Thank you for your feedback on our manuscript; we made numerous improvements to improve the readability. Hence, we decided to not cut the supplement short or split it into more papers. We chose eLife because we wanted to publish this study in one complete manuscript. This has three benefits: (1) The reader can find all information in one well-edited paper at one publisher that is open-access and high-quality. (2) The first author works in industry and gets no benefits from publishing multiple papers, and hence he opted to publish one with support of the author team. (3) The senior author is not interested in fragmented publishing. Rather, he writes fewer, more comprehensive integrative papers because that is ultimately more informative for the reader: one trusted published source has all that is important to know based on this completed research project. Overall, we weren’t able to find technical information that shouldn't go in the paper using the lens of reproducibility, so the supplement is relatively long. Combining three methods (kinematics, forces, muscles), of which two are only available in the senior author’s lab, and extensive math (two new integrative models plus scaling laws) requires sharing the information needed for replication for all approaches we combine.

Also, some figure captions are very long and some of the content might have been included in the main text.

Complied: thank you for helping us streamline the captions. We reviewed all the figure captions and removed material that is repeated in the main text, but not essential to understanding the figures. However, because of the length of the manuscript and our desire to make the manuscript readable and clear, we left all other text in the captions intact so they remain readable independently of the main text. This way, the reader does not have to go searching for information in the main text just to make sense of the figures. This is especially important because readers often read the figures first before deciding if they want to read the main text completely. In addition, we moved two panels from Figure 2 into its associated figure supplement, because it was not a main point in the text, and hence this helped reduce the length of the caption in figure 2.

eLife assessment

This important study combines experiments and mathematical modelling to enhance our understanding of the interplay between the two flight muscles in birds during slow flight. The evidence for the findings is compelling, derived from new methods for measuring wing shape and force production combined with previously validated methods in muscle physiology. This work will be of broad interest to comparative biomechanists.

Reviewer #1 (Public Review):

The authors sought to resolve the coordinated functions of the two muscles that primarily power flight in birds (supracoracoideus and pectoralis), with particular focus on the pectoralis. Technology has limited the ability to resolve some details of pectoralis function, so the authors developed a model that can make accurate predictions about this muscle's function during flight. The authors first measured aerodynamic forces, wing shape changes, and pectoralis muscle activity in flying doves. They used cutting-edge techniques for the aerodynamic and wing shape measurements and they used well-established methods to measure activity and length of the pectoralis muscle. The authors then developed two mathematical models to estimate the instantaneous force vector produced by the pectoralis throughout the wing stroke. Finally, the authors applied their mathematical models to other-sized birds in order to compare muscle physiology across species.

The strength of the methods is that they smoothly incorporate techniques from many complementary fields to generate a comprehensive model of pectoralis muscle function during flight. The high-speed structured-light technique for quantifying surface area during flight is novel and cutting-edge, as is the aerodynamic force platform used. These methods push the boundaries of what has historically been used to quantify their respective aspects of bird flight and their use here is exciting. The methods used for measuring muscle activation and length are standard in the field. Together, these provide both a strong conceptual foundation for the model and highlight its novelty. This model allows for estimations of muscle function that are not feasible to measure in live birds during flight at present. The weakness of this approach is that it relies heavily on a series of assumptions. While the research presented in this paper makes use of powerful methods from multiple fields, those methods each have assumptions inherent to them that simplify the biological system of study. This reduction in the complexity of phenomena allows specific measurements to be made. In joining the techniques of multiple fields to study greater complexity of the phenomenon of interest, the assumptions are all incorporated also. Furthermore, assumptions are inherent to mathematical modelling of biological phenomena. That being said, the authors acknowledge and justify their assumptions at each step and their model seems to be quite good at predicting muscle function.

Indeed, the authors achieve their aims. They effectively integrate methods from multiple disciplines to explore the coordination and function of the pectoralis and supracoracoideus muscles during flight. The conclusions that the authors derive from their model address the intended research aim.

The authors demonstrate the value of such interdisciplinary research, especially in studying complex behaviors that are difficult or infeasible to measure in living animals. Additionally, this work provides predictions for muscle function that can be tested empirically. These methods are certainly valuable for understanding flight, but also have implications for biologists studying movement and muscle function more generally.

Reviewer #2 (Public Review):

In this work, the authors investigated the pectoralis work loop and the function of the supracoracoideus muscle in the down stroke during slow flight in doves. The aim of this study was to determine how aerodynamic force is generated, using simultaneous high-speed measurements of the wings' kinematics, aerodynamics, and activation and strain of pectoralis muscles during slow flight. The measurements show a reduction in the angle of attack during mid-downstroke, which induces a peak power factor and facilitates the tensioning of the supracoracoideus tendon with pectoralis power, which then can be released in the up-stroke. By combining the data with a muscle mechanics model, the timely tuning of elastic storage in the supracoracoideus tendon was examined and showed an improvement of the pectoralis work loop shape factor. Finally, other bird species were integrated into the model for a comparative investigation.

The major strength of the methods is the simultaneous application of four high-speed techniques - to quantify kinematics, aerodynamics and muscle activation and strain - as well as the implementation of the time-resolved data into a muscle mechanics model. With a thorough analysis which supports the conclusions convincingly, the authors achieved their goal of reaching an improved understanding of the interplay of the pectoralis and supracoracoideus muscles during slow flight and the resulting energetic benefits.

Author Response

The authors wish to thank the Reviewers for valuable and constructive comments that will help up improve the paper’s quality.

Public Reviews:

Reviewer #1 (Public Review):

Summary:

This manuscript builds upon the authors' previous work on the cross-talk between transcription initiation and post-transcriptional events in yeast gene expression. These prior studies identified an mRNA 'imprinting' phenomenon linked to genes activated by the Rap1 transcription factor (TF), a surprising role for the Sfp1 TF in promoting RNA polymerase II (RNAPII) backtracking, and a role for the non-essential RNAPII subunits Rpb4/7 in the regulation of mRNA decay and translation. Here the authors aimed to extend these observations to provide a more coherent picture of the role of Sfp1 in transcription initiation and subsequent steps in gene expression. They provide evidence for (1) a physical interaction between Sfp1 and Rpb4, (2) Sfp1 binding and stabilization of mRNAs derived from genes whose promoters are bound by both Rap1 and Sfp1 and (3) an effect of Sfp1 on Rpb4 binding or conformation during transcription elongation.

Strengths:

This study provides evidence that a TF (yeast Sfp1), in addition to stimulating transcription initiation, can at some target genes interact with their mRNA transcripts and promote their stability. Sfp1 thus has a positive effect on two distinct regulatory steps. Furthermore, evidence is presented indicating that strong Sfp1 mRNA association requires both Rap1 and Sfp1 promoter binding and is increased at a sequence motif near the polyA track of many target mRNAs. Finally, they provide compelling evidence that Sfp1-bound mRNAs have higher levels of RNAPII backtracking and altered Rpb4 association or conformation compared to those not bound by Sfp1.

Weaknesses:

The Sfp1-Rpb4 association is supported only by a two-hybrid assay that is poorly described and lacks an important control. Furthermore, there is no evidence that this interaction is direct, nor are the interaction domains on either protein identified (or mutated to address function).

Indeed, our two hybrid, immunoprecipitation and imaging results do not allow us to conclusively discern whether the interaction between Rpb4 and Sfp1 is direct or indirect. While the interaction holds significance, we consider the direct versus indirect distinction to be of secondary importance in the context of this paper. We intend to give more attention to this matter in our revised paper. In addition, we will make an effort to investigate an in vitro interaction between Sfp1 and Rpb4 by employing purified Sfp1 and Rpb4 proteins.

The contention that Sfp1 nuclear export to the cytoplasm is transcription-dependent is not well supported by the experiments shown, which are not properly described in the text and are not accompanied by any primary data.

We note that this assay has been developed and published in prior research by Lee, M. S., M. Henry, and P. A. Silver. (G&D, 1996) and was reported in a number of subsequent papers. Reassuringly, our conclusion is supported by the observation that Sfp1 binds to Pol II transcripts co-transcriptionally suggesting that Sfp1 is exported in the context of the mRNA.

The presence of Sfp1 in P-bodies is of unclear relevance and the authors do not ask whether Sfp1-bound mRNAs are also present in these condensates.

In the revised paper, we will indicate that we do not know whether RP mRNAs are present in the actual foci shown in Fig. 1B.

Further analysis of Sfp1-bound mRNAs would be of interest, particularly to address the question of whether those from ribosomal protein genes and other growth-related genes that are known to display Sfp1 binding in their promoters are regulated (either stabilized or destabilized) by Sfp1.

Fig. 4A, C and D show that RP mRNAs become destabilized in sfp1Δ cells.

The authors need to discuss, and ideally address, the apparent paradox that their previous findings showed that Rap1 acts to destabilize its downstream transcripts, i.e. that it has the opposite effect of Sfp1 shown here.

We would like to thank Reviewer 1 for this valuable comment. In the revised paper, we will delve into our hypothesis suggesting that Rap1 is likely responsible for regulating the imprinting of other proteins, that, in turn, lead to the destabilization of mRNAs, such as Rpb4.

Finally, recent studies indicate that the drugs used here to measure mRNA stability induce a strong stress response accompanied by rapid and complex effects on transcription. Their relevance to mRNA stability in unstressed cells is questionable.

Half-lives were determined mainly by the GRO analysis of optimally proliferating cells. This method does not requires any drug or stressful treatment. The results obtained by this method were consistent with the those obtained after thiolutin addition. Nevertheless, in our revised manuscript, we plan to supplement the half-life data with results obtained by subjecting cells to a temperature shift to 42°C, a natural method to block transcription in wild-type (WT) cells. This approach to determine half-lives has been previously reported in our publications, such as Lotan et al. (2005, 2007) and Goler Baron et al. (2008). This may rule out effects of the drug on halfe-life.

Reviewer #2 (Public Review):

Summary:

The manuscript by Kelbert et al. presents results on the involvement of the yeast transcription factor Sfp1 in the stabilisation of transcripts whose synthesis it stimulates. Sfp1 is known to affect the synthesis of a number of important cellular transcripts, such as many of those that code for ribosomal proteins. The hypothesis that a transcription factor can remain bound to the nascent transcript and affect its cytoplasmic half-life is attractive, but the methods used to demonstrate the half-life effects and the association of Sfp1 with cytoplasmic transcripts remain to be fully validated, as explained in my comments on the results below:

Comments on methodology and results:

1. A two-hybrid-based assay for protein-protein interactions identified Sfp1, a transcription factor known for its effects on ribosomal protein gene expression, as interacting with Rpb4, a subunit of RNA polymerase II. Classical two-hybrid experiments depend on the presence of the tested proteins in the nucleus of yeast cells, suggesting that the observed interaction occurs in the nucleus. Unfortunately, the two-hybrid method cannot determine whether the interaction is direct or mediated by nucleic acids.

Please see our response to comment 1 of Reviewer 1.

1. Inactivation of nup49, a component of the nuclear pore complex, resulted in the redistribution of GFP-Sfp1 into the cytoplasm at the temperature non-permissive for the nup49-313 strain, suggesting that GFP-Sfp1 is a nucleo-cytoplasmic shuttling protein. This observation confirmed the dynamic nature of the nucleo-cytoplasmic distribution of Sfp1. For example, a similar redistribution to the cytoplasm was previously reported following rapamycin treatment and under starvation (Marion et al., PNAS 2004). In conjunction with the observation of an interaction with Rpb4, the authors observed slower nuclear import kinetics for GFP-Sfp1 in the absence of Rpb4 when cells were transferred to a glucose-containing medium after a period of starvation. Since the redistribution of GFP-Sfp1 was abolished in an rpb1-1/nup49-313 double mutant, the authors concluded that Sfp1 localisation to the cytoplasm depends on transcription. The double mutant yeast cells may show a variety of non-specific effects at the restrictive temperature, and whether transcription is required for Sfp1 cytoplasmic localisation remains incompletely demonstrated.

We concur with Reviewer 2 that any heat inactivation of a temperature-sensitive (ts) protein can result in non-specific effects. In the instance of rpb1-1, these non-specific effects are anticipated because of the transcriptional arrest, which can eventually lead to a reduction in protein content. However, it is worth noting that this process takes some time, whereas the impact on export is more rapid. We note that that this assay has been developed and published in prior research by Pam Silver (op. cit.) and was reported in a number of subsequent papers. Reassuringly, our conclusion is supported by the observation that Sfp1 binds to Pol II transcripts co-transcriptionally.

1. Under starvation conditions, which led to the presence of Sfp1 in the cytoplasm and have previously been correlated with a decrease in the transcription of Sfp1 target genes, the authors observed that a plasmid-based expressed GFP-Sfp1 accumulated in cytoplasmic foci. These foci were also labelled by P-body markers such as Dcp2 and Lsm1. The quality of the microscopic images provided does not allow to determine whether Rpb4-RFP colocalises with GFP-Sfp1.

The submitted PDF figure is of low quality. We believe that high quality figure will be convincing.

