2,836 Matching Annotations
  1. Last 7 days
    1. Reviewer #1 (Public Review):

      Insect chemosensory receptors function as ligand-gated ion channels, while vertebrate and nematode chemoreceptors are G-protein coupled receptors. This difference led to multiple questions. One was whether there are vertebrate homologs of insect chemosensory receptors or receptor-like proteins. This manuscript of Benton and Himmel titled "Structural screens identify candidate human homologs of insect chemoreceptors and cryptic Drosophila gustatory receptor-like proteins" addressed this key question. First, it showed consistent results using the new tool for protein structure prediction, AlphaFold2, and confirmed the previously identified OR, GR, GRL, and DUF proteins in the 7TMIC superfamily as structural homologs of Orco. Then the authors identified human/vertebrate homologs: PHTF, but the function of this protein is not clear. Finally, they further expanded drosophilid-specific GRL proteins. It is great to see new members of the 7TMIC superfamily!

    2. Reviewer #2 (Public Review):

      The chemosensory systems of vertebrates and insects share a lot of structural and functional similarities. However, looking deeper into their molecular components reveals that these similarities likely represent remarkable examples of convergent evolution. For instance, receptor molecules that detect odors are unrelated between vertebrates and insects - vertebrates use G-protein coupled receptors while insects use ligand-gated ion channels. The latter was long regarded as specific to insects, but later studies identified putative homologs in other animals, (but not in vertebrates), some unicellular eukaryotes, and plants, raising the possibility that it is an ancient family. Still, the evolution of this protein family is notoriously difficult to analyze due to a high degree of sequence divergence between the genes despite the shared structural features of the proteins they encode. Here, the authors make use of the recent explosion of high-quality structural predictions produced by AlphaFold to conduct a deep search for previously undiscovered homologs of insect odorant and gustatory receptors.

      The study describes two major findings:<br /> 1. In contrast to the previous idea that vertebrates lack any homologs of the insect receptors, two proteins in vertebrates turn out to display a similar structure (Fig. 2B).<br /> 2. The authors describe a previously uncharacterized family of Drosophila "gustatory receptor-like" proteins with a putative function in chemoreception as suggested by expression data (Fig 3A, G).

      All analyses are extremely thorough, the logic of the narrative is very clear, and I find all conclusions well supported by data. The authors clearly favor a hypothesis that the family that includes insect odorant and gustatory receptors has a very deep evolutionary origin, and the homologous genes in other animals and non-animals have strongly diverged at the level of the sequence but retained detectable structural homology. However, they also acknowledge the limitations of some of their arguments and they discuss an alternative whereby the observed structural similarity is the result of convergence (which would be equally interesting). Overall, this study represents a major advance in our understanding of protein evolution and opens several avenues of research into the question of how functional demands steer the preservation of structural features of proteins while allowing their amino acid sequences to diverge.

    3. Reviewer #3 (Public Review):

      The Odorant Receptor and Gustatory Receptor families of 7 Transmembrane domain Ion channels were previously believed to have no family members in vertebrates. This paper uses the recent advances in protein folding prediction tools to first validate previous discoveries and confirm their approach with genes of known function. They then search for new family members and discover additional related genes in insects, where both ORs and IRs were previously known to exist. The most striking finding of the paper is that they identify genes related to these protein families in vertebrates, including humans. They propose a model for the evolution of this gene family based on their data.

      Overall, the data in this paper is strong, the data presentation is clear and the text is well-written and scholarly. The main weaknesses of the paper are that they have no functional analysis of any of their newly discovered proteins. This paper would benefit from experimental evidence that these are functional ligand-gated ion channels. The authors discuss this limitation at the end of the paper and note the challenges that conducting a functional analysis of these channels would represent. We agree that this could take years and that it is beyond the scope of the current paper, although we eagerly await a follow-up study where those experiments might be done.

    1. Reviewer #1 (Public Review):

      The authors sequence some of the oldest maize macroremains found to date, from lowland Peru. They find evidence that these specimens were already domesticated forms. They also find a lack of introgression from wild maize populations. Finally, they find evidence the Par_N16 sample already carried alleles for lowland adaptation.<br /> Overall I think this is an interesting topic, the study is well-written and executed for the most part.

      I have a variety of comments, most important of which revolve around methodological clarity. I will give those comments first.

      The authors should say in the Results section how "alleles previously reported to be adaptive to highlands and lowlands, specifically in Mesoamerica or South America" were identified in Takuno et al. 2015. What method was used? I see this partly comes in the Discussion eventually, but it would help to have it in the Results with more detail. The answer to this question would help a skeptical reader decide the appropriateness of the resource, given that many selection scans have been performed on maize genomes, the choice would ideally not be arbitrary.

      How were the covered putative adaptive SNPs distributed in the genome? Were any clustered and linked? The random sampled SNPs should be similarly distributed to give an appropriate null.

      How is genetic similarity calculated? It should be briefly described in the Results.

      It would help for the authors to state why they focus on Par_N16, I did not see this in my reading. Presumably, the analyses done are because of the higher quality data, but it would also help to mention why Par_N16 was sequenced in an additional run.

      In the sections on phylogenetic analysis, introgression, and D statistics, the authors could do a better job specifically indicating how the results support their conclusions.

    2. Reviewer #2 (Public Review):

      In this foundational article, the authors conduct an ancient DNA characterization of maize unearthed in archaeological contexts from Paredones and Huaca Prieta in the Chicama river valley of Peru. These maize specimens were recovered by painstakingly controlled excavation. Their context would appear to be beyond reproach though the individual radiocarbon determinations should be subject to further scrutiny.

      Radiocarbon determination for at least one of the maize cobs analyzed for aDNA is not a direct date, but dates associated material. The authors should provide a table of the direct dates on the specimens that were analyzed for ancient DNA. They should also specify the type and quantity of material sent and whether the cob, glumes, pith, or husks were submitted for dates. Include δ13C determinations for each cob with laboratory analysis numbers because there is justifiable concern that at least one of these cob dates has a δ13C value suggesting the material dated is not maize. Generally, the δ13C for maize ranges from -14 to -7. One or more of the specimens subjected to ancient DNA analysis in this paper have δ13C values far outside of this confidence interval.

      From the perspective of future scientists being able to repeat the analyses performed here, I would hope that all details of specimen treatment, extraction methods, read length and quality would need to be assiduously described. Routine analytical results should be reported so that comparisons with earlier and future results are facilitated, and not made difficult to decipher or search for.

      The aDNA analysis may or may not be affected by the anomalous δ13C values but one would anticipate that standard aDNA extraction and analysis protocols would provide a means by which the specimen's preservation of the specimens could be ascertained, for example, perhaps deamination and fragmentation rates could be compared or average read length evaluated with modern-contemporary materials so that preservation of the Paredones samples relative to that of maize in the CIMMYT germplasm bank and the San Marcos specimens investigated by the same researchers can be evaluated.

      The size and shape of the cobs depicted are similar to specimens occurring much later in Mesoamerican assemblages. For example, the approximate rachis diameter of the San Marcos specimens depicted by Valle-Bueno et al. (2016: Fig.1) averages less than 0.5cm while the specimens depicted in Valle-Bueno et al. (this manuscript) average 1.0 cm. The former - San Marcos - specimens are dated at 5300-4970 BP cal while the larger - Paredones - specimens date roughly 6777 - 5324 BP cal. The considerable disparity among the smaller more recent specimens compared to the very much larger putatively older specimens suggests the Paredones specimen's radiocarbon determinations are equivocal. The authors point this out but repeatedly state these cobs are the most ancient; a conundrum that should be resolved.

      I would suggest the authors consider redating these three specimens and if they do, hope that they will prepare the laboratory personnel with depositional environment information. MacNeish was skeptical about late dates on maize at Tehuacan, at first. Adovasio was initially certain about maize's associated dates from Meadowcroft. One would prefer to be reasonably certain the foundation this article creates is solid; the author's repeated reference to these cobs as the most ancient in the Americas should be reaffirmed so retraction will not be necessary.

    1. Reviewer #1 (Public Review):

      The authors investigated state-dependent changes in evoked brain activity, using electrical stimulation combined with multisite neural activity across wakefulness and anesthesia. The approach is novel, and the results are compelling. The study benefits from an in-depth sophisticated analysis of neural signals. The effects of behavioral state on brain responses to stimulation are generally convincing.

      It is possible that the authors' use of "an average reference montage that removed signals common to all EEG electrodes" could also remove useful components of the signal, which are common across EEG electrodes, especially during deep anesthesia. For example, it is possible (in fact from my experience I would be surprised if it is not the case) that under isoflurane anesthesia, electrical stimulation induces a generalized slow wave or a burst of activity across the brain. Subtracting the average signal will simply remove that from all channels. This does not only result in signals under anesthesia being affected more by the referencing procedure than during waking but also will have different effects on different channels, e.g. depending on how strong the response is in a specific channel.

    2. Reviewer #2 (Public Review):

      This study reports a novel role of thalamic activity in the late components of a cortical event-related potential (ERP). To show this association, the authors used high-density EEG together with multiple deep electrophysiological recordings combined with electrical stimulation of superficial and deep cortical layers. Stimulation of deep layers elicits a late ERP component that is closely related to bursts of thalamic activity during quiet wakefulness. This relationship is quite noticeable when deep layers of the cortex are stimulated, and it does depend on the arousal state, being maximal during quiet wakefulness, diminished during active wakefulness, and absent during anesthesia.

      The study is very well performed, with a high number of subjects and appropriate methodology. Performing simultaneous recording of EEG and several neuropixels probes together with cortical microstimulation is no small feat considering the size of the mouse head and the fact that mice are freely behaving in many of the experiments. It is also noticeable how the authors use a seemingly outdated technique (electrical microstimulation) to produce compelling and significant research. The conclusions regarding the thalamic contributions to the ERP components are strongly supported by the data.

      The spatiotemporal complexity is almost a side point compared to what seems to be the most important point of the paper: showing the contribution of thalamic activity to some components of the cortical ERP. Scalp ERPs have long been regarded as purely cortical phenomena, just like most EEGs, and this study shows convincing evidence to the contrary.

      The data presented seemingly contradicts the results presented by Histed et al. (2009), who assert that cortical microstimulation only affects passing fibers near the tip of the electrodes, and results in distant, sparse, and somewhat random neural activation. In this study, it is clear that the maximum effect happens near the electrodes, decays with distance, and is not sparse at all, suggesting that not only passing fibers are activated but that also neuronal elements might be activated by antidromic propagation from the axonal hillock. This appears to offer proof that microstimulation might be much more effective than it was thought after the publication of Histed 2009, as the uber-successful use of DBS to treat Parkinson's disease has also shown.

    1. Reviewer #1 (Public Review):

      This manuscript studies the representation by gender and name origin of authors from Nature and Springer Nature articles in Nature News. The representation of author identities is an important step towards equality in science, and the authors found that women are underrepresented in news quotes and mentions with respect to the proportion of women authors.


      The research is rigorously conducted. It presents relevant questions and compelling answers. The documentation of the data and methods is thoroughly done, and the authors provide the code and data for reproduction.


      The article is not so clearly structured, which makes it hard to follow. A better framing, contextualization, and conceptualization of their analysis would help the readers to better understand the results. There are some unclear definitions and wrong wording of key concepts.

    2. Reviewer #2 (Public Review):

      This paper set out to investigate disparities in how authors of scientific papers are quoted in the context of science journalism. Quotations, the authors argue, reveal who a science journalist approaches as a source and thus who is considered an expert. At the same time, quotation in the news legitimizes experts and signals the importance of their perspective and opinions. It is therefore important to identify disparities in a quotation, both as a matter of justice and to ensure the representation of diverse viewpoints in journalism.

      Here, the authors investigate disparities in quotation based on the gender and national origin of experts. They focus on science journalism in non-research articles published in the journal Nature. Articles are scraped from the Nature website and using established NLP tools the article content is parsed for quotations and the names of scientists being quoted. The gender and national origin of scientists are inferred based on their names and gendered pronouns used in the text. The rates of quotation based on gender/national origin are then compared to the demographics of authors (also inferred) of research articles published in Nature; this establishes a baseline to compare who is quoted vs. who is actually doing research. Based on these data, a variety of analyses are presented showing various aspects of bias and disparity in who is quoted in science journalism.

      From their analysis, the authors make the following claims:

      • Authors inferred as men were over-represented in quotations in journalistic Nature articles relative to their share of first and last authors in Nature.

      • A quotation is sharply trending towards gender parity, with variation by the type of article.

      • Authors with names inferred as originating from Celtic/English regions were over-represented, whereas authors with names inferred as originating from East Asia were heavily under-represented in quotations.

      • The representation of authors with inferred East Asian names has increased faster among the last authors of research articles in Nature than it has in a journalistic quotation.

      Claims 2-4 are solidly supported by the evidence presented in the manuscript. Claim 1 is supported by the evidence, but with some caveats. Support for Claim 1 depends on whether Nature's first or last authors are the most appropriate comparison set; if the last authors are the most appropriate, then Claim 1 only holds for 2005 through 2010. I expand on this point below.

      I praise the manuscript and the authors for their commitment to reproducibility. Supplied with the paper is all the data (where possible) and code necessary to reproduce the results, as well as a Docker image that ensures that it can be re-executed far into the future.

      The analyses conducted are methodologically rigorous. The authors provide bootstrapped confidence intervals for all analyzed values, choose appropriate baselines, and validate their name inference approach. In addition, I found their analysis comprehensive. By this I mean that they sufficiently explored their data to support their claims; nearly every caveat or limitation I could think of while reading was appropriately addressed either in the main or in a supplemental figure or table.