1. To understand to which RNA Sfp1 might bind, the authors used an N-terminally tagged fusion protein in a cross-linking and purification experiment. This method identified 264 transcripts for which the CRAC signal was considered positive and which mostly correspond to abundant mRNAs, including 74 ribosomal protein mRNAs or metabolic enzyme-abundant mRNAs such as PGK1. The authors did not provide evidence for the specificity of the observed CRAC signal, in particular, what would be the background of a similar experiment performed without UV cross-linking. In a validation experiment, the presence of several mRNAs in a purified SFP1 fraction was measured at levels that reflect the relative levels of RNA in a total RNA extract. Negative controls showing that abundant mRNAs not found in the CRAC experiment were clearly depleted from the purified fraction with Sfp1 would be crucial to assessing the specificity of the observed protein-RNA interactions. The CRAC-selected mRNAs were enriched for genes whose expression was previously shown to be upregulated upon Sfp1 overexpression (Albert et al., 2019). The presence of unspliced RPL30 pre-mRNA in the Sfp1 purification was interpreted as a sign of co-transcriptional assembly of Sfp1 into mRNA, but in the absence of valid negative controls, this hypothesis would require further experimental validation.

We argue that the 264 CRAC+ genes represent a distinct group with many unique features. Moreover, many CRAC+ genes do not fall into the category of highly transcribed genes.

The biological significance of the 264 CRAC+ mRNAs was demonstrated by various experiments; all are inconsistent with technical flaws. Some examples are:

1. Fig. 2a and B show that most reads of CRAC+ mRNA were mapped to specific location – close the pA sites.

2. Fig. 2C shows that most reads of CRAC+ mRNA were mapped to specific RNA motif.

3. Most RiBi CRAC+ promoter contain Rap1 binding sites (p= 1.9x10-22), whereas the vast majority of RiBi CRAC- promoters do not contain Rap1 binding site. (Fig. 3C).

4. Fig. 4A shows that RiBi CRAC+ mRNAs become destabilized due to Sfp1 deletion, whereas RiBi CRAC- mRNAs do not. Fig. 4B shows similar results due to

5. Fig. 6B shows that the impact of Sfp1 on backtracking is substantially higher for CRAC+ than for CRAC- genes. This is most clearly visible in RiBi genes.

6. Fig. 7A shows that the Sfp1-dependent changes along the transcription units is substantially more rigorous for CRAC+ than for CRAC-.

7. Fig. S4B Shows that chromatin binding profile of Sfp1 is different for CRAC+ and CRAC- genes

Moreover, only a portion of the RiBi mRNAs binds Sfp1, despite similar expression of all RiBi.

Most importantly, these genes do not all fall into the category of highly transcribed genes. On the contrary, as depicted in Figure 6A (green dots), it is evident that CRAC+ genes exhibit a diverse range of Rpb3 ChIP and GRO signals. Furthermore, as illustrated in Figure 7A, when comparing CRAC+ to Q1 (the most highly transcribed genes), it becomes evident that the Rpb4/Rpb3 profile of CRAC+ genes is not a result of high transcription levels. In our revised paper, we will give increased attention to this matter in the Discussion section.

1. To address the important question of whether co-transcriptional assembly of Spf1 with transcripts could alter their stability, the authors first used a reporter system in which the RPL30 transcription unit is transferred to vectors under different transcriptional contexts, as previously described by the Choder laboratory (Bregman et al. 2011). While RPL30 expressed under an ACT1 promoter was barely detectable, the highest levels of RNA were observed in the context of the native upstream RPL30 sequence when Rap1 binding sites were also present. Sfp1 showed better association with reporter mRNAs containing Rap1 binding sites in the promoter region. However, removal of the Rap1 binding sites from the reporter vector also led to a drastic decrease in reporter mRNA levels. Whether the fraction of co-purified RNA is nuclear and co-transcriptional or not cannot be inferred from these results.

The proposed co-transcriptional binding of Sfp1 is based on the findings presented in Figure 5C and Figure S2D, as well as the observed binding of Sfp1 to transcripts containing introns, as shown in Figures 2D and 3B. Our conclusion, which we still uphold, was drawn from the results presented in Figure 3. These results led us to the assertion that the "RNA-binding capacity of Sfp1 is regulated by Rap1-binding sites located at the promoter." We maintain our stance on this conclusion. Indeed, the Rap1 binding site does impact mRNA levels, as highlighted by Reviewer 2. However, "construct E," which possesses a promoter with a Rap1 binding site, exhibits lower transcript levels compared to "construct F," which lacks such a binding site in its promoter. Despite this difference in transcript levels, Sfp1 was able to pull down the former transcript but not the latter, even though expression of the former gene is relatively low. Thus, the results appear to be more reliant on the specific capacity of Sfp1 to interact with the transcript rather than on the transcript's expression level.

1. To complement the biochemical data presented in the first part of the manuscript, the authors turned to the deletion or rapid depletion of SFP1 and used labelling experiments to assess changes in the rate of synthesis, abundance, and decay of mRNAs under these conditions. An important observation was that in the absence of Sfp1, mRNAs encoding ribosomal protein genes not only had a reduced synthesis rate but also an increased degradation rate. This important observation needs careful validation, as genomic run-on experiments were used to measure half-lives, and this particular method was found to give results that correlated poorly with other measures of half-life in yeast (e.g. Chappelboim et al., 2022 for a comparison). Similarly, the use of thiolutin to block transcription as a method of assessing mRNA half-life has been reported to be problematic, as thiolutin can specifically inhibit the degradation of ribosomal protein mRNA (Pelechano & Perez-Ortin, 2008). Specific repressible reporters, such as those used by Baudrimont et al. (2017), would need to be tested to validate the effect of Sfp1 on the half-life of specific mRNAs. Also, it would be very difficult to infer from the images presented whether the rate of deadenylation is altered by Sfp1.

Various methods exist for assessing mRNA half-lives (HLs), and each of them carries its own set of challenges and biases. Consequently, it becomes problematic to directly compare HL values of a specific mRNA when different methods are employed. The superiority of one particular method over others remains unclear. However, they all exhibit a high degree of reliability when it comes to comparing different strains under the identical conditions using a single method.

Estimating half-lives through the GRO approach is a non-invasive method, applied on optimally proliferating cells, which has been employed in numerous publications. While no method is without its limitations, we consider this approach to be among the most dependable. Our HL determination using thiolutin to block transcription provided results that were consistent with the values obtained by the GRO approach.

Nevertheless, in our revised manuscript, we plan to supplement the HL data, obtain by thiolutin, with results obtained by subjecting cells to a temperature shift to 42°C, a natural method to block transcription in wild-type (WT) cells. This approach to determine HLs has been previously reported in our publications, such as Lotan et al. (2005, 2007) and Goler Baron et al. (2008).

1. The effects of SFP1 on transcription were investigated by chromatin purification with Rpb3, a subunit of RNA polymerase, and the results were compared with synthesis rates determined by genomic run-on experiments. The decrease in polII presence on transcripts in the absence of SFP1 was not accompanied by a marked decrease in transcript output, suggesting an effect of Sfp1 in ensuring robust transcription and avoiding RNA polymerase backtracking. To further investigate the phenotypes associated with the depletion or absence of Sfp1, the authors examined the presence of Rpb4 along transcription units compared to Rpb3. One effect of spf1 deficiency was that this ratio, which decreased from the start of transcription towards the end of transcripts, increased slightly. The results presented are largely correlative and could arise from the focus on very specific types of mRNAs, such as those of ribosomal protein genes, which are sensitive to stress and are targeted by very active RNA degradation mechanisms activated, for example, under heat stress (Bresson et al., 2020).

Figure 7A illustrates a significant reduction in Rpb4/Rpb3 ratios along the transcription unit in WT cells. This reduction is notably more pronounced in CRAC+ genes compared to the highly transcribed quartile (Q1), which includes all ribosomal protein (RP) genes, and it is completely absent in sfp1∆ cells. Furthermore, it's important to highlight that the CRAC+ gene group displays a wide range of transcription rates, as measured by either Rpb3 ChIP or GRO (Figure 6A). Given these observations, it is challenging to reconcile how the heightened sensitivity of RP mRNA degradation in response to stress could account for the more pronounced differences in the configuration of the Pol II elongation complex that are detected in CRAC+ genes under standard culture conditions in wt cells.

Correlative studies are particularly informative when a gene mutation eliminates a correlation, and this is precisely the type of study depicted in Figure 7B-C. The configuration of elongating Pol II (as reflected by Rpb4/Rpb3 ratios) and the backtracking index are both transcriptional outputs. It is difficult to envision how stress-induced destabilization of RP mRNAs could explain the twofold higher correlation between these two parameters observed in CRAC+ genes under non-stressful conditions in WT cells (Figure 7B).

Furthermore, it's worth noting that in WT cells, CRAC+ genes did not display any apparent unusual destabilization, but rather exhibited higher (not lower) mRNA stability compared to CRAC- genes (Figure 7C).

Strengths: - Diversity of experimental approaches used - Validation of large-scale results with appropriate reporters

Weaknesses: - Choice of evaluation method to test mRNA half-life - Lack of controls for the CRAC results

eLife assessment

This study shows that the yeast transcription factor Sfp1 binds to a subset of its target gene mRNAs, increases their half-lives, and affects RNA polymerase II backtracking. These, and other related findings, provide important new insights into mechanisms by which a transcription factor can affect post-transcriptional steps in gene regulation. The main claims are partially backed by the evidence presented. However, the evidence remains incomplete as the methods used to estimate RNA degradation rates and the biochemistry of Sfp1-RNA complexes require further validation.

Reviewer #1 (Public Review):

Summary:<br /> This manuscript builds upon the authors' previous work on the cross-talk between transcription initiation and post-transcriptional events in yeast gene expression. These prior studies identified an mRNA 'imprinting' phenomenon linked to genes activated by the Rap1 transcription factor (TF), a surprising role for the Sfp1 TF in promoting RNA polymerase II (RNAPII) backtracking, and a role for the non-essential RNAPII subunits Rpb4/7 in the regulation of mRNA decay and translation. Here the authors aimed to extend these observations to provide a more coherent picture of the role of Sfp1 in transcription initiation and subsequent steps in gene expression. They provide evidence for (1) a physical interaction between Sfp1 and Rpb4, (2) Sfp1 binding and stabilization of mRNAs derived from genes whose promoters are bound by both Rap1 and Sfp1 and (3) an effect of Sfp1 on Rpb4 binding or conformation during transcription elongation.

Strengths:<br /> This study provides evidence that a TF (yeast Sfp1), in addition to stimulating transcription initiation, can at some target genes interact with their mRNA transcripts and promote their stability. Sfp1 thus has a positive effect on two distinct regulatory steps. Furthermore, evidence is presented indicating that strong Sfp1 mRNA association requires both Rap1 and Sfp1 promoter binding and is increased at a sequence motif near the polyA track of many target mRNAs. Finally, they provide compelling evidence that Sfp1-bound mRNAs have higher levels of RNAPII backtracking and altered Rpb4 association or conformation compared to those not bound by Sfp1.

Weaknesses:<br /> The Sfp1-Rpb4 association is supported only by a two-hybrid assay that is poorly described and lacks an important control. Furthermore, there is no evidence that this interaction is direct, nor are the interaction domains on either protein identified (or mutated to address function).

The contention that Sfp1 nuclear export to the cytoplasm is transcription-dependent is not well supported by the experiments shown, which are not properly described in the text and are not accompanied by any primary data.<br /> The presence of Sfp1 in P-bodies is of unclear relevance and the authors do not ask whether Sfp1-bound mRNAs are also present in these condensates.

Further analysis of Sfp1-bound mRNAs would be of interest, particularly to address the question of whether those from ribosomal protein genes and other growth-related genes that are known to display Sfp1 binding in their promoters are regulated (either stabilized or destabilized) by Sfp1.

The authors need to discuss, and ideally address, the apparent paradox that their previous findings showed that Rap1 acts to destabilize its downstream transcripts, i.e. that it has the opposite effect of Sfp1 shown here.

Finally, recent studies indicate that the drugs used here to measure mRNA stability induce a strong stress response accompanied by rapid and complex effects on transcription. Their relevance to mRNA stability in unstressed cells is questionable.

Reviewer #1 (Public Review):

Summary:<br /> This manuscript explores the impact of serotonin on olfactory coding in the antennal lobe of locusts and odor-evoked behavior. The authors use serotonin injections paired with an odor-evoked palp-opening response assay and bath application of serotonin with intracellular recordings of odor-evoked responses from projection neurons (PNs).

Strengths:<br /> The authors make several interesting observations, including that serotonin enhances behavioral responses to appetitive odors in starved and fed animals, induces spontaneous bursting in PNs, and uniformly enhances PN responses to odors. Overall, I had no technical concerns.

Weaknesses:<br /> While there are several interesting observations, the conclusions that serotonin enhanced sensitivity specifically and that serotonin had feeding-state-specific effects, were not supported by the evidence provided. Furthermore, there were other instances in which much more clarification was needed for me to follow the assumptions being made and inadequate statistical testing was reported.

Major concerns.<br /> -To enhance olfactory sensitivity, the expected results would be that serotonin causes locusts to perceive each odor as being at a relatively higher concentration. The authors recapitulate a classic olfactory behavioral phenomenon where higher odor concentrations evoke weaker responses which is indicative of the odors becoming aversive. If serotonin enhanced the sensitivity to odors, then the dose-response curve should have shifted to the left, resulting in a more pronounced aversion to high odor concentrations. However, the authors show an increase in response magnitude across all odor concentrations. I don't think the authors can claim that serotonin enhances the behavioral sensitivity to odors because the locusts no longer show concentration-dependent aversion. Instead, I think the authors can claim that serotonin induces increased olfactory arousal.