      While a good paper, it is not without weaknesses. The paper is generally well-written, and the visualizations do a good job of communicating results. There is, of course, room to improve on both. In some cases, the manuscript lacks consistency in terminology, and uses word choice that is strange (e.g., "enrichment" and "depletion" when discussion representation). While this paper is methodologically rigorous and professional in its presentation, I feel that the authors could have done a better job of interpreting and contextualizing their findings. Specifically, readers should be aware of the caveats regarding Claim 1 (listed above), the limits of generalizing these findings to other areas of science journalism, and a somewhat shallow discussion section that I believe detracts from the study's significance. I outline these points in more detail below.

      Despite these quibbles, the authors find solid support for their claims and achieve their goals. This paper, I believe will be of general interest to scientists and science communicators, to those interested in science communication as a field, to meta-scientists, and to those aiming to improve diversity and equity in the scientific process.

      Caveats to Claim Claim 1:

      One of the claims made by the authors (Claim 1) is that quotations in the dataset skew towards men. I find this true, but with two related caveats: that it depends on the choice of comparator set, and that it changes over time.

      The authors assess the representation of quotation by comparison to either Nature's first authors, or last authors. However, the authors do not discuss whether one is more appropriate, and what is implied if, say, quotations match the last author but not the first authors. In most scientific fields, the last author corresponds to the conceptual lead of a paper and is often the corresponding author who is most likely to be contacted to discuss the paper's significance. First authors, in contrast, will often represent the "driver" of the project-basically the person doing most of the actual work and is usually a student or more junior researcher. This distinction is important because cases could be made for either being a more appropriate comparator - last authors due to their seniority, first authors due to their closeness to the study, and (typically) greater diversity.

      The choice of comparator set becomes an issue because, as per Claim 2, the representation of women is increasing over time. Claim 1 only holds for the last authors from 2005 through 2010, and after 2018 women have higher representation given the demographics of the last authors. For the first authors, Claim 1 holds through 2017, after which they are representative or slightly over-representative of women authors.

      So while Claim 1 holds, it does not hold for all comparator sets and for all years. I don't think this is critical of the paper-the authors do discuss the trend in Claim 2-but interpretation of this claim should take care of these caveats, and readers should consider the important differences in first and last authorship.

      Generalizability to other contexts of science journalism:

      Journalistic articles in Nature may not be representative of all contexts of science journalism. Nature has a unique readership, consisting of scientists from many disciplines who have not only a generalist interest in science but also an interest in aspects of science as a profession. Science journalism as a whole, however, is part of the broader landscape of mainstream media, consisting of outlets such as ABC, BBC, and Scientific American. The audiences for these outlets will be more general, less interested in science as a career, and will likely have a different appetite for direct quotations and for more technical topics.

      This does not make the study bad. On the contrary, the author's focus on Nature allowed for many interesting analyses-but their findings should still be understood as coming from a specific context. While the authors outline many limitations of their study, they do not grapple with the limits of its generalizability, and what aspects of their analysis might translate to other contexts of science journalism. For example, part of the trend towards gender parity in a quotation is explained by the higher representation of women in the "Career Feature" article type. However, this article type will likely not be present in more general-interest contexts, which would affect the representation of women.

      Shallow discussion:

      I feel that the authors missed an opportunity to use their discussion to not only properly contextualize their results, but also explore their significance. In broad terms, there is literature on science journalism, its consequences for science, and the impact on public perceptions, as well as a continuous meta-discourse on journalistic ethics and best practices. The authors pay lip service to some of these themes but do little to actually place their findings in the broader discourse. Below, I provide a few specific points that could be further discussed:

      What might be the downstream impacts on the public stemming from the under-representation of scientists with East Asian names?

      The authors highlight gender parity in career features, but why exactly is there gender parity in this format of Representation in quotations varies by first and last author, most certainly as a result of the academic division of labor in the life sciences. However, what does it say about the scientific quotation that it appears first authors are more often to be quoted? Does this mean that the division of labor is changing such that the first authors are the lead scientists? Or does it imply that senior authors are being skipped over, or giving away their chance to comment on a study to the first author?

      Moreover, there are several findings in the study which are notable but don't seem to have been mentioned at all in the discussion.

      Below I highlight a few:

      • According to Figure 3d, not only are East Asian names under-represented in quotations, but they are becoming more under-represented over time as they appear as authors in a greater number of Nature publications.

      • Those with European names are proportionately represented in quotations given their share of authors in Nature. Why might this be, especially seeing as Anglo names are heavily over-represented?

    1. Reviewer #3 (Public Review):

      In this study, the authors present the first comprehensive transcriptome map of the human locus coeruleus using two independent but complementary approaches, spatial transcriptomics and single nucleus RNA sequencing. Several canonical features of locus coeruleus neurons that have been described in rodents were conserved, but potentially important species differences were also identified. This work lays the foundation for future descriptive and experimental approaches to understand the contribution of the locus coeruleus to healthy brain function and disease.

      This study has many strengths. It is the first reported comprehensive map of the human LC transcriptome, and uses two independent but complementary approaches (spatial transcriptomics and snRNA-seq). Some of the key findings confirmed what has been described in the rodent LC, as well as some intriguing potential genes and modules identified that may be unique to humans and have the potential to explain LC-related disease states. The main limitations of the study were acknowledged by the authors and include the spatial resolution probably not being at the single cell level and the relatively small number of samples (and questionable quality) for the snRNA-seq data. Overall, the strengths greatly outweigh the limitations. This dataset will be a valuable resource for the neuroscience community, both in terms of methodology development and results that will no doubt enable important comparisons and follow-up studies.

      Major comments:

      Overall, the discovery of some cells in the LC region that express serotonergic markers is intriguing. However, no evidence is presented that these neurons actually produce 5-HT.

      Concerning the snRNA-seq experiments, it is unclear why only 3 of the 5 donors were used, particularly given the low number of LC-NE nuclear transcriptomes obtained, why those 3 were chosen, and how many 100 um sections were used from each donor. It is also unclear if the 295 nuclei obtained truly representative of the LC population or whether they are just the most "resilient" LC nuclei that survive the process.

      The LC displays rostral/caudal and dorsal/ventral differences, including where they project, which functions they regulate, and which parts are vulnerable in neurodegenerative disease (e.g. Loughlin et al., Neuroscience 18:291-306, 1986; Dahl et al., Nat Hum Behav 3:1203-14, 2019; Beardmore et al., J Alzheimer's Dis 83:5-22, 2021; Gilvesy et al., Acta Neuropathol 144:651-76, 2022; Madelung et al., Mov Disord 37:479-89, 2022). It was not clear which part(s) of the LC was captured for the SRT and snRNAseq experiments.

      The authors mention that in other human SRT studies, there are typically between 1-10 cells per expression spot. I imagine that this depends heavily on the part of the brain being studied and neuronal density, but it was unclear how many LC cells were contained in each expression spot.

      Regarding comparison of human LC-associated genes with rat or mouse LC-associated genes (Fig. 2D-F), the authors speculate that the modest degree of overlap may be due to species differences between rodents and human and/or methodological differences (SRT vs microarray vs TRAP). Was there greater overlap between mouse and rat than between mouse/rat and human? If so, that is evidence for the former. If not, that is evidence for the latter. Also would be useful for more in-depth comparison with snRNA-seq data from mouse LC: https://www.biorxiv.org/content/10.1101/2022.06.30.498327v1.

      The finding of ACHE expression in LC neurons is intriguing, especially in light of work from Susan Greenfield suggesting that ACHE has functions independent of ACH metabolism that contributes to cellular vulnerability in neurodegenerative disease.

      High mitochondrial reads from snRNA-seq can indicate lower quality. It was not clear why, given the mitochondrial read count, the authors are confident in the snRNA-seq data from presumptive LC-NE neurons.

    2. Reviewer #1 (Public Review):

      Weber et al. collect locus coeruleus (LC) tissue blocks from 5 neurotypical European men, dissect the dorsal pons around the LC and prepare 2-3 tissue sections from each donor on a slide for 10X spatial transcriptomics. From three of these donors, they also prepared an additional section for 10x single nucleus sequencing. Overall, the results validate well-known marker genes for the LC (e.g. DBH, TH, SLC6A2), and generate a useful resource that lists genes which are enriched in LC neurons in humans, with either of these two techniques. A comparison with publicly available mouse and rat datasets identifies genes that show reliable LC-enrichment across species. Their analyses also support recent rodent studies that have identified subgroups of interneurons in the region surrounding the LC, which show enrichment for different neuropeptides. In addition, the authors claim that some LC neurons co-express cholinergic markers, and that a population of serotonin (5-HT) neurons is located within or near the LC. These last two claims must be taken with great caution, as several technological limitations restrict the interpretation of these results. Overall, there is limited integration between the spatial and single-nucleus sequencing, thus the data does not yet provide a conclusive list of bona fide LC-specific genes. The authors transparently present limitations of their work in the discussion, but some points discussed below warrant further attention.

      Specific comments:

      1) snRNAseq:

      a. Major concerns with the snRNAseq dataset are A) the low recovery rate of putative LC-neurons in the snRNAseq dataset, B) the fact that the LC neuron cluster is contaminated with mitochondrial RNA, and C) that a large fraction of the nuclei cannot be assigned to a clear cell type (presumably due to contamination or damaged nuclei). The authors chose to enrich for neurons using NeuN antibody staining and FACS. But it is difficult to assess the efficacy of this enrichment without images of the nuclear suspension obtained before FACS, and of the FACS results. As this field is in its infancy, more detail on preliminary experiments would help the reader to understand why the authors processed the tissue the way they did. It would be nice to know whether omitting the FACS procedure might in fact result in higher relative recovery of LC-neurons, or if the authors tried this and discovered other technical issues that prompted them to use FACS.

      b. It is unclear what percentage of cells that make up each cluster.

      c. The number of subjects used in each analysis was not always clear. Only 3 subjects were used for snRNAseq, and one of them only yielded 4 LC-nuclei. This means the results are essentially based on n=2. The authors report these numbers in the corresponding section, but the first sentence of the results section (and Figure 1C specifically!) create the impression that n=5 for all analyses. Even for spatial transcriptomics, if I understood it correctly, 1 sample had to be excluded (n=4).

      2) Spatial transcriptomics:

      a. It is not clear to me what the spatial transcriptomics provides beyond what can be shown with snRNAseq, nor how these two sets of results compare to each other. It would be more intuitive to start the story with snRNAseq and then try to provide spatial detail using spatial transcriptomics. The LC is not a homogeneous structure but can be divided into ensembles based on projection specificity. Spatial transcriptomics could - in theory - offer much-needed insights into the spatial variation of mRNA profiles across different ensembles, or as a first step across the spatial (rostral/caudal, ventral/dorsal) extent of the LC. The current analyses, however, cannot address this issue, as the orientation of the LC cannot be deduced from the slices analyzed.

      b. Unfortunately, spatial transcriptomics itself is plagued by sampling variability to a point where the RNAscope analyses the authors performed prove more powerful in addressing direct questions about gene expression patterns. Given that the authors compare their results to published datasets from rodent studies, it is surprising that a direct comparison of genes identified with spatial transcriptomics vs snRNAseq is lacking (unless this reviewer missed this comparison). Supplementary Figure 17 seems to be a first step in that direction, but this is not a gene-by-gene comparison of which analysis identifies which LC-enriched genes. Such an analysis should not compare numbers of enriched genes using artificial cutoffs for significance/fold-change, but rather use correlations to get a feeling for which genes appear to be enriched in the LC using both methods. This would result in one list of genes that can serve as a reference point for future work.

      c. Maybe the spatial transcriptomics could be useful to look at the peri-LC region, which has generated some excitement in rodent work recently, but remains largely unexplored in humans.

      3) The comparison of snRNAseq data to published literature is laudable. Although the authors mention considerable methodological differences between the chosen rodent work and their own analyses, this needs to be further explained. The mouse dataset uses TRAPseq, which looks at translating mRNAs associated with ribosomes, very different from the nuclear RNA pool analyzed in the current work. The rat dataset used single-cell LC laser microdissection followed by microarray analyses, leading to major technical differences in terms of tissue processing and downstream analyses. The authors mention and reference a recent 10x mouse LC dataset (Luskin et al, 2022), however they only pick some neuropeptides from this study for their analysis of interneuron subtypes (Figure S13). Although this is a very interesting part of the manuscript, a more in-depth analysis of these two datasets would be very useful. It would likely allow for a better comparison between mouse and human, given that the technical approach is more similar (albeit without FACS), and Luskin et al have indicated that they are willing to share their data.

      4) Statements in the manuscript about the unexpected identification of a 5-HT (serotonin) cell-cluster seem somewhat contradictory. Figure S14 suggests that 5-HT markers are expressed in the LC-regions just as much as anywhere else, but the RNAscope image in Figure S15 suggests spatial separation between these two populations. And Figure S17 again suggests almost perfect overlap between the LC and 5HT clusters. Maybe I misunderstood, in which case the authors should better clarify/explain these results.

    3. Reviewer #2 (Public Review):

      The data generated for this paper provides an important resource for the neuroscience community. The locus coeruleus (LC) is the known seed of noradrenergic cells in the brain. Due to its location and size, it remains scarcely profiled in humans. Despite the physically minute structure containing these cells, its impact is wide-reaching due to the known neuromodulatory function of norepinephrine (NE) in processes like attention and mood. As such, profiling NE cells has important implications for most neurological and neuropsychiatric disorders. This paper generates transcriptomic profiles that are not only cell-specific but which also maintain their spatial context, providing the field with a map for the cells within the region.


      Using spatial transcriptomics in a morphologically distinct region is a very attractive way to generate a map. Overlaying macroscopic information, i.e. a region with greater pigmentation, with its corresponding molecular profile in an unbiased manner is an extremely powerful way to understand the specific cellular and molecular composition of that brain structure.