-The authors report that 5-HT causes PNs to change from tonic to bursting and conclude that this stems from a change in excitability. However, excitability tests (such as I/V plots) were not included, so it's difficult to disambiguate excitability changes from changes in synaptic input from other network components.

-There is another explanation for the theoretical discrepancy between physiology and behavior, which is that odor coding is further processing in higher brain regions (ie. Other than the antennal lobe) not studied in the physiological component of this study. This should at least be discussed.

-The authors cannot claim that serotonin underlies a hunger state-dependent modulation, only that serotonin impacts responses to appetitive odors. Serotonin enhanced PORs for starved and fed locusts, so the conclusion would be that serotonin enhances responses regardless of the hunger state. If the authors had antagonized 5-HT receptors and shown that feeding no longer impacts POR, then they could make the claim that serotonin underlies this effect. As it stands, these appear to be two independent phenomena.

eLife assessment

The work shows that the experimental application of serotonin to locust antennal lobes induces an increased feeding-related response to some odorants (even in food-satiated animals). To explain how the odorant-specific effects are seen despite similar consequences of 5-HT modulation on all projection neuronal types analyses by electrophysiology, the authors propose a simple quantitative model built around PNs with different downstream connections. These convincing observations are useful to guide further studies of serotonin and other modulatory mechanisms in the olfactory system.

Reviewer #2 (Public Review):

Summary:<br /> The authors investigate the influence of serotonin on feeding behavior and electrophysiological responses in the antennal lobe of locusts. They find that serotonin injection changes behavior in an odor-specific way. In physiology experiments, they can show that antennal lobe neurons generally increase their baseline firing and odor responses upon serotonin injection. Using a modeling approach the authors propose a framework on how a general increase in antennal lobe output can lead to odor-specific changes in behavior. The authors finally suggest that serotonin injection can mimic a change in a hunger state.

Strengths:<br /> This study shows that serotonin affects feeding behavior and odor processing in the antennal lobe of locusts, as serotonin injection increases activity levels of antennal lobe neurons. This study provides another piece of evidence that serotonin is a general neuromodulator within the early olfactory processing system across insects and even phyla.

Weaknesses:<br /> I have several concerns regarding missing control experiments, unclear data analysis, and interpretation of results.

A detailed description of the behavioral experiments is lacking. Did the authors also provide a mineral oil control and did they analyze the baseline POR response? Is there an increase in baseline response after serotonin exposure already at the behavioral output level? It is generally unclear how naturalistic the chosen odor concentrations are. This is especially important as behavioral responses to different concentrations of odors are differently modulated after serotonin injection (Figure 2: Linalool and Ammonium).

Regarding recordings of potential PNs - the authors do not provide evidence that they did record from projection neurons and not other types of antennal lobe neurons. Thus, these claims should be phrased more carefully.

The presented model suggests labeled lines in the antennal lobe output of locusts. Could the presented model also explain a shift in behavior from aversion to attraction - such as seen in locusts when they switch from a solitarious to a gregarious state? The authors might want to discuss other possible scenarios, such as that odor evaluation and decision-making take place in higher brain regions, or that other neuromodulators might affect behavioral output. Serotonin injections could affect behavior via modulation of other cell types than antennal lobe neurons. This should also be discussed - the same is true for potential PNs - serotonin might not directly affect this cell type, but might rather shut down local inhibitory neurons.

Finally, the authors claim that serotonin injection can mimic the starved state behavioral response. However, this is only shown for one of the four odors that are tested for behavior (HEX), thus the data does not support this claim.

Reviewer #1 (Public Review):

Summary:<br /> Animals in natural environments need to identify predator-associated cues and respond with the appropriate behavioral response to survive. In rodents, some chemical cues produced by predators (e.g., cat saliva) are detected by chemosensory neurons in the vomeronasal organ (VNO). The VNO transmits predator-associated information to the accessory olfactory bulb, which in turn projects to the medial amygdala and the bed nucleus of the stria terminalis, two regions implicated in the initiation of antipredator defensive behaviors. A downstream area to these two regions is the ventromedial hypothalamus (VMH), which has been shown to control both active (i.e., flight) and passive (i.e, freezing) antipredator defensive responses via distinct efferent projections to the anterior hypothalamic nucleus or the periaqueductal gray, respectively. However, whether differences in predator-associated sensory information initially processed in the VNO and further conveyed to the VMH can trigger different types of behavioral responses remained unexplored. To address this question, here the authors investigated the behavioral responses of mice exposed to either fresh or old cat saliva, and further compared the underlying neural circuits that are activated by cat saliva with different freshness.

The scientific question of the study is valid, the experiments were well-performed, and the statistical analyses are appropriate. However, there are some concerns that may directly affect the main interpretation of the results.

Major Concerns:<br /> 1. An important point that the authors should clarify in this study is whether mice are detecting qualitative or quantitative differences between fresh and old cat saliva. Do the environmental conditions in which the old saliva was maintained cause degradation of Fel d 4, the main protein known for inducing a defensive response in rodents? (see Papes et al, 2010 again). If that is the case, one would expect that a lower concentration of Fel d 4 in the old saliva after protein degradation would result in reduced antipredator responses. Alternatively, if the authors believe that different proteins that are absent in the old saliva are contributing to the increased defensive responses observed with the fresh saliva, further protein quantification experiments should be performed. An important experiment to differentiate qualitative versus quantitative differences between the two types of saliva would be diluting the fresh saliva to verify if the amount of protein, rather than the type of protein, is the main factor regulating the behavioral differences.

2. The authors claim that fresh saliva is recognized as an immediate danger by rodents, whereas old saliva is recognized as a trace of danger. However, the study lacks empirical tests to support this interpretation. With the current experimental tests, the behavioral differences between animals exposed to fresh vs. old saliva could be uniquely due to the reduced amount of the exact same protein (e.g., Fel d 4) in the two samples of saliva.

3. In Figure 4H, the authors state that there were no significant differences in the number of cFos-positive cells between the two saliva-exposed groups. However, this result disagrees with the next result section showing that fresh and old saliva differentially activate the VMH. It is unclear why cFos quantification and behavioral correlations were not performed in other upstream areas that connect the VNO to the VMH (e.g., BNST, MeA, and PMCo). That would provide a better understanding of how brain activity correlates with the different types of behaviors reported with the fresh vs. old saliva.

4. The interpretation that fresh and old saliva activates different subpopulations of neurons in the VMH based on the observation that cFos positively correlates with freezing responses only with the fresh saliva lacks empirical evidence. To address this question, the authors should use two neuronal activity markers to track the response of the same population of VHM cells within the same animals during exposure to fresh vs. old saliva. Alternatively, they could use single-cell electrophysiology or imaging tools to demonstrate that cat saliva of distinct freshness activates different subpopulations of cells in the VMH. Any interpretation without a direct within-subject comparison or the use of cell-type markers would become merely speculative. Furthermore, the authors assume that differential activations of mitral cells between fresh and old saliva result in the differential activation of VMH subpopulations (page 13, line 3). However, there are intermediate structures between the mitral cells and the VMH, which are completely ignored in this study (e.g., BNST, medial amygdala).

5. The authors incorrectly cited the Papes et al., 2010 article on several occasions across the manuscript. In the introduction, the authors cited the Papes et al 2010 study to make reference to the response of rodents to chemical cues, but the Papes et al. study did not use any of the chemical cues listed by the authors (e.g., fox feces, snake skin, cat fur, and cat collars). Instead, the Papes et al. 2010 article used the same chemical cue as the present study: cat saliva. The Papes et al. 2010 article was miscited again in the results section where the authors cited the study to make reference to other sources of cat odor that differ from the cat saliva such as cat fur and cat collars. Because the Papes et al. 2010 article has previously shown the involvement of Trpc2 receptors in the VNO for the detection of cat saliva and the subsequent expression of defensive behaviors by using Trpc2-KO mice, the authors should properly cite this study in the introduction and across the manuscript when making reference to their findings.

6. In the introduction, the authors hypothesized that the VNO detects predator cues and sends sensory signals to the VMH to trigger defensive behavioral decisions and stated that direct evidence to support this hypothesis is still missing. However, the evidence that cat saliva activates the VMH and that activity in the VMH is necessary for the expression of antipredator defensive response in rodents has been previously demonstrated in a study by Engelke et al., 2021 (PMID: 33947849), which was entirely omitted by the authors.

7. In the discussion, the authors stated that their findings suggest that the induction of robust freezing behavior is mediated by a distinct subpopulation of VMH neurons. The authors should cite the study by Kennedy et al., 2020 (PMID: 32939094) that shows the involvement of VMH in the regulation of persistent internal states of fear, which may provide an alternative explanation for why distinct concentrations of saliva could result in different behavioral outcomes.

8. The anatomical connectivity between the olfactory system and the ventromedial hypothalamus (VMH) in the abstract is unclear. The authors should clarify that the VMH does not receive direct inputs from the vomeronasal organ (VNO) nor the accessory olfactory bulb (AOB) as it seems in the current text.

Reviewer #2 (Public Review):

In this study, Nguyen et al. showed that cat saliva can robustly induce freezing behavior in mice. This effect is mediated through the accessory olfactory system as it requires physical contact and is abolished in Trp2 KO mice. The authors further showed that V2R-A4 cluster is responsive to cat saliva. Lastly, they demonstrated c-Fos induction in AOB and VMHdm/c by the cat saliva. The c-Fos level in the VMHdm/c is correlated with the freezing response.

Strength:<br /> The study opens an interesting direction. It reveals the potential neural circuit for detecting cat saliva and driving defense behavior in mice. The behavior results and the critical role of the accessory olfactory system in detecting cat saliva are clear and convincing.

Weakness:<br /> The findings are relatively preliminary. The identities of the receptor and the ligand in the cat saliva that induces the behavior remain unclear. The identity of VMH cells that are activated by the cat saliva remains unclear. There is a lack of targeted functional manipulation to demonstrate the role of V2R-A4 or VMH cells in the behavioral response to cat saliva.

eLife assessment

This valuable study shows that, in mice, fresh cat saliva elicits a greater defensive response compared to old cat saliva. Additionally, the authors implicate the vomeronasal organ (VNO) and ventromedial hypothalamus (VMH) as part of a circuit that underlies this process. While the study has potential, the results are somewhat preliminary, and as such the evidence presented is incomplete.

Reviewer #3 (Public Review):

Summary:<br /> Nguyen et al show data indicating that the vomeronasal organ (VNO) and ventromedial hypothalamus (VMH) are part of a circuit that elicits defensive responses induced by predator odors. They also show that using fresh or old predator saliva may be a method to change the perceived imminence of predation. The authors also identify a family of VNO receptors that are activated by cat saliva. Next, the authors show how different components of this defensive circuit are activated by saliva, as measured by fos expression. Though interesting, the findings are not all integrated into a single narrative, and some of the results are only replications of earlier findings using modern methods. Overall, these findings provide incremental advance.

Strengths:<br /> 1 Predator saliva is a stimulus of high ethological relevance<br /> 2 The authors performed a careful quantification of fos induction across the anterior-posterior axis in Figure 6.

Weaknesses:<br /> 1 It is unclear if fresh and old saliva indeed alter the perceived imminence predation, as claimed by the authors. Prior work indicates that lower imminence induces anxiety-related actions, such as re-organization of meal patterns and avoidance of open spaces, while slightly higher imminence produces freezing. Here, the authors show that fresh and old predator saliva only provoke different amounts of freezing, rather than changing the topography of defensive behaviors, as explained above. Another prediction of predatory imminence theory would be that lower imminence induced by old saliva should produce stronger cortical activation, while fresh saliva would activate the amygdala, if these stimuli indeed correspond to significantly different levels of predation imminence.

2 It is known that predator odors activate and require AOB, VNO, and VMH, thus replications of these findings are not novel, decreasing the impact of this work.

3 There is a lack of standard circuit dissection methods, such as characterizing the behavioral effects of increasing and decreasing the neural activity of relevant cell bodies and axonal projections, significantly decreasing the mechanistic insights generated by this work.

4 The correlation shown in Figure 5c may be spurious. It appears that the correlation is primarily driven by a single point (the green square point near the bottom left corner). All correlations should be calculated using Spearman correlation, which is non-parametric and less likely to show a large correlation due to a small number of outliers. Regardless of the correlation method used, there are too few points in Figure 5c to establish a reliable correlation. Please add more points to 5c.

5 Some of the findings are disconnected from the story. For example, the authors show that V2R-A4-expressing cells are activated by predator odors. Are these cells more likely to be connected to the rest of the predatory defense circuit than other VNO cells?

6 Were there other behavioral differences induced by fresh compared to old saliva? Do they provoke differences in stretch-attend risk evaluation postures, number of approaches, the average distance to odor stimulus, the velocity of movements towards and away from the odor stimulus, etc?