      The technologies were used with an astute awareness of their limitations, as such, multiple technologies were leveraged to paint a more complete and resolved picture of the cellular composition of the region. For example, the lack of resolution in the spatial transcriptomic platform was compensated by complementary snRNA-seq and single molecule FISH.

      This work has been made publicly available and accessible through a user-friendly application such that any interested researcher can investigate the level of expression of their gene of interest within this region.

      Two important implications from this work are 1) the potential that the gene regulatory profiles of these cells are only partially conserved across species, humans, and rodents, and 2) that there may be other neuromodulatory cell types within the region that were otherwise not previously localized to the LC


      Given that the markers used to identify cells are not as specific as they need to be to definitively qualify the desired cell type, the results may be over-interpreted. Specifically, TH is the primary marker used to qualify cells as noradrenergic, however, TH catalyzes the synthesis of L-DOPA, a precursor to dopamine, which in turn is a precursor for epinephrine and norepinephrine suggesting some of the cells in the region may be dopaminergic and not NE cells. Indeed, there are publications to support the presence of dopaminergic cells in the LC (see Kempadoo et al. 2016, Takeuchi et al., 2016, Devoto et al. 2005). This discrepancy is further highlighted by the apparent lack of overlap per given Visium spots with TH, SCL6A2, or DBH. While the single-nucleus FISH confirms that some of the cells in the region are noradrenergic, others very possibly represent a different catecholamine. As such it is suggested that the nomenclature for the cells be reconsidered.

      The authors are unable to successfully implement unsupervised clustering with the spatial data, this greatly reduces the impact of the spatial technology as it implies that the transcriptomic data generated in the study did not have enough resolution to identify individual cell types.

      The sample contribution to the results is highly unbalanced, which consequently, may result in ungeneralizable findings in terms of regional cellular composition, limiting the usefulness of the publicly available data.

      This study aimed to deeply profile the LC in humans and provide a resource to the community. The combination of data types (snRNA-seq, SRT, smFISH) does in fact represent this resource for the community. However, due to the limitations, of which, some were described in the manuscript, we should be cautious in the use of the data for secondary analysis. For example, some of the cellular annotations may lack precision, the cellular composition also may not reflect the general population, and the presence of unexpected cell types may represent the accidental inclusion of adjacent regions, in this case, serotonergic cells from the Raphe nucleus.

      Nonetheless having a well-developed app to query and visualize these data will be an enormous asset to the community especially given the lack of information regarding the region in general.

    1. Reviewer #1 (Public Review):

      Ibar and colleagues investigate the function of spectrin in Drosophila wing imaginal discs and its effect on the Hippo pathway and myosin activity. The authors find that both βH-Spec and its canonical binding partner α-Spec reduce junctional localization of the protein Jub and thereby restrict Jub's inhibitory effect on Hippo signaling resulting in activation of the Hippo effector Yorkie regulating tissue shape and organ size. From genetic epistasis analysis and analysis of protein localization, the authors conclude that βH-Spec and α-Spec act independently in this regulation. The major point of this study is that the apical localization of βH-Spec and myosin is mutually exclusive and that the proteins antagonize each other's activity in wing discs. In vitro co-sedimentation assays and in silico structural modeling suggest that this antagonization is due to a competition of βH-Spec and myosin for F-actin binding.

      The study's strengths are the genetic perturbation that is the basis for the epistasis analysis which includes specific knockdowns of the genes of interest as well as an elegant CRISPR-based overexpression system with great tissue specificity. The choice of the model for such an in-depth analysis of pathway dependencies in a well-characterized tissue makes it possible to identify and characterize quantitative differences between closely entangled and mutually dependent components. The method of quantifying protein localization and abundance is common for multiple figures which makes it easy to assess differences across experiments.

      A weakness in the methodology is the link to tissue tension and conclusions about tissue mechanics. Methods that directly affect tissue tension and a more thorough and systematic application of laser ablation experiments would be needed to profoundly investigate mechanosensation and consequential effects on tissue tension by the various genetic perturbations. While the in-silico analysis of competing for F-actin binding sites for βH-Spec and myosin appears logical and supports the authors' claims, no point mutation or truncations were used to test these results in vivo. In its current structure the manuscript's strength, the genetic perturbations, is compromised by missing clear assessments of knockdown efficiencies early in the manuscript and other controls such as the actual effect on myosin by ROCK overactivation.

      The flow of experiments is logical and in general, the author's conclusions are supported by the presented data. The findings are very well embedded into the context of relevant literature and both confronting and confirming literature are discussed.

      The study shows how components of the cytoskeleton are directly involved in the regulation of the mechanosensitive Hippo pathway in vivo and thus ultimately regulate organ size supporting previous data in other contexts. The molecular mechanism regulating myosin activity by out-competing it for F-actin binding has been observed for small actin-binding proteins such as cofilin but is a new mode for such a big, membrane-associated actin-binding protein. This may inspire future experiments in different morphogenetic contexts for the investigation of similar mechanisms. For example, the antagonistic activity of βH-Spec and myosin in this tissue context might help explain phenomena in other systems such as spectrin-dependent ratcheting of apical constriction during mesoderm invagination (as the authors discuss). Against the classical view, the work shows that βH-Spec can act independently of α-Spec. Together the results will be of interest to the cell biology community with a focus on the cytoskeleton and mechanotransduction.

    2. Reviewer #2 (Public Review):

      Ibar and colleagues address the role of the spectrin cytoskeleton in the regulation of tissue growth and Hippo signaling in an attempt to elucidate the underlying molecular mechanism(s) and reconcile existing data. Previous reports in the field have suggested three distinct mechanisms by which the Spectrin cytoskeleton regulates Hippo signaling and this is, at least in part, due to the fact that different groups have mainly focused on different spectrins (alpha, beta, or beta-heavy) in previous reports.

      The authors start their investigation by trying to reconcile their previous data on the role of Ajuba in the regulation of Hippo signaling via mechanotransduction and previous observations suggesting that Spectrins affect Hippo signaling independently of any effect on myosin levels or Ajuba localization. Contrary to previous reports, the authors reveal that, indeed, depletion of alpha- and beta-heavy-spectrin leads to an increase in myosin levels at the apical membrane. Moreover, the authors also reveal that the depletion of spectrins leads to an increase in Ajuba levels.

      The authors suggest that Ajuba is required for the effect of beta-heavy spectrin. However, it is still formally possible that this could be a parallel pathway that is being masked by the strong phenotype of Ajuba RNAi flies.

      One of the major points of the manuscript is the observation that alpha- and beta-heavy-spectrin are potentially working independently and not as part of a spectrin tetramer. This is mostly dependent on the observation that alpha- and beta-heavy-spectrin appear to have non-overlapping localizations at the membrane and the fact that alpha- and beta-heavy-spectrin localize at the membrane seemingly independently. It is not entirely obvious that a potential lack of colocalization and the fact that protein localization at the membrane is not affected when the other partner is absent is sufficient to argue that alpha- and beta-heavy-spectrin do not form a complex. Moreover, it is possible that the spectrin complexes are only formed in specific conditions (e.g. by modulating tissue tension).

      If indeed spectrins function independently, would it not be expected to see additive effects when both spectrins are depleted?

      Related to the two previous points, the fact that the authors suggest that both alpha- and beta-heavy-spectrin regulate Hippo signaling via Ajuba would be consistent with the necessity of an alpha- and beta-heavy-spectrin complex being formed. How would the authors explain that both spectrins require Ajuba function but work independently?

      Another major point of the manuscript is the potential competition between beta-heavy-spectrin and myosin for F-actin binding. The authors suggest that there is a mutual antagonism between the two proteins regarding apical F-actin. However, this has not been formally assessed. Moreover, despite the arguments put forward in the discussion, it seems hard to justify a competition for F-actin when beta-heavy-spectrin seems to be unable to compete with myosin. Myosin can displace beta-heavy-spectrin from F-actin but the reciprocal effect seems unlikely given the in vitro data.

    1. Reviewer #1 (Public Review):

      The authors sought to assess how not only RNA but also protein changes across the developmental time course of cortical organoid development. The methods used included reporter lines to label progenitor and neuronal populations, RNA-sequencing, protein quantification using mass spectrometry, and analysis of these results. The primary findings included the identification of RNA sequences that impact translation, the most significant of which was a 5'-TOP cassette that is mediated by mTOR.

      Strengths of the paper include strong experimental design, replicates, and images to show the quality of the organoids used in the studies. Additionally, the analysis of elements regulating translation was strong, and the polysome experiments exploring an impact when TSC is deleted were interesting.

      Potential limitations include technical challenges related to the specificity of the reporters, ambiguity about the impact of normalization on the actual protein/RNA data, and potential over-interpretation of the TSC result to encompass all of the mTOR signalings.

      The paper validates already observed and documented results in translational regulation whereby RNA does not fully predict protein levels. The impact of the specific examples upon functional significance in cortical development is currently unclear but this work could set the stage for additional future impactful work.

    2. Reviewer #2 (Public Review):

      This work by Sidhaye, Trepte et al. systematically investigates the relationship between transcript and protein abundance across the genome in human neurogenesis. Through analysis of the transcriptome and proteome in brain organoids, they find that for specific gene modules, transcript and protein abundance are highly disconnected. While there are already several anecdotic examples of this phenomenon in the literature, highlighting the role of post-transcriptional gene regulation in corticogenesis, Sidhaye, Trepte et al. for the first time systematically explore the pervasiveness of this phenomenon in a genome-wide manner at different stages of human neurogenesis using a dual reporter cell line to isolate neural progenitor cells and neurons.

      The authors then focus on one of the modules that is characterized by the enrichment of the 5'TOP (terminal oligopyrimidine) motif in the 5'UTR of transcripts and enriched in ribosomal proteins and translation initiation factors. The authors show that partial inhibition of the translation of ribosomal genes in neural progenitor cells inhibits the translation of differentiation genes, a process that involves mTOR-mediated regulation.


      The integration of transcriptome and proteome data enables an unbiased systemic analysis revealing gene modules that follow similar trajectories, and as such may share common regulatory principles. For one of the modules, the authors dissect the posttranscriptional regulatory cascade using an elegant combination of fluorescent reporter human pluripotent stem cell lines in combination with gene knockouts.

      Overall, the data presented in this work is of a very high standard and supports the conclusions put forward by the authors. The processed omics data sets are made available via a Shiny app web interface for easy access and therefore promote exploration by the scientific community.


      This study uses a large range of specific reporter and knockout hPSC lines generated in the context of this work, however, very limited information is provided on these lines. For example, do the lines remain karyotypically normal throughout the targeting procedure? Does reporter gene expression faithfully recapitulate the activity of the promoters controlling their expression? Specifically, it appears that a significant GFP signal is detected within the neuronal layer (Figure 1B) and that there is a much larger double reporter-positive population than expected (Figure S2A).

      The authors propose that stress-associated translational regulation takes place in early neural progenitors, involving the sequestration of transcripts in stress granule-like structures. However, given that at least some human brain organoid protocols have been reported to lead to ectopic activation of cellular stress pathways (Bhaduri et al., Nature 2019), it would be desirable to see this aspect of the study confirmed in primary tissue (mouse or human).

    3. Reviewer #3 (Public Review):

      The manuscript by Sidhaye et al. aims to integrate proteomic and transcriptomic analyses of human stem cell-derived cortical brain organoids to identify post-transcriptional regulatory mechanisms during human cortical development. The authors use an innovative and useful dual-reporter strategy to isolate NPCs and neurons separately and integrate proteomic and transcriptomic analyses in each cell type. The data analysis is robust and identifies gene modules with cell class specificity.

      While there is no large overlap between the proteomic and transcriptomic datasets, the authors focus additional experiments on one candidate pathway, mTOR-mediated regulation of translation in progenitors, and validate this pathway's role in progenitor development.

      The authors also identified a stress-related role for processes in corticogenesis, although, without comparison to human tissue, it's possible that some of the results are due to the artificial nature of the organoids as they have been reported to have elevated stress (Bhaduri et al.,).

      The data is from organoids from one human stem cell line, the female H9 human embryonic stem cell line and so it is critical to validate the results on 1-2 additional stem cell lines, to rule out the possibility that these results are unique to this one cell line.

      The major concerns in this paper can be addressed through validation of the results in other systems (e.g. human tissue) or in additional cell lines.

      The results provide a valuable resource and address some of the limitations of current organoid and tissue single-cell data by focusing on proteomics.

    1. Reviewer #1 (Public Review):

      This paper presents the results of two fragment screens of PTP1B using room-temperature (RT) crystallography, and compares these results with a previously published fragment screen of PTP1b using cryo-temperature crystallography. The RT screen identified fewer fragment hits and lower occupancy compared to the cryo screen, consistent with prior publications on other proteins. The authors attempted to identify additional hits by applying two additional layers of data processing, which resulted in a doubling in the number of possible hits in one of the screens. Because I am not an expert in panDDA modeling, however, I am unable to evaluate the reproducibility and potential potency of these fragment hits as protein binders or their potential use as starting points for follow-up chemistry.

      The fragment library used in this study was larger than those used in previously published RT crystallography experiments. Among the cryo hits that bound in RT, most fragments bound in the same manner as they did in cryo, while some bound in altered orientations or conformations, and two bound at different locations in RT compared to cryo. This level of variability is not surprising. However, one fragment was observed to bind covalently to lysines in RT, even though it showed no density in the cryo crystallization attempt. It is unclear from the provided information whether this fragment decayed during storage or if the higher temperatures accelerated the covalent chemistry. The authors also observed temperature-dependent changes in the solvation shell, and modifications to the protein structure upon fragment binding, including a distal modification.

      The current version of the paper is somewhat repetitive in its presentation of the results and could be clearer in its presentation of the variations and comparisons of the two different protocols. It would be helpful to have a more concise summary of the differences between the two protocols in the current paper, as well as a discussion of how they compare to the protocol used in the previously published cryo-temperature fragment screen.