Reviewer #3 (Public Review):

The main problem with the work is that the results are only descriptive and do not allow any inferences or conclusions about the importance of the function of G4 structures. The discussion and conclusions are poor. The results are preliminary and in order to try to make the analysis more interesting, it should be further extended and the data must be explored in a much greater depth.

eLife assessment

This fundamental study explores the relationship between guanine-quadruplex structures and pathogenicity islands in 89 pathogenic strains. Guanine-quadruplex structures were found to be non-randomly distributed within pathogenicity islands and conserved within the same strains. Positive correlations were observed between Guanine-quadruplex structures and GC content across various genomic features, suggesting a link between these structures and GC-rich regions. These compelling findings shed light on the molecular mechanisms of Guanine-quadruplex structure-pathogenicity island interactions and will be of interest to all microbiologists.

Reviewer #1 (Public Review):

Summary:

This study explores the relationship between guanine-quadruplex (G4) structures and pathogenicity islands (PAIs) in 89 pathogenic strains. G4 structures were found to be non-randomly distributed within PAIs and conserved within the same strains. Positive correlations were observed between G4s and GC content across various genomic features, suggesting a link between G4 structures and GC-rich regions. Differences in GC content between PAIs and the core genome underscored the unique nature of PAIs. High-confidence G4 structures in Escherichia coli's regulatory regions were identified, influencing DNA integration within PAIs. These findings shed light on the molecular mechanisms of G4-PAI interactions, enhancing our understanding of bacterial pathogenicity and G4 structures in infectious diseases.

Strengths:

The findings of this study hold significant implications for our understanding of bacterial pathogenicity and the role of guanine-quadruplex (G4) structures.

Molecular Mechanisms of Pathogenicity: The study highlights that G4 structures are not randomly distributed within pathogenicity islands (PAIs), suggesting a potential role in regulating pathogenicity. This insight into the uneven distribution of G4s within PAIs provides a basis for further research into the molecular mechanisms underlying bacterial pathogenicity.

Conservation of G4 Structures: The consistent conservation of G4 structures within the same pathogenic strains suggests that these structures might play a vital and possibly conserved role in the pathogenicity of these bacteria. This finding opens doors for exploring how G4s influence virulence across different pathogens.

Unique Nature of PAIs: The differences in GC content between PAIs and the core genome underscore the unique nature of PAIs. This distinction suggests that factors such as DNA topology and G4 structures might contribute to the specialized functions and characteristics of PAIs, which are often associated with virulence genes.

Regulatory Role of G4s: The identification of high-confidence G4 structures within regulatory regions of Escherichia coli implies that these structures could influence the efficiency or specificity of DNA integration events within PAIs. This finding provides a potential mechanism by which G4s can impact the pathogenicity of bacteria.

Weaknesses:

No weaknesses were identified by this reviewer.

Overall, the study provides fundamental insights into the pathogenicity island and conservation of G4 motifs.

Reviewer #2 (Public Review):

Summary:

In the manuscript entitled "The Intricate Relationship of G-Quadruplexes and Pathogenicity Islands: A Window into Bacterial Pathogenicity" Bo Lyu explored the interactions between guanine-quadruplex (G4) structures and pathogenicity islands (PAIs) in 89 bacterial genomes through a rigorous computational approach. This paper handles an intriguing and complex topic in the field of pathogenomics. It has the potential to contribute significantly to the understanding of G4-PAI interactions and bacterial pathogenicity.

Strengths:

- The chosen research area.<br /> - The summarizing of the results through neat illustrations.

Weaknesses:

This reviewer did not find any significant weaknesses.

Reviewer #2 (Public Review):

Summary:

This study by Sun et al. identifies a novel role for IBTK in promoting cancer protein translation, through regulation of the translational helicase eIF4A1. Using a multifaceted approach, the authors demonstrate that IBTK interacts with and ubiquitinates eIF4A1 in a non-degradative manner, enhancing its activation downstream of mTORC1/S6K1 signaling. This represents a significant advance in elucidating the complex layers of dysregulated translational control in cancer.

Strengths:

A major strength of this work is the convincing biochemical evidence for a direct regulatory relationship between IBTK and eIF4A1. The authors utilize affinity purification and proximity labeling methods to comprehensively map the IBTK interactome, identifying eIF4A1 as a top hit. Importantly, they validate this interaction and the specificity for eIF4A1 over other eIF4 isoforms by co-immunoprecipitation in multiple cell lines. Building on this, they demonstrate that IBTK catalyzes non-degradative ubiquitination of eIF4A1 both in cells and in vitro through the E3 ligase activity of the CRL3-IBTK complex. Mapping IBTK phosphorylation sites and showing mTORC1/S6K1-dependent regulation provides mechanistic insight. The reduction in global translation and eIF4A1-dependent oncoproteins upon IBTK loss, along with clinical data linking IBTK to poor prognosis, support the functional importance.

Weaknesses:

While these data compellingly establish IBTK as a binding partner and modifier of eIF4A1, a remaining weakness is the lack of direct measurements showing IBTK regulates eIF4A1 helicase activity and translation of target mRNAs. While the effects of IBTK knockout/overexpression on bulk protein synthesis are shown, the expression of multiple eIF4A1 target oncogenes remains unchanged.

Summary:

Overall, this study significantly advances our understanding of how aberrant mTORC1/S6K1 signaling promotes cancer pathogenic translation via IBTK and eIF4A1. The proteomic, biochemical, and phosphorylation mapping approaches established here provide a blueprint for interrogating IBTK function. These data should galvanize future efforts to target the mTORC1/S6K1-IBTK-eIF4A1 axis as an avenue for cancer therapy, particularly in combination with eIF4A inhibitors.

eLife assessment

The findings in this fundamental study identify a novel substrate and mediator of oncogenesis downstream of mTORC1 and advance our understanding of the mechanistic basis of mTORC1-regulated cap-dependent translation and protein synthesis. The authors present convincing data using an array of biochemical, proteomic, and functional assays. These studies are of broad relevance to biochemists and cancer biologists and have potential translational relevance in cancer.

Reviewer #1 (Public Review):

In this study, the authors examined the role of IBTK, a substrate-binding adaptor of the CRL3 ubiquitin ligase complex, in modulating the activity of the eiF4F translation initiation complex. They find that IBTK mediates the non-degradative ubiquitination of eiF4A1, promotes cap-dependent translational initiation, nascent protein synthesis, oncogene expression, and tumor cell growth. Correspondingly, phosphorylation of  IBTK by mTORC1/ S6K1 increases eIF4A1 ubiquitination and sustains oncogenic translation.

Strengths:

This study utilizes multiple biochemical, proteomic, functional, and cell biology assays to substantiate their results.  Importantly, the work nominates IBTK as a unique substrate of mTORC1, and further validates eiF4A1 ( a crucial subunit of the ei44F complex) as a promising therapeutic target in cancer. Since IBTK interacts broadly with multiple members of the translational initial complex - it will be interesting to examine its role in eiF2alpha-mediated ER stress as well as eiF3-mediated translation. Additionally, since IBTK exerts pro-survival effects in multiple cell types, it will be of relevance to characterize the role of IBTK in mediating increased mTORC1 mediated translation in other tumor types, thus potentially impacting their treatment with eiF4F inhibitors.

Limitations/Weaknesses:

The findings are mostly well supported by data, but some areas need clarification and could potentially be enhanced with further experiments:

1) Since eiF4A1 appears to function downstream of IBTK1, can the effects of IBTK1 KO/KD in reducing puromycin incorporation (in Fig 3A),  cap-dependent luciferase reporter activity (Fig 3G), reduced oncogene expression ( Fig 4A) or 2D growth/ invasion assays (Fig 4) be overcome or bypassed by overexpressing eiF4A1? These could potentially be tested in future studies. 

2) The decrease in nascent protein synthesis in puromycin incorporation assays in Figure 3A suggest that the effects of IBTK KO are comparable to and additive with silvesterol. It would be of interest to examine whether silvesterol decreases nascent protein synthesis or increases stress granules in the IBTK KO cells stably expressing IBTK as well. 

3) The data presented in Figure 5 regarding the role of mTORC1 in IBTK-mediated eiF4A1 ubiquitination needs further clarification on several points:

- It is not clear if the experiments in Figure 5F with Phos-tag gels are using the FLAG-IBTK deletion mutant or the peptide containing the mTOR sites as it is mentioned on line 517, page 19 "To do so, we generated an IBTK deletion mutant (900-1150 aa) spanning the potential mTORC1-regulated phosphorylation sites" This needs further clarification.

-It may be of benefit to repeat the Phos tag experiments with full-length FLAG-IBTK and/or endogenous IBTK with molecular weight markers indicating the size of migrated bands.

-Additionally, torin or Lambda phosphatase treatment may be used to confirm the specificity of the band in separate experiments.

-Phos-tag gels with the IBTK CRISPR KO line would also help confirm that the non-phosphorylated band is indeed IBTK. 

-It is unclear why the lower, phosphorylated bands seem to be increasing (rather than decreasing) with AA starvation/ Rapa in Fig 5H.

eLife assessment

This manuscript describes valuable new findings on the impact of chromatin context on the outcomes of microhomology-mediated end joining of DNA double-strand breaks (DSBs), specifically a preference for DSB-proximal microhomologies in repair within a heterochromatic compared to a euchromatic locus. The authors develop the Drosophila spermatogonia as a model for repair at induced DSBs in a mitotically-active tissue and leverage this system to provide convincing evidence that the local environment impacts the preference for repair mechanism and outcome. The work could be strengthened by the use of additional euchromatin insertion(s) to robustly validate the findings.

Reviewer #1 (Public Review):

Summary:<br /> In this study, the authors investigated the mechanisms to repair DSBs induced in euchromatic (Eu) or heterochromatic (Het) contexts in Drosophila. They used a previously described reporter construct that can be used to differentiate between HR, SSA, and mutagenic end joining in response to an I-SceI-induced DSB. Different sub-pathways of end joining (NHEJ, MMEJ, and SD-MMEJ) could be further distinguished by DNA sequence analysis. The main findings of the study are: (1) HR repair is more frequent in Het than in the Eu context; (2) mutagenic EJ repair is more frequent than HR in both contexts; (3) sub-pathways of mutagenic EJ are variable even within the same chromatin domain; and (4) SD-MMEJ repair is associated with larger deletions in the Eu than within the Het compartment.

Strengths:<br /> Overall, the study is well designed and the use of the Bam promoter to drive I-SceI removes some of the variability observed in previous studies. Importantly, the observation of different repair outcomes using the same reporter integrated at different genomic sites suggests that repair is influenced by chromatin state in addition to local DNA sequence context.

Weaknesses:<br /> The main concern I have is the use of only one Eu site versus four for the Het insertions. Given the variability observed between the Het insertions, analysis of a second Eu insertion would give more confidence that the differences observed are significant. One puzzling finding is that HR is increased when the reporter is inserted within the Het domain relative to the Eu domain, suggesting more end resection, yet deletions are smaller for the Het insertions. Bright Ddc2/ATRIP focus formation at DSBs induced in the Het domain is consistent with extensive end resection in this compartment. The authors speculate that this finding could indicate differences in the density of RPA loading or recruitment of Pol theta near ends. I recognize that measuring RPA density on single-stranded DNA would be extremely challenging, but is it known if Pol theta is recruited to DSBs within the Het domain before they move to the periphery?

Reviewer #2 (Public Review):

Summary:<br /> The authors seek to vary the integration site of a double-strand break repair reporter and assess how the chromatin state of different reporter integration sites impacts the contribution of various DSB repair pathways.

Strengths:<br /> It addresses repair in vivo. The reporter improves assay reliability (relative to previous fly DSB repair substrates) by inducing I-SceI within a more narrow and well-defined expression window. The authors' characterization of the spectrum of a-EJ products by sequencing is largely rigorous and thorough, and this often difficult to communicate data is presented in a clear and easily digested manner.

Weaknesses:<br /> The use of the single euchromatic site undercuts their ability to generalize the impact of chromatin state. This concern is minor when considering repair by HR, as repair efficiency appears to vary little when comparing repair across the 4 different heterochromatic sites. Still, it is possible the single euchromatic site they used is an outlier in its sparing use of HR. The assessment of repair by alt-EJ is more problematic, though, since the character of repair appears to vary as much across the different heterochromatic sites as it does comparing a given heterochromatic site vs. the euchromatic site. For example, focusing on their central argument (decreased deletion during SD-MMEJ at heterochromatic sites), the difference between Het2 and all other sites appears to be more dramatic than the difference between Het1 and the single euchromatic site (Figure 5A, Supp Fig 2).

Reviewer #3 (Public Review):

Summary:<br /> In this manuscript, Chiolo and colleagues adapt a Drosophila induced-DSB repair outcome assay to the spermatogonia. In order to compare the outcomes in H3K9me-rich centromeric heterochromatin with a euchromatic site they use a cross to a silencing mutant to reveal the sequence changes in the reporter, which otherwise are not expressed. The authors corroborate that homologous recombination (HR) is up-regulated in this chromatin context, consistent with prior studies. Applying sequencing to mutagenic products the authors reveal context-dependent preferences in mutagenic end joining pathways and mechanisms, although these seem less categorical in terms of hetero- and euchromatin and instead sensitive to more subtle aspects of the local chromatin landscape. One theme, however, is that the microhomologies used for synthesis-dependent end joining are nearer to the induced DSB in heterochromatin than seen for the euchromatic DSB.