      While I appreciate the speculative nature of the discussion at the end of the paper, the evidence presented by the authors does not instil confidence that these results will correspond to meaningful binders that could be used to train future machine learning models. However, depending on the intended use, it may be acceptable to train ML models to predict expected densities under typical experimental conditions.

    2. Reviewer #2 (Public Review):

      The authors set out to understand how a room-temperature X-Ray crystallography-based chemical-fragment screen against a drug target may differ from a cryo screen. They carried out two room-temperature screens and compared the results with that of a cryo screen they previously performed. With a substantial set of crystallographic evidence they showed that the modes of protein-fragment binding are affected by temperature. The conclusion of the work is compelling. It suggests that temperature provides another dimension in X-ray crystallography-based fragment screening. In a practical sense, it suggests that room-temperature fragment screen is a promising new avenue for hit identification in drug discovery and for obtaining insights into the fragment binding. Room-temperature screening carries unique advantage over cryo screening. This work is confirmative to the notion, which seems not yet universally considered, that very weak protein-small molecule binding may be inherently fluid structurally, and that crystal structures of such weak binding, especially cryo structures, cannot be taken for granted without cross validation.

    1. Reviewer #1 (Public Review):

      Autoantibodies to nuclear proteins are commonly associated with autoimmune conditions. Since their discovery, several reports have suggested that T-follicular regulatory cells (Tfr) Tfr cells have the capacity to preferentially suppress autoimmune antibody responses. Tfr have a TCR repertoire strongly skewed to self-antigens and in this report Ke et al. probe the idea that Tfr directly recognize nuclear proteins and inhibit nuclear protein specific B-cells. They find that vaccination of mice with an ongoing GC reaction to a foreign antigen using nuclear proteins causes expansion of Tfr and a Tfr dependent inhibition of the germinal center. Overall, this is a well written paper that significantly advances the idea that Tfr can control autoreactive B-cells in a selective manner. Most experiments are convincing. Some of the novel methods regarding the use of nuclear proteins during sequential vaccinations in mice or Tfr-B-cell doublet formation will be of interest to members of the same fields.

      A primary weakness of the paper is that despite detailed analysis of cells involved in antibody production, there is very little analysis of the antibodies themselves. Particularly when Tfr deficient mice are used in figure 5 analysis of both anti-SA and anti-NucPr antibodies between the Tfr cKO and other groups would significantly advance the findings.

    2. Reviewer #2 (Public Review):

      In this study, the authors developed a mouse model to specifically investigate whether GC B cells that present nuclear protein (NucPr) could be specifically suppressed by Tfr cells. Most current mouse models that have been used in investigating Tfr functions are based on the overall readout of autoantibody production in the scenario of loss-of-function of Tfr cells. The proposed model of gain-of-function of Tfr cells is novel and valuable.

      The authors mainly compared two boosting immunizations by Strepatividin (SA) alone or SA-conjugated with nuclear proteins (SA-NucPr) and demonstrated SA-NucPr boosting immunization was able to expand Tfr cells, suppress overall and SA-specific GC/memory/plasma cell responses. The results are mostly convincing.

      One major concern is the conditions and controls used in the study. The control group (SA boosting immunization) would have enhanced T and B cell responses by this boosting. Unfortunately, there was no non-boosting control group so the level was unclear. It is therefore to strictly match such boosting condition in the SA-NucPr group. Notably, both SA and SA-NucPr were used at 10ug for boosting immunization. Considering NucPr were comparable or much larger (Nucleosome, about 200KDa) than SA (about 60KDa), the dose of SA in the SA-NucPr group was far less than that in the SA group. Due to this cavity, it is difficult to judge the difference between two groups was due to less SA boosting immunization or NucPr-induced Tfr function. This was a fundamental issue weakens the conclusion.

      The single cell analyses clearly demonstrated the expansion of Tfr clones. It remains unclear why other Treg populations other than Tfr cells were not expanded? The Treg cells in the CXCR5intPD-1int population were recently activated and should be able to respond to the boosting immunization. On an alternative explanation, the changes in Tfr cells could be indirectly driven by the changes in Tfh cells. For example, Tfh can produce IL-21 and restrict Tfr expansion (Jandl C, et al.2017). This could be the case of the reduction in Tfr cells in the SA-OVA group as compared to the SA group.

    1. Reviewer #1 (Public Review):

      The authors have succeeded in demonstrating that they can further extend the methodology and value of Mendelian randomization by combining their two recently developed novel approaches to Mendelian randomization studies (1) Lifecourse MR which relates the genetic instruments to the outcome, eg obesity, at different stages of life eg childhood and adulthood and (2) Tissue partitioned MR to determine if the genetic instruments have different effects on different tissues such as the brain and adipose tissue. They have successfully combined these two to investigate the influence of adiposity on circulating leptin to demonstrate the value/proof of concept of these techniques in extending the use of MR.

      This is a very clearly presented and well-conducted work showing both new methodology and clear-cut results on the impact of adiposity at age 10 and in middle life and the weight gain in between on leptin levels and that the effect is mediated via the brain. They show that childhood obesity has a direct effect on leptin levels at age 10 years and an indirect effect on adult leptin along a causal pathway involving adulthood body size. They also show that BMI exerts its effect on leptin levels at both life stages via brain-tissue-mediated pathways.

      Major strengths are the well-characterized data sets used and in particular, having a comprehensive data set for children and the successful use of a new approach to address a complex issue. There are no major weaknesses

      The authors have achieved their two aims - the use of the new methodology and its application to the specific issue to demonstrate how it works ie proof of concept. Their results support their conclusions.

      The main advance here is a demonstration of a new further enhanced approach to Mendelian randomization. This is likely to end up being used by other researchers to address complex questions.

    2. Reviewer #2 (Public Review):

      In this proof-of-concept study, Richardson et al explore lifecourse effects of adiposity on leptin levels using life course Mendelian randomization and perform a tissue-partitioned MR to study the effects of tissue-specific BMI genetic instruments on leptin levels. The methods are solid and they have been nicely applied in the context of the present study. The results are important, revealing differences in the impact of adiposity on leptin levels in childhood vs adulthood, and highlighting the importance of the adipose-brain pathway in leptin homeostasis.

      Additional MR analyses are suggested to explore bidirectional associations between leptin levels and adiposity, due to the interrelation of these two markers. Also, the fact that the MR instruments for childhood adiposity are based on self-reported body size, while the MR instruments for adult adiposity are based on measured adult BMI should be highlighted in the manuscript, and the possible impact of this in the findings should be discussed.

      In summary, this important study is a proof of concept of life course and tissue-partitioned MR, while providing interesting insights into the regulation of leptin homeostasis by adiposity in different life stages.

    1. Reviewer #1 (Public Review):

      This work endeavours to delineate the relationship between IL-7R+ and IL-7R- ILC1 in the liver. They elegantly utilize a PLZF reporting system to identify the progenitor/product relationship between ILC subsets and show that ILC1s emerge separately from NK cells and LTi cells.