Strengths:<br /> 1. The use of the mitotically active spermatogonia and transient expression of the I-SceI to induce the DSB mitigates some caveats of prior experimental approaches including the fact that the cells are universally mitotically active. This approach also enables the outcomes to be assayed in the next generation, which is necessary for reporters expressed within heterochromatin. Thus, this is a technological tool that will be useful to other groups.

2. The observations suggest that MMEJ within heterochromatin (inferred to be Pol theta-dependent) prefers to use microhomologies close to the DSB. This suggests that either DSB end resection or RPA loading/removal is modulated by chromatin context, which is a new finding.

Weaknesses:<br /> 1. The observation that HR is preferred in heterochromatin has been documented in many prior systems.

2. Although the conclusions of the authors are well-supported by the data, the study is somewhat limited in mechanistic detail and would be strengthened by additional use of the genetic tools in the model system, particularly with regard to whether the preference for using microhomologies near the DSB in heterochromatin arises due to modulation of resection or RPA loading stability (the latter is the preferred interpretation of the authors, but goes untested). Nucleosome stability, presence of HP1, etc. seem attractive.

3. Given the variability observed for EJ pathway usage at the four heterochromatic genomic sites probed in the manuscript there is some concern that a single euchromatic site may not be sufficient for rigorous comparisons. This is particularly true because there seems to be little transcription at the "euchromatic" region (Fig. S5). Given that we do not know what matters to dictate the outcomes (epigenetic modifications and/or transcriptional status), this is concerning.

4. (Minor) Some caution should be stated in comparing the HR frequency between this system (low single digits) and prior induction/tissue systems (~20%) because the time domain of cut and repair cycles is vastly different.

5. (Minor) While there are certainly strengths to using the spermatogonia system, one also wonders if it might not have some unique biology given the importance of maintaining genome integrity in this tissue (e.g. the piRNA pathways to repress transposon mobilization). A comment on this point would be welcomed.

6. (Minor) The authors argue that alt-EJ is less mutagenic as a consequence of the observed use of microhomologues closer to the DSB, but what they really mean perhaps is that less sequence is lost? A mutagenic outcome can be equally deleterious in other cases if 1, 5, or 20+ bps are lost, depending on the context.

eLife assessment

This in principle useful study suggests that the G-protein subunit Gng13 is required for limiting injury and inflammation following H1N1 influenza infection via anti-inflammatory effects from ectopic tuft cells. There appears to be support for Gng13 helping to limit influenza injury in the transgenic mouse models used here, but evidence for these effects being mediated by tuft cells is incomplete, giving conflicting data from mice that lack tuft cells entirely.

Reviewer #1 (Public Review):

Li et al. report here on the expression of a G-protein subunit Gng13 in ectopic tuft cells that develop after severe pulmonary injury in mice. By deleting this gene in ectopic tuft cells as they arise, the authors observed worsened lung injury and greater inflammation after influenza infection, as well as a decrease in the overall number of ectopic tuft cells. This was in stark contrast to the deletion of Trpm5, a cation channel generally thought to be required for all functional gustatory signaling in tuft cells, where no phenotype is observed. Strengths here include a thorough assessment of lung injury via a number of different techniques. Weaknesses are notable: confusingly, these findings are at odds with reports from other groups demonstrating no obvious phenotype upon influenza infection in mice lacking the transcription factor Pou2f3, which is essential for all tuft cell specification and development. The authors speculate that heterogeneity within nascent tuft cell populations, specifically the presence of pro- and anti-inflammatory tuft cells, may explain this difference, but they do not provide any data to support this idea.

Reviewer #2 (Public Review):

Summary:<br /> The study by Li et al. aimed to demonstrate the role of the G𝛾13-mediated signal transduction pathway in tuft cell-driven inflammation resolution and repairing injured lung tissue. The authors showed a reduced number of tuft cells in the parenchyma of G𝛾13 null lungs following viral infection. Mice with a G𝛾13 null mutation showed increased lung damage and heightened macrophage infiltration when exposed to the H1N1 virus. Their further findings suggested that lung inflammation resolution, epithelial barrier, and fibrosis were worsened in G𝛾13 null mutants.

Strengths:<br /> The beautiful immunostaining findings do suggest that the number of tuft cells is decreased in Gr13 null mutants.

Weaknesses:<br /> The description of phenotypes, and the approaches used to measure the phenotypes are problematic. Rigorous investigation of the mouse lung phenotypes is needed to draw meaningful conclusions.

eLife assessment

This paper presents what could be a useful approach for association testing, using the output of neural networks that have been trained to predict functional changes from DNA sequences. The approach presented by the author is an interesting addition to statistical genetics. It is, however, unclear whether the method not only detects more associations but also whether the quality of these associations (i.e., the likelihood that they are causal associations) is as good or better than what one finds with conventional methods. The enrichment analyses are encouraging but without rigorous assessment of statistical power and a better understanding of the pitfalls of the method, the evidence for this being an advance that will find application in the field remains incomplete.

Reviewer #1 (Public Review):

Summary:<br /> In this paper, Song, Shi, and Lin use an existing deep learning-based sequence model to derive a score for each haplotype within a genomic region, and then perform association tests between these scores and phenotypes of interest. The authors then perform some downstream analyses (fine-mapping, various enrichment analyses, and building polygenic scores) to ensure that these associations are meaningful. The authors find that their approach allows them to find additional associations, the associations have biologically interpretable enrichments in terms of tissues and pathways, and can slightly improve polygenic scores when combined with standard SNP-based PRS.

Strengths:<br /> - I found the central idea of the paper to be conceptually straightforward and an appealing way to use the power of sequence models in an association testing framework.<br /> - The findings are largely biologically interpretable, and it seems like this could be a promising approach to boost power for some downstream applications.

Weaknesses:<br /> - The methods used to generate polygenic scores were difficult to follow. In particular, a fully connected neural network with linear activations predicting a single output should be equivalent to linear regression (all intermediate layers of the network can be collapsed using matrix-multiplication, so the output is just the inner product of the input with some vector). Using the last hidden layer of such a network for downstream tasks should also be equivalent to projecting the input down to a lower dimensional space with some essentially randomly chosen projection. As such, I am surprised that the neural network approach performs so well, and it would be nice if the authors could compare it to other linear approaches (e.g., LASSO or ridge regression for prediction; PCA or an auto-encoder for converting the input to a lower dimensional representation).

- A very interesting point of the paper was the low R^2 between the HFS scores in adjacent windows, but the explanation of this was unclear to me. Since the HFS scores are just deterministic functions of the SNPs, it feels like if the SNPs are in LD then the HFS scores should be and vice versa. It would be nice to compare the LD between adjacent windows to the average LD of pairs of SNPs from the two windows to see if this is driven by the fact that SNPs are being separated into windows, or if sei is somehow upweighting the importance of SNPs that are less linked to other SNPs (e.g., rare variants).

- There were also a number of robustness checks that would have been good to include in the paper. For instance, do the findings change if the windows are shifted? Do the findings change if the sequence is reverse-complemented?

- It was also difficult to contextualize the present work in terms of recent results showing that sequence models tend to not do very well at predicting cross-individual expression changes (and such results presumably hold for predicting cross-individual chromatin changes). In particular, it would be good for the authors to contrast their findings with the work of Alex Sasse and colleagues (https://www.biorxiv.org/content/10.1101/2023.03.16.532969.abstract) and Connie Huang and colleagues (https://www.biorxiv.org/content/10.1101/2023.06.30.547100.abstract).

Reviewer #2 (Public Review):

Summary:<br /> In this work, Song et al. propose a locus-based framework for performing GWAS and related downstream analyses including finemapping and polygenic risk score (PRS) estimation. GWAS are not sufficiently powered to detect phenotype associations with low-frequency variants. To overcome this limitation, the manuscript proposes a method to aggregate variant impacts on chromatin and transcription across a 4096 base pair (bp) loci in the form of a haplotype function score (HFS). At each locus, an association is computed between the HFS and trait. Computing associations at the level of imputed functional genomic scores should enable the integration of information across variants spanning the allele frequency spectrum and bolster the power of GWAS.

The HFS for each locus is derived from a sequence-based predictive model. Sei. Sei predicts 21,907 chromatin and TF binding tracks, which can be projected onto 40 pre-defined sequence classes ( representing promoters, enhancers, etc.). For each 4096 bp haplotype in their UKB cohort, the proposed method uses the Sei sequence class scores to derive the haplotype function score (HFS). The authors apply their method to 14 polygenic traits, identifying ~16,500 HFS-trait associations. They finemap these trait-associated loci with SuSie, as well as perform target gene/pathway discovery and PRS estimation.

Strengths:<br /> Sequence-based deep learning predictors of chromatin status and TF binding have become increasingly accurate over the past few years. Imputing aggregated variant impact using Sei, and then performing an HFS-trait association is, therefore, an interesting approach to bolster power in GWAS discovery. The manuscript demonstrates that associations can be identified at the level of an aggregated functional score. The finemapping and pathway identification analyses suggest that HFS-based associations identify relevant causal pathways and genes from an association study. Identifying associations at the level of functional genomics increases the portability of PRSs across populations. Imputing functional genomic predictions using a sequence-based deep learning model does not suffer from the limitation of TWAS where gene expression is imputed from a limited-size reference panel such as GTEx.

However, there are several major limitations that need to be addressed.

Major concerns/weaknesses:<br /> 1. There is limited characterization of the locus-level associations to SNP-level associations. How does the set of HFS-based associations differ from SNP-level associations?

2. A clear advantage of performing HFS-trait associations is that the HFS score is imputed by considering variants across the allele frequency spectrum. However, no evidence is provided demonstrating that rare variants contribute to associations derived by the model. Similarly, do the authors find evidence that allelic heterogeneity is leveraged by the HFS-based association model? It would be useful to do simulations here to characterize the model behavior in the presence of trait-associated rare variants.

3. Sei predicts chromatin status / ChIP-seq peaks in the center of a 4kb region. It would therefore be more relevant to predict HFS using overlapping sequence windows that tile the genome as opposed to using non-overlapping windows for computing HFS scores. Specifically, in line 482, the authors state that "the HFS score represents overall activity of the entire sequence, not only the few bp at the center", but this would not hold given that Sei is predicting activity at the center for any sequence.

4. Is the HFS-based association going to miss coding variation and several regulatory variants such as splicing variants? There are also going to be cases where there's an association driven by a variant that is correlated with a Sei prediction in a neighboring window. These would represent false positives for the method, it would be useful to identify or characterize these cases.

Additional minor concerns:<br /> 1. It's not clear whether SuSie-based finemapping is appropriate at the locus level, when there is limited LD between neighboring HFS bins. How does the choice of the number of causal loci and the size of the segment being finemapped affect the results and is SuSie a good fit in this scenario?

2. It is not clear how a single score is chosen from the 117 values predicted by Sei for each locus. SuSie is run assuming a single causal signal per locus, an assumption which may not hold at ~4kb resolution (several classes could be associated with the trait of interest). It's not clear whether SuSie, run in this parameter setting, is a good choice for variable selection here.

3.. A single HFS score is being chosen from amongst multiple tracks at each locus independently. Does this require additional multiple-hypothesis correction?

4. The results show that a larger number of loci are identified with HFS-based finemapping & that causal loci are enriched for causal SNPs. However, it is not clear how the number of causal loci should relate to the number of SNPs. It would be really nice to see examples of cases where a previously unresolved association is resolved when using HFS-based GWAS + finemapping.

5. Sequence-based deep learning model predictions can be miscalibrated for insertions and deletions (INDELs) as compared to SNPs. Scaling INDEL predictions would likely improve the downstream modeling.

Reviewer #1 (Public Review):

Summary:<br /> Shakhawat et al., investigated how enhancement of plasticity and impairment could result in the same behavioral phenotype. The authors tested the hypothesis that learning impairments result from saturation of plasticity mechanisms and had previously tested this hypothesis using mice lacking two class I major histocompatibility molecules. The current study extends this work by testing the saturation hypothesis in a Purkinje-cell (L7) specific Fmr1 knockout mouse mice, which have enhanced parallel fiber-Purkinje cell LTD. The authors found that L7-Fmr1 knockout mice are impaired on an oculomotor learning task and both pre-training, to reverse LTD, and diazepam, to suppress neural activity, eliminated the deficit when compared to controls.

Strengths:

This study tests the "saturation hypothesis" to understand plasticity in learning using a well-known behavior task, VOR, and an additional genetic mouse line with a cerebellar cell-specific target, L7-Fmr1 KO. This hypothesis is of interest to the community as it evokes a novel inquisition into LTD that has not been examined previously.

Utilizing a cell-specific mouse line that has been previously used as a genetic model to study Fragile X syndrome is a unique way to study the role of Purkinje cells and the Fmr1 gene. This increases the understanding in the field in regards to Fragile X syndrome and LTD.

The VOR task is a classic behavior task that is well understood, therefore using this metric is very reliable for testing new animal models and treatment strategies. The effects of pretraining are clearly robust and this analysis technique could be applied across different behavior data sets.

The rescue shown using diazepam is very interesting as this is a therapeutic that could be used in clinical populations as it is already approved.

There was a proper use of controls and all animal information was described. The statistical analysis and figures are clear and well describe the results.

Weaknesses:<br /> While the proposed hypothesis is tested using genetic animal models and the VOR task, LTD itself is not measured. This study would have benefited from a direct analysis of LTD in the cerebellar cortex in the proposed circuits.

Diazepam was shown to rescue learning in L7-Fmr1 KO mice, but this drug is a benzodiazepine and can cause a physical dependence. While the concentrations used in this study were quite low and animals were dosed acutely, potential side-effects of the drug were not examined, including any possible withdrawal. This drug is not specific to Purkinje cells or cerebellar circuits, so the action of the drug on cerebellar circuitry is not well understood for the study presented.

It was not mentioned if L7-Fmr1 KO mice have behavior impairments that worsen with age or if Purkinje cells and the cerebellar microcircuit are intact throughout the lifespan. Connections between Purkinje cells and interneurons could also influence the behavior results found.

While males and females were both used for the current study, only 7 of each sex were analyzed, which could be underpowered. While it might be justified to combine sexes for this particular study, it would be worth understanding this model in more detail.

Training was only shown up to 30 minutes and learning did not seem to plateau in most cases. What would happen if training continued beyond the 30 minutes? Would L7-Fmr1 KO mice catch-up to WT littermates?

The pathway discussed as the main focus for VOR in this learning paradigm was connections between parallel fibers (PF) and Purkinje cells, but the possibility of other local or downstream circuitry being involved was not discussed. PF-Purkinje cell circuits were not directly analyzed, which makes this claim difficult to assess.

The authors mostly achieved their aim and the results support their conclusion and proposed hypothesis. This work will be impactful on the field as it uses a new Purkinje-cell specific mouse model to study a classic cerebellar task. The use of diazepam could be further analyzed in other genetic models of neurodevelopmental disorders to understand if effects on LTD can rescue other pathways and behavior outcomes.

Reviewer #2 (Public Review):

This manuscript explores the seemingly paradoxical observation that enhanced synaptic plasticity impairs (rather than enhances) certain forms of learning and memory. The central hypothesis is that such impairments arise due to saturation of synaptic plasticity, such that the synaptic plasticity required for learning can no longer be induced. A prior study provided evidence for this hypothesis using transgenic mice that lack major histocompatibility class 1 molecules and show enhanced long-term depression (LTD) at synapses between granule cells and Purkinje cells of the cerebellum. The study found that a form of LTD-dependent motor learning-increasing the gain of the vestibulo-ocular reflex (VOR)-is impaired in these mice and can be rescued by manipulations designed to "unsaturate" LTD. The present study extends this line of investigation to another transgenic mouse line with enhanced LTD, namely, mice with the Fragile X gene knocked out. The main findings are that VOR gain increased learning is selectively impaired in these mice but can be rescued by specific manipulations of visuomotor experience known to reverse cerebellar LTD. Additionally, the authors show that a transient global enhancement of neuronal inhibition also selectively rescues gain increases learning. This latter finding has potential clinical relevance since the drug used to boost inhibition, diazepam, is FDA-approved and commonly used in the clinic. The evidence provided for the saturation is somewhat indirect because directly measuring synaptic strength in vivo is technically difficult. Nevertheless, the experimental results are solid. In particular, the specificity of the effects to forms of plasticity previously shown to require LTD is remarkable. The authors should consider including a brief discussion of some of the important untested assumptions of the saturation hypothesis, including the requirement that cerebellar LTD depends not only on pre- and postsynaptic activity (as is typically assumed) but also on the prior history of synaptic activation.

eLife assessment

This manuscript provides solid evidence for the involvement of membrane actin, and its regulatory proteins, mDia1/3, RhoA, and Rac1 in the mechanism of synaptic vesicle re-uptake (endocytosis). These valuable data fill a gap in the understanding of how the regulation of actin dynamics and endocytosis are linked. The manuscript will be of interest to all scientists working on cellular trafficking and membrane remodeling

Reviewer #1 (Public Review):

Summary: The authors set out to clarify the molecular mechanism of endocytosis (re-uptake) of synaptic vesicle (SV) membrane in the presynaptic terminal following release. They have examined the role of presynaptic actin, and of the actin regulatory proteins diaphanous-related formins ( mDia1/3), and Rho and Rac GTPases in controlling the endocytosis. They successfully show that presynaptic membrane-associated actin is required for normal SV endocytosis in the presynaptic terminal and that the rate of endocytosis is increased by activation of mDia1/3. They show that RhoA activity and Rac1 activity act in a partially redundant and synergistic fashion together with mDia1/3 to regulate the rate of SV endocytosis. The work adds substantially to our understanding of the molecular mechanisms of SV endocytosis in the presynaptic terminal.

Strengths: The authors use state-of-the-art optical recording of presynaptic endocytosis in primary hippocampal neurons, combined with well-executed genetic and pharmacological perturbations to document effects of alteration of actin polymerization on the rate of SV endocytosis. They show that removal of the short amino-terminal portion of mDia1 that associates with the membrane interrupts the association of mDia1 with membrane actin in the presynaptic terminal. They then use a wide variety of controlled perturbations, including genetic modification of the amount of mDia1/3 by knock-down and knockout, combined with inhibition of activity of RhoA and Rac1 by pharmacological agents, to document the quantitative importance of each agent and their synergistic relationship in regulation of endocytosis.<br /> The analysis is augmented by ultrastructural analyses that demonstrate the quantitative changes in numbers of synaptic vesicles and in uncoated membrane invaginations that are predicted by the optical recordings.<br /> The manuscript is well-written and the data are clearly explained. Statistical analysis of the data is strengthened by the very large number of data points analyzed for each experiment.

Weaknesses: There are no major weaknesses. The optical images as first presented are small and it is recommended that the authors provide larger, higher-resolution images.

Reviewer #2 (Public Review):

Summary:<br /> This manuscript expands on previous work from the Haucke group which demonstrated the role of formins in synaptic vesicle endocytosis. The techniques used to address the research question are state-of-the-art. As stated above there is a significant advance in knowledge, with particular respect to Rho/Rac signalling.

Strengths:<br /> The major strength of the work was to reveal new information regarding the control of both presynaptic actin dynamics and synaptic vesicle endocytosis via Rho/Rac cascades. In addition, there was further mechanistic insight regarding the specific function of mDia1/3. The methods used were state-of-the-art.

Weaknesses:<br /> There are a number of instances where the conclusions drawn are not supported by the submitted data, or further work is required to confirm these conclusions.

eLife assessment

This is an important paper on the role of engrams and relevant conditions that influence memory and forgetting. The variety of methods used, namely, behavioural, labeling, interrogation, immunohistochemistry, microscopy, pharmacology, and computational, are exemplary and provide solid evidence for the role of engrams in the dentate gyrus in memory retrieval and forgetting. This examination will be of interest broadly across behavioural and neural science.

Reviewer #1 (Public Review):

Summary:<br /> O'Leary and colleagues present data identifying several procedures that alter discrimination between novel and familiar objects, including time, environmental enrichment, Rac-1, context reexposure, and brief reminders of the familiar object. This is complimented with an engram approach to quantify cells that are active during learning to examine how their activation is impacted following each of the above procedures at test. With this behavioral data, authors apply a modeling approach to understand the factors that contribute to good and poor object memory recall.

Strengths:<br /> • Authors systematically test several factors that contribute to poor discrimination between novel and familiar objects. These results are extremely interesting and outline essential boundaries of incidental, nonaversive memory.<br /> • These results are further supported by engram-focused approaches to examine engram cells that are reactivated in states with poor and good object recognition recall.

Weaknesses:<br /> • For the environmental enrichment, authors seem to suggest objects in the homecage are similar to (or reminiscent of) the familiar object. Thus, the effect of improved memory may not be related to enrichment per se as much as it may be related to the preservation of an object's memory through multiple retrievals, not the enriching experiences of the environment itself. This would be consistent with the brief retrieval figure. Authors should include a more thorough discussion of this.

• Authors should justify the marginally increased number of engram cells in the non-enrichment group that did not show object discrimination at test, especially relative to other figures. More specific cell counting criteria may be helpful for this. For example, was the DG region counted for engram and cfos cells or only a subsection?

• It is unclear why the authors chose a reactivation time point of 1hr prior to testing. While this may be outside of the effective time window for pharmacological interference with reconsolidation for most compounds, it is not necessarily outside of the structural and functional neuronal changes accompanied by reconsolidation-related manipulations.

• Figure 5: Levels of exploration at test are inconsistent between manipulations. This is problematic, as context-only reexposures seem to increase exploration for objects overall in a manner that I'm unsure resembles 'forgetting'. Instead, cross-group comparisons would likely reveal increased exploration time for familiar and novel objects. While I understand 'forgetting' may be accompanied by greater exploration towards objects, this is inconsistent across and within the same figure. Further, this effect is within the period of time that rodents should show intact recognition. Instead, context-only exposures may form a competing (empty context) memory for the familiar object in that particular context.

• I am concerned at the interpretation that a memory is 'forgotten' across figures, especially considering the brief reminder experiments. Typically, if a reminder session can trigger the original memory or there is rapid reacquisition, then this implies there is some savings for the original content of the memory. For instance, multiple context retrievals in the absence of an object reminder may be more consistent with procedures that create a distinct memory and subsequently recruit a distinct engram.

• Authors state that spine density decreases over time. While that may be generally true, there is no evidence that mature mushroom spines are altered or that this is consistent across figures. Additionally, it's unclear if spine volume is consistently reduced in reactivated and non-reactivated engram cells across groups. This would provide evidence that there is a functionally distinct aspect of engram cells that is altered consistently in procedures resulting in poor recognition memory (e.g. increased spine density relative to spine density of non-reactivated engram cells and non-engram cells)

• Authors should discuss how the enrichment-neurogenesis results here are compatible with other neurogenesis work that supports forgetting.

Reviewer #2 (Public Review):

Summary:<br /> The manuscript examines an important question about how an inaccessible, natural forgotten memory can be retrieved through engram ensemble reactivation. It uses a variety of strategies including optogenetics, behavioral and pharmacological interventions to modulate engram accessibility. The data characterize the time course of natural forgetting using an object recognition task, in which animals can retrieve 1 day and 1 week after learning, but not 2 weeks later. Forgetting is correlated with lower levels of cell reactivation (c-fos expression during learning compared to retrieval) and reduction in spine density and volume in the engram cells. Artificial activation of the original engram was sufficient to induce recall of the forgotten object memory while artificial inhibition of the engram cells precluded memory retrieval. Mice housed in an enriched environment had a slower rate of forgetting, and a brief reminder before the retrieval session promoted retrieval of a forgotten memory. Repeated reintroduction to the training context in the absence of objects accelerated forgetting. Additionally, activation of Rac1-mediated plasticity mechanisms enhanced forgetting, while its inhibition prolonged memory retrieval. The authors also reproduce the behavioral findings using a computational model inspired by Rescorla-Wagner model. In essence, the model proposes that forgetting is a form of adaptive learning that can be updated based on prediction error rules in which engram relevancy is altered in response to environmental feedback.

Strengths:<br /> 1) The data presented in the current paper are consistent with the authors claim that seemingly forgotten engrams sometimes remain accessible. This suggests that retrieval deficits can lead to memory impairments rather than a loss of the original engram (at least in some cases).

2) The experimental procedures and statistics are appropriate, and the behavioral effects appear to be very robust. Several key effects are replicated multiple times in the manuscript.

Weaknesses:<br /> 1) My major issue with the paper is the forgetting model proposed in Figure 7. Prior work has shown that neutral stimuli become associated in a manner similar to conditioned and unconditioned stimuli. As a result, the Rescorla-Wagner model can be used to describe this learning (Todd & Homes, 2022). In the current experiments, the neutral context will become associated with the unpredicted objects during training (due to a positive prediction error). Consequently, the context will activate a memory for the objects during the test, which should facilitate performance. Conversely, any manipulation that degrades the association between the context and object should disrupt performance. An example of this can be found in Figure 5A. Exposing the mice to the context in the absence of the objects should violate their expectations and create a negative prediction error. According to the Rescorla-Wagner model, this error will create an inhibitory association between the context and the objects, which should make it harder for the former to activate a memory of the latter (Rescorla & Wagner, 1972). As a result, performance should be impaired, and this is what the authors find. However, if the cells encoding the context and objects were inhibited during the context-alone sessions (Figure 5D) then no prediction error should occur, and inhibitory associations would not be formed. As a result, performance should be intact, which is what the authors observe.

What about forgetting of the objects that occurs over time? Bouton and others have demonstrated that retrieval failure is often due to contextual changes that occur with the passage of time (Bouton, 1993; Rosas & Bouton, 1997, Bouton, Nelson & Rosas, 1999). That is, both internal (e.g. state of the animal) and external (e.g. testing room, chambers, experimenter) contextual cues change over time. This shift makes it difficult for the context to activate memories with which it was once associated (in the current paper, objects). To overcome this deficit, one can simply re-expose animals to the original context, which facilitates memory retrieval (Bouton, 1993). In Figure 2D, the authors do something similar. They activate the engram cells encoding the original context and objects, which enhances retrieval.