      Furthermore, ILC1 are enriched in the liver. Extending this work in Rora-deficient mice, they demonstrate that over time, these cells are poorly replaced in the liver, and that IL-7R+ cells did not convert into IL-7R- cells at steady-state. Fetal liver IL-7R+ ILC1s were shown to partially contribute to mature ILC1s. Interestingly, they show that there were localization changes between ILC1 precursors and mature ILC1s in the liver. They then analysed the factors that might underpin these different localizations by examining IL-15 which is highly produced by macrophages and endothelial cells. They identify that hepatocyte-derived IL-15 supports the development of 7R− ILC1s in the parenchyma to maintain adult 7R− ILC1s within the sinusoids. Finally, the authors addressed the discrepancy in understanding of cytotoxicity expressed by ILC1s and identify that constitutive expression of mTOR was necessary to effect this function, thereby providing a mechanistic explanation for variable cytotoxicity observed in other studies. Overall, this study advances our knowledge of how ILC1 are generated and maintained in the liver, and how they acquire their effector functions.

    2. Reviewer #2 (Public Review):

      The authors are aiming to characterize the developmental process and functional heterogeneity of liver ILC1s. The role of liver ILC1's in human health is still unknown. ILC1's are abundant in fetal liver and decline throughout development into adults.

      The authors have gone to great lengths to establish the relationship between IL-7R expression and ILC1 ontogeny.

      The study provides insight into the complex ILC1 ontogeny by revealing relationships among heterogenous ILC1 subsets.

      The data suggests intrinsic cytotoxic programs of 7R− ILC1s differ to NK cells, proposing them as critical steady-state sentinels against infection prevention and tumor surveillance.

      The big unresolved question is why ILC1 dominate fetal innate lymphocytes versus NK cells in adult life.

    1. Reviewer #1 (Public Review):

      In this manuscript, Scagliotti and colleagues investigate the role of Dlk1 in regulating pituitary size in multiple mouse models with different Dlk1 gene dosages in order to understand the mechanisms of organ size control. They find that overexpression of Dlk1 leads to pituitary overgrowth and loss of Dlk1 causes undergrowth. Authors find two compartments of Dlk1 expression in the pituitary, in the marginal zone stem cell compartment and the parenchymal differentiated cell compartment, and by combing genetic mouse models show that a specific interaction of Dlk1 expression in both regions is necessary to affect pituitary organ size. They present to suggest that Dlk1 may repress Wnt signaling during development to control a shift from progenitor proliferation to differentiation. The data are meticulous, high quality, and clear.

      I have some questions about the interpretation of their data regarding the mechanism of Dlk1 regulation of pituitary organ size, as I believe there could be potential alternative explanations for their observations:

      I was wondering about the cause of the enlargement of the pituitary gland in Fig 1E, and whether it is caused by an increased number of cells (hyperplasia), an increased cell size (hypertrophy), or both. Line 104 states it is hyperplasia, and that cell size was not affected in WT-TG ('not shown', line 121). However, line 444 says the TG is hypertrophic. It would be good if the authors could elaborate on this and show or state how cell size was determined. Figs 5/6 show that WT-Tg proliferation is generally similar to WT, which suggests the increased size is not hyperplasia. It would be good to know whether this is correct. Some previous studies have shown that in pregnancy, lactotroph hypertrophy can be responsible for pituitary enlargement without hyperplasia (Castrique 2010, Hodson 2012).

      Related to the organ size question above, I had a question about the cell number and proportions in Fig 1D/E/F, which shows the maintenance of endocrine cell proportions and an increase in the volume of ~30% in WT-Tg. For the cell proportions to be maintained, I thought the increase in volume per cell type (Fig 1G) would therefore have to also increase proportionally in every cell type, while 1G appears to show an increase in GH (sig) and PRL/TSH cells (ns). It would be good if the authors could discuss this briefly.

      This study is impactful and will be of interest to several research communities, including those interested in pituitary development and function, organ size control, and gene imprinting mechanisms.

    2. Reviewer #2 (Public Review):

      Scagliotti et al address how organ size is regulated by imprinted genes. Using a series of mouse models to modulate the dosage of the paternally expressed gene, Dlk1, the authors demonstrate that DLK1 is important for the maintenance of the stem cell compartment leading to the growth of the pituitary gland and the expansion of growth hormone-producing cells. The authors show that overexpression of Dlk1 leads to pituitary hyperplasia while deletion of the paternal allele leads to reduced pituitary size. Reduced pituitary size is accompanied by reduced cell proliferation in the cleft at e13.5 and an increase in the number of POU1F1+ cells, suggesting that loss of Dlk1 alters the balance between the number of cells remaining in the replicating stem cell pool and those differentiating into the POU1F1 lineage. An elegant caveat of this paper is the rescue of Dlk1 expression in the population of cells expressing Pou1f1 but not in SOX2+ stem cells. Expression of Dlk1 only in POU1F1+ cells is not sufficient to rescue pituitary size. The authors suggest that this is because DLK1 must be present in stem cells which then activate paracrine WNT signaling to promote cell proliferation in POU1F1+ cells.


      This is an important study that provides a mechanistic understanding of how the imprinted gene, Dlk1, regulates organ size. The study employs an elegant experimental design to address the dosage requirement for Dlk1 in regulating pituitary gland size. Rescuing Dlk1 in the POU1F1+ cells, but not the marginal zone SOX2+ cells provides intriguing results about a possible role for DLK1 in paracrine signaling between these different pituitary cell types. The study uses publicly available scRNAseq and ChIPseq data to further support their findings and identify Dlk1 as a likely target of POU1F1.


      The study only analyzes females for the adult time point. For embryonic and postnatal time points sexes are pooled. Gender differences in pituitary gene expression embryonically or postnatally could potentially affect experimental outcomes.

      The authors employ a mouse model that rescues Dlk1 expression starting at e15.5 in POU1F1+ parenchymal cells but not in marginal zone stem cells. Rescuing Dlk1 expression in a specific population of cells is one of the strengths of this study. Based on this information and the fact that overexpression of Dlk1 leads to increased pituitary size, the authors suggest that DLK1+ marginal zone stem cells and DLK+ parenchymal cells may interact to promote postnatal proliferation. However, the ability to more carefully parse out the complex spatial and temporal contributions of DLK1 to pituitary size would be enhanced by the addition of a mouse model that rescues Dlk1 expression only in SOX2+ cells and a model that rescues expression in both stem cells and POU1F1+ cells.

    1. Reviewer #1 (Public Review):

      The authors of this study sought to test whether the optogenetic induction of context-related freezing behavior could be enhanced by synchronizing light pulses to the ongoing hippocampal theta rhythm. Theta is a hippocampus-wide oscillation that strongly modulates almost every cell in this structure, which suggests that causal interventions locked to theta could have a more pronounced impact than open-loop ones. Indeed, the authors found that activating engram-associated dentate