Therefore, the forgetting effects presented in the current paper can be explained by changes in the context and the associations it has formed with the objects (excitatory or inhibitory). The results are perfectly predicted by the Rescorla-Wagner model and the context-change findings of Bouton and others. As a result, the authors do not need to propose the existence of a new "forgetting" variable that is driven by negative prediction errors. This does not add anything novel to the paper as it is not necessary to explain the data (Figures 7 and 8).

2) I also have an issue with the conclusions drawn from the enriched environment experiment (Figure 3). The authors hypothesize that this manipulation alleviates forgetting because "Experiencing extra toys and objects during environmental enrichment that are reminiscent of the previously learned familiar object might help maintain or nudge mice to infer a higher engram relevancy that is more robust against forgetting.". This statement is completely speculative. A much simpler explanation (based on the existing literature) is that enrichment enhances synaptic plasticity, spine growth, etc., which in turn reduces forgetting. If the authors want to make their claim, then they need to test it experimentally. For example, the enriched environment could be filled with objects that are similar or dissimilar to those used in the memory experiments. If their hypothesis is correct, only the similar condition should prevent forgetting.

3) It is well-known that updating can both weaken or strengthen memory. The authors suggest that memory is updated when animals are exposed to the context in the absence of the objects. If the engram is artificially inhibited (opto) during context-only re-exposures, memory cannot be updated. To further support this updating idea, it would be good to run experiments that investigate whether multiple short re-exposures to the training context (in the presence of the objects or during optogenetic activation of the engram) could prevent forgetting. It would also be good to know the levels of neuronal reactivation during multiple re-exposures to the context in the absence versus context in the presence of the objects.

4) There are a number of studies that show boundary conditions for memory destabilization/reconsolidation. Is there any evidence that similar boundary conditions exist to make an inaccessible engram accessible?

5) More details about how the quantification of immunohistochemistry (c-fos, BrdU, DAPI) was performed should be provided (which software and parameters were used to consider a fos positive neurons, for example).

6) Duration of the enrichment environment was not detailed.

Reviewer #3 (Public Review):

Summary: The manuscript by Ryan and colleagues uses a well-established object recognition task to examine memory retrieval and forgetting. They show that memory retrieval requires activation of the acquisition engram in the dentate gyrus and failure to do so leads to forgetting. Using a variety of clever behavioural methods, the authors show that memories can be maintained and retrieval slowed when animals are reared in environmental enrichment and that normally retrieved memories can be forgotten if exposed to the environment in which the expected objects are no longer presented. Using a series of neural methods, the authors also show that activation or inhibition of the acquisition engram is key to memory expression and that forgetting is due to Rac1.

Strengths:<br /> This is an exemplary examination of different conditions that affect successful retrieval vs forgetting of object memory. Furthermore, the computational modelling that captures in a formal way how certain parameters may influence memory provides an important and testable approach to understanding forgetting.<br /> The use of the Rescorla-Wagner model in the context of object recognition and the idea of relevance being captured in negative prediction error are novel (but see below).<br /> The use of gain and loss of function approaches are a considerable strength and the dissociable effects on behaviour eliminate the possibility of extraneous variables such as light artifacts as potential explanations for the effects.

Weaknesses:<br /> Knowing what process (object retrieval vs familiarity) governed the behavioural effect in the present investigation would have been of even greater significance.

The impact of the paper is somewhat limited by the use of only one sex.

While relevance is an interesting concept that has been operationalized in the paper, it is unclear how distinct it is from extinction. Specifically, in the case where the animals are exposed to the context in the absence of the object, the paper currently expresses this as a process of relevance - the objects are no longer relevant in that context. Another way to think about this is in terms of extinction - the association between the context and the objects is reduced results in a disrupted ability of the context to activate the object engram.

eLife assessment

This important study advances our understanding of the potential mechanisms of deep-sea adaptation and sheds light on the evolutionary history of hadal snailfish. Through comparative genomic analysis, the authors provide convincing evidence and propose hypotheses on the timing of trench colonization, population structure, and adaptations to the hadal snailfish genome in response to their environment.

Reviewer #1 (Public Review):

This manuscript provides an important case study for in-depth research on the adaptability of vertebrates in deep-sea environments. Through analysis of the genomic data of the hadal snailfish, the authors found that this species may have entered and fully adapted to extreme environments only in the last few million years. Additionally, the study revealed the adaptive features of hadal snailfish in terms of perceptions, circadian rhythms and metabolisms, and the role of ferritin in high-hydrostatic pressure adaptation. Besides, the reads mapping method used to identify events such as gene loss and duplication avoids false positives caused by genome assembly and annotation. This ensures the reliability of the results presented in this manuscript. Overall, these findings provide important clues for a better understanding of deep-sea ecosystems and vertebrate evolution.

Reviewer #2 (Public Review):

This paper presents improved, chromosome level assemblies of the hadal snailfish and Tanaka's snailfish. This is an extension and update of previous work from the group on the hadal snailfish genome. The chromosomal assemblies allow comparisons of genome architecture between a shallow water snailfish and the hadal snailfish to aid inference on timing of colonization of trenches and genomic changes that may have been adaptive for that move.

The comparisons in genomic architecture are compelling: genes present in Tanaka's snailfish that are lost in hadal snailfish that involve whole regions of the genome that no longer map even though adjacent regions do map between the species and across a large evolutionary distance to stickleback. Or genes that are duplicated in hadal snailfish but only appear as single copy in other fishes. The paper focuses on genes in the eye, in hearing, in circadian rythms, and in ROS scavaging. These are all functions that could play a role in adapting to the hadal environment.

The genomic comparisons all seem sound. Stylistically I would prefer if the authors could introduce the gene product and protein function every time they introduce a gene locus. They introduce a gene and general function, but don't usually note what the protein encoded by the gene is and what it's specific function is.

I found the paper generally well written, and the data compelling and creatively displayed.

Upon revision, the authors have commendably addressed all reviewer comments and added a slew of additional analyses. I find the paper stronger, better argued and have no further questions or comments.

eLife assessment

This is an important study that addresses a significant question in microbiome research. The authors provide convincing evidence that certain bacterial groups within the fly microbiome have critical functions for host development. Additionally, dietary aspects such as microbial community progression in a natural food source are integrated into their host-microbe interaction analyses.

Reviewer #1 (Public Review):

Summary:<br /> This valuable study analyzes the contribution of fungal and bacterial microbiota species to the growth and development of Drosophila. The authors use bacterial and fungal species associated with Drosophila in the wild to analyze their respective contributions in promoting larval growth in a decaying banana, mimicking the natural niche of fruit fly. They found that some fungal species and some fungus/bacteria combinations effectively promote growth by supplementing key branched amino acids in the food substratum. Production of these amino acids by Drosophila itself is not sufficient, and only fungal species that secrete these amino acids into the medium can sustain Drosophila growth. Thus, the authors clarify how facultative symbionts contribute to host growth in a natural setting by changing the food substratum in a dynamic manner.

Strengths:<br /> The natural setting developed by the authors to analyze the impact of the microbiota is clearly valuable, as is the focus on the role of fungal microbiota species. This complements studies of Drosophila microbiota that have previously focused on bacterial species and used a lab setting.

While there has been an extensive focus on obligate endosymbionts or gut symbionts, this study analyzes how facultative symbionts shape the food substratum and influence host growth.<br /> A last strength of this study is that it analyzes the contribution of Drosophila microbiota over a dynamic timeframe, analyzing how microbial species change in decaying fruit over time.

Weaknesses:<br /> 1) The author should better review what we know of fungal Drosophila microbiota species as well as the ecology of rotting fruit. Are the microbiota species described in this article specific to their location/setting? It would have been interesting to know if similar species can be retrieved in other locations using other decaying fruits. The term 'core' in the title suggests that these species are generally found associated with Drosophila but this is not demonstrated. The paper is written in a way that implies the microbiota members they have found are universal. What is the evidence for this? Have the fungal species described in this paper been found in other studies? Even if this is not the case, the paper is interesting, but there should be a discussion of how generalizable the findings are.

2) Can the author clearly demonstrate that the microbiota species that develop in the banana trap are derived from flies? Are these species found in flies in the wild? Did the authors check that the flies belong to the D. melanogaster species and not to the sister group D. simulans?

3) Did the microarrays highlight a change in immune genes (ex. antibacterial peptide genes)? Whatever the answer, this would be worth mentioning. The authors described their microarray data in terms of fed/starved in relation to the Finke article. This is fine they should clarify if they observed significant differences between species (differences between species within bacteria or fungi, and more generally differences between bacteria versus fungi).

4) The whole paper - and this is one of its merits - points to a role of the Drosophila larval microbiota in processing the fly food. Are these bacterial and fungal species found in the gut of larvae/adults? Are these species capable of establishing a niche in the cardia of adults as shown recently in by the Ludington lab (Dodge et al.,)? Previous studies have suggested that microbiota members stimulate the Imd pathway leading to an increase in digestive proteases (Erkosar/Leulier). Are the microbiota species studied here affecting gut signaling pathways beyond providing branched amino acids?

Reviewer #2 (Public Review):

Summary:<br /> In this manuscript, Mure et al investigated host-microbe interactions in wild-mimicked settings. They analyzed microbiome composition using bananas that had been fed on by wild larvae and found that the microbiota composition shifted from the early stage of feeding to the later stage of the fermentation process proceeded. They isolated several yeast and bacterial species from the food, and examined larval growth on banana-based food, mimicking natural setting where germ-free larvae cannot grow on it. The authors found that a yeast, Hanseniaspora uvarum, can support larval growth sufficiently, and insists that branched-chain amino acids (BCAAs) provided by the yeast may partly be accounted for the growth support. Interestingly, other isolated yeast species, some were non-supportive strains in terms of larval growth, can assist larval development when they were heat-killed. Besides, they showed that acetic acid bacteria, isolated from well-fermented banana (later-stage food), is sufficiently supportive but their presence depended on other microbes, lactic acid bacteria or yeast.

Strengths:<br /> So far, host-microbe studies using Drosophila melanogaster have relatively less focused on the roles of fungi and many studies used only "model" yeasts. In the experimental setting where natural conditions may be well mimicked, the authors successfully isolated wild yeast species and convincingly showed that wild yeast plays a critical role in promoting host growth. In addition, the authors provided intriguing observations that all of the heat-killed yeast promoted larval growth even though some of the yeast never support the development when they were alive, suggesting that wild yeasts produce the necessary nutrients for larval development, but the nutrients of non-supportive yeasts are not accessible to the host. This might be an interesting indication for further studies revealing host-fungi interactions.

Weaknesses:<br /> The experimental setting that, the authors think, reflects host-microbe interactions in nature is one of the key points. However, it is not explicitly mentioned whether isolated microbes are indeed colonized in wild larvae of Drosophila melanogaster who eat bananas. Another matter is that this work is rather descriptive. A molecular level explanation is missing in "interspecies interactions" between lactic acid bacteria (or yeast) and acetic acid bacteria that assure their inhabitation.

Reviewer #3 (Public Review):

Summary: In this manuscript, Mure et al. describe interactions between diet, microbiome, and host development using Drosophila as a model. By characterizing microbial communities in food sources and animals, the authors showed that microbial community dynamics in the food source is critical for host development.

Strengths: This is a very interesting study where authors managed to tackle a difficult question in an elegant manner. How the interactions between different microbial species within the microbiome shape host physiology is an area of great interest but equally challenging due to the complexity of intercellular interactions in complex, host-associated microbial communities. By using a simplified model and interrogating not only microbe-microbe and host-microbe interactions, but also the role played by diet, authors were able to identify significant interactions during fly development.

Weaknesses: All weaknesses observed in the original manuscript have been corrected in the current version.

eLife assessment

This study explores the activation mechanisms of members of the kinesin-3 family, demonstrating common and unique regulation modes with solid evidence. The findings make for valuable contributions to the field of kinesin activation and regulation.

eLife assessment

This study explores the activation mechanisms of members of the kinesin-3 family, demonstrating common and unique regulation modes with solid evidence. The findings make for valuable contributions to the field of kinesin activation and regulation.

eLife assessment

The manuscript investigates how the tandem reader domains in BPTF co-recognize two types of modifications present on histone tails, H3K4me3 and H3 acetylation. The authors interpret their results in the context of the conformational restriction of histone tails due to interactions with nucleosomal DNA. The findings contribute new insights into how the nucleosomal context regulates the recognition of multiple histone modifications by tandem reader domains and should be of interest to the broader chromatin field.

Reviewer #1 (Public Review):

The manuscript investigates the binding of PHD-BD, a tandem of reader domains in the C-terminus of BPTF, to modified histone tail peptides and nucleosomes. It focuses on the differences in binding affinity between peptide and nucleosome substrates for BPTF PHD-BD. Using the dCypher approach, they find that multi-modified peptide substrates (both acetylation and methylation) do not increase PHD-BD binding affinity. They argue that histone peptide substrates do not support the histone code model, which champions that multivalent engagement by PHD-BD with a multi-modified substrate would lead to stronger binding when compared to the engagement of each domain alone. In contrast, when using nucleosome substrates, even though the overall affinity is reduced, the affinity for H3K4me3triac (double modification) is tighter than either modification on its own. This is consistent with the histone code model.

A strength of the manuscript is that it further delineates the contribution of each domain by again using dCypher to compare peptide and nucleosome binding of the PHD and BD domains alone, as well as tandem domain constructs where each domain has been inactivated by a point mutation (W2891A for the PHD and N3007A for the BD). PHD alone had a lower affinity for nucleosomes than peptides overall. With peptide substrates, PHD had the highest affinity for H3K4me3 and reduced affinity for H3K4me3triac; while with nucleosomes this trend was reversed. BD alone showed an affinity for acetylated H3 and H4 peptides but surprisingly was unable to bind nucleosomes. PHD requires the combination of H3K4 methylation and H3 tail acetylation for binding, and when partnered with BD, which is not able to bind nucleosomes alone, interestingly confers specificity for K14ac and K18ac. The in vivo relevance is argued using CUT&RUN analysis.

NMR spectroscopy is further used to show that PHD-BD binds acetylated H3 in a multivalent manner while forming a unique complex with H3K4me3triac. Deleting the N-terminal A1 region of H3 abolishes the binding of PHD-BD, implying its importance for recognition. The authors also discuss a "fuzzy complex" that forms between H3 and DNA, as well as H4 and DNA, which explains the occlusion of histone tail accessibility in the nucleosome. By changing the sidechain charge, such as with PTMs, this interaction can be weakened and allow PHD in this case to bind to the modified H3 tail. Comparisons between spectra of the H4 tail, H4 tail with DNA, and the H4 tail in the nucleosome are made and used to argue for H4-DNA interactions in the nucleosome.

The conclusions of the manuscript are very well-supported by the data and reveal a lot of insight into how the two reader domains of BPTF interact with modified nucleosomes. In many places, however, the manuscript is written more generally as if the conclusions apply in all cases (e.g. the title, abstract, and introduction) and this remains to be determined. It is also overstated that there is a belief that peptides perfectly recapitulate nucleosomes. It should also be pointed out that the nucleosomes are multi-valent and the data cannot discriminate binding of a single PHD-BD to single or multiple tails, and that the work is limited as it is using a construct of BPTF and in fact, there is at least one other reader domain involved.

Reviewer #2 (Public Review):

This manuscript by Musselman and coworkers uses a commercial library of modified histone peptides and mononucleosomes to probe the substrate specificity of the PHD-bromodomain combination of the BPTF protein. They arrive at the conclusion that BPTF preferably binds H3K4me3 and H3K18ac in the H3 tail. By using NMR with lableled H4 protein in nucleosomes they show that the H4 tail interacts with DNA, which may limit its ability to interact with BPTF. Finally, experiments in cells demonstrate that BPTF, H3K4me3, and H3K18ac occupy overlapping regions of chromatin. The authors suggest that recruitment of BPTF to specific regions of chromatin is driven by the co-binding of H3K4me3 and H3K18ac by BPTF. This study is of interest to readers interested in understanding the functions of the BPTF protein in cells.

In this reviewer's opinion, the manuscript needs some revision and the inclusion of some missing information.

1) The authors seem to have overlooked the fact that mononucleosome substrates have been in use for determining the substrate specificity and mechanisms of quite a few enzymes that simply do not act on peptide substrates. For example, Dot1L doesn't do anything with peptides nor does COMPASS/Set1, both of which require intact nucleosomal substrates to measure their activity in response to ubiquitylated H2B. Thus, the authors' refinement of the "histone code hypothesis" is unnecessary and overdone. I would suggest that they instead cite examples where nucleosome substrates have provided answers that cannot be obtained from peptide substrates alone. For example, extensive work from the Muir and Allis labs.

2) Ruthenburg and Allis in Cell 2011 conducted similar experimentation and concluded that H3K4me3-H4K16ac is a modification state bound by BPTF in cells. They also showed co-localization in ChIP-seq experiments and demonstrated preferential pulldowns with BPTF and semisynthetic methylated and acetylated nucleosomes. The authors have entirely ignored these previous results in their own discussions. Readers would benefit from a side-by-side comparison of the two acetylation states to get a sense of which is a stronger interaction and why both seemingly correlate in CUTnRUN or ChIP-seq.

3) The idea that electrostatics may modulate tail accessibility was reported by Musselman and coworkers for the H3 tail in eLife 2018. Yet the PHD domain of BPTF clearly binds H3K4me3 in nucleosomes. In light of this prior observation, the NMR experiments now with H4 tail seem repetitive and not informative regarding BPTF's bromodomain binding. Also, missing is the effect of H4K16acetylation on H4 tail dynamics, which would be pertinent to addressing the hypothesis regarding the BPTF bromodomain binding H4K16ac

4) The NMR experiments are all undertaken with 150mM KCl with no NaCl present. While NMR experimental constraints are understandable, the authors should avoid sweeping statements from NMR experiments regarding the dynamism of histone tails in chromatin, unless specific experiments are cited/conducted to demonstrate the same in cells. Many factors may contribute to the exclusion of BPTF from modified histone tails in cells, including the binding of other reader proteins, and the precise genomic localization of these modifications vis-a-vis BPTF. The important role of anchoring proteins must also be taken into account when considering binding/non-binding of substrates by CAPs. Thus, the NMR experiments presented in the manuscript do not report on whether BPTF binds H4K16ac in cells or indeed in vitro. If the PHD domain is capable of ultimately binding the H3 tail despite the tail's fuzzy interaction with DNA, the question remains as to why the bromodomain may not do so for acetylated H4 tails?

This manuscript reports several interesting elements regarding BPTF regulation, but as presented it is missing some key comparisons with prior information that makes it hard for readers to assess the relevance of the results presented.

eLife assessment

This valuable paper examines the link between the neuropeptide cholecystokinin (CCK) and motor learning and neural plasticity in the motor cortex. While CCK was known to be involved in neural plasticity in other brain regions and behavioral contexts, this study is the first to provide evidence that CCK manipulation causes deficits in motor learning. However, the evidence for specific effects regarding behavior, activity, and pathways is currently incomplete.

Reviewer #1 (Public Review):

This paper combines an array of techniques to study the role of cholecystokinin (CCK) in motor learning. Motor learning in a pellet reaching task is shown to depend on CCK, as both global and locally targeted CCK manipulations eliminate learning. This learning deficit is linked to reduced plasticity in the motor cortex, evidenced by both slice recordings and two-photon calcium imaging. Furthermore, CCK receptor agonists are shown to rescue motor cortex plasticity and learning in knockout mice. While the behavioral results are clear, the specific effects on learning are not directly tested, nor is the specificity pathway between rhinal CCK neurons and the motor cortex. In general, the results present interesting clues about the role of CCK in motor learning, though the specificity of the claims is not fully supported.

Since all CCK manipulations were performed throughout learning, rather than after learning, it is not clear whether it is learning that is affected or if there is a more general motor deficit. Related to this point, Figure 1D appears to show a general reduction in reach distance in CCK-/- mice. A general motor deficit may be expected to produce decreased success on training day 1, which does not appear to be the case in Figure 1C and Figure 2B, but may be present to some degree in Figure 5B. Or, since the task is so difficult on day 1, a general motor deficit may not be observable. It is therefore inconclusive whether the behavioral effect is learning-specific.

The paper implicates motor cortex-projecting CCK neurons in the rhinal cortex as being a key component in motor learning. However, the relative importance of this pathway in motor learning is not pinned down. The necessity of CCK in the motor cortex is tested by injecting CCK receptor antagonists into the contralateral motor cortex (Figure 2), though a control brain region is not tested (e.g. the ipsilateral motor cortex), so the specificity of the motor cortex is not demonstrated. The learning-related source of CCK in the motor cortex is also unclear, since even though it is demonstrated that CCK neurons in the rhinal cortex project to the motor cortex in Figure 4D, Figure 4C shows that there is also a high concentration of CCK neurons locally within the motor cortex. Likewise, the importance of the projection from the rhinal cortex to the motor cortex is not specifically tested, as rhinal CCK neurons targeted for inactivation in Figure 5 include all CCK cells rather than motor cortex-projecting cells specifically.

CCK is suggested to play a role in producing reliable activity in the motor cortex through learning through two-photon imaging experiments. This is useful in demonstrating what looks like normal motor cortex activity in the presence of CCK receptor antagonist, indicating that the manipulations in Figure 2 are not merely shutting off the motor cortex. It is also notable that, as the paper points out, the activity appears less variable in the CCK manipulations (Figure 3G). However, this could be due to CCK manipulation mice having less-variable movements throughout training. The Hausdorff distance is used for quantification against this point in Figure 1E, though the use of the single largest distance between trajectories seems unlikely to give a robust measure of trajectory similarity, which is reinforced by the CCK-/- traces looking much less variable than WT traces in Figure 1D. The activity effects may therefore be expected from a general motor deficit if that deficit prevented the mice from normal exploratory movements and restricted the movement (and activity) to a consistently unsuccessful pattern.

Finally, slice experiments are used to demonstrate the lack of LTP in the motor cortex following CCK knockout, which is rescued by CCK receptor agonists. This is a nice experiment with a clear result, though it is unclear why there are such striking short-term depression effects from high-frequency stimulation observed in Figure 6A that are not observed in Figure 1H. Also, relating to the specificity of the proposed rhinal-motor pathway, these experiments do not demonstrate the source of CCK in the motor cortex, which may for example originate locally.

Reviewer #2 (Public Review):

This study aims to test whether and if so, how cholecystokinin (CCK) from the mice rhinal cortex influences neural activity in the motor cortex and motor learning behavior. While CCK has been previously shown to be involved in neural plasticity in other brain regions/behavioral contexts, this work is the first to demonstrate its relationship with motor cortical plasticity in the context of motor learning. The anatomical projection from the rhinal cortex to the motor cortex is also a novel and important finding and opens up new opportunities for studying the interactions between the limbic and motor systems. I think the results are convincing to support the claim that CCK and in particular CCK-expressing neurons in the rhinal cortex are critical for learning certain dexterous movements such as single pellet reaching. However, more work needs to be done, or at least the following concerns should be addressed, to support the hypothesis that it is specifically the projection from the rhinal cortex to the motor cortex that controls motor learning ability in mice.

1) Because CCK is expressed in multiple brain regions, as the authors recognized, results from the CCK knock-out mice could be due to a global loss of neural plasticity. In comparison, the antagonist experiment is in my opinion the most convincing result to support the specific effect of CCK in the motor cortex. However, it is unclear to me whether the CCK knock-out mice exhibited an impaired ability to learn in general, i.e., not confined to motor skills. For instance, it would be very valuable to show whether these mice also had severe memory deficits; this would help the field to understand different or similar behavioral effects of CCK in the case of global vs. local loss of function. If the CCK knock-out mice only exhibited motor learning deficits, that would be surprising but also very interesting given previous studies on its effect in other brain areas.

2) Related to my last point, I believe that normal neural plasticity should be essential to motor skill learning throughout development not just during the current task. Thus, it would be important to show whether these CCK knock-out mice present any motor deficits that could have resulted from a lack of CCK-mediated neural plasticity during development. If not, the authors should explain how this normal motor learning during development is consistent with their major hypothesis in this study (e.g., is CCK not critical for motor learning during early development).

3) Lines 198-200 and Fig. 2C: The authors found that the vehicle group showed significantly increased "no grasp" behavior, and reasoned that the implantation of a cannula may have caused injuries to the motor cortex. In order to support their reasoning and make the control results more convincing, I think it would be helpful to show histology from both the antagonist and control groups and demonstrate motor cortical injury in some mice of the vehicle group but not the antagonist group. Otherwise, I'm a bit concerned that the methods used here could be a significant confounding factor contributing to motor deficits.

4) The authors showed that chemogenetic inhibition of CCK neurons in the rhinal cortex impaired motor skill learning in the pellet-reaching task. However, we know that the rhinal cortex projects to multiple brain regions besides the motor cortex (e.g., other cortical areas and the hippocampus). Thus, the conclusion/claim that the observed behavioral deficits resulted from inhibited rhinal-motor cortical projections is not strongly supported without more targeted loss-of-function or rescue experiments.

It would also be very informative to the field to compare the specific behavioral deficits, if any, of inhibiting specific downstream targets of the rhinal CCK neurons. As a concrete example, the hippocampus may be involved in learning more sophisticated motor skills (as the authors pointed out in the Discussion) besides the motor cortex. It would be a critical result if the authors could either show or exclude the possibility that the motor learning deficits observed in CCK-/- mice were at least partially due to the inhibition of hippocampal plasticity. This echoes my earlier point (point 1) that it is unclear whether the effect of lacking CCK in knock-out mice is specific in the motor cortex or engages multiple brain regions.

Lastly, because Fig. 4 only showed histology in the rhinal and motor cortices, I am not sure whether the motor cortex solely receives CCK input from the rhinal cortex. A more comprehensive viral tracing result could be important to both supporting the circuit-specificity of the observed behavior in this study and providing a clearer picture of where the motor cortex receives CCK inputs.

5) I am glad to see the CCK4 rescue experiment to demonstrate the sufficiency of CCK in promoting motor learning. However, the rescue experiment lacked specificity: IP injection did not allow specific "gain of function" in the motor cortex but instead, the improved learning ability in CCK knock-out mice could be a result of a global effect of CCK4 across multiple brain regions. CCK4 injection specifically targeted at the motor cortex would be necessary to support the sufficiency of CCK-regulated neuroplasticity in the motor cortex to promote motor learning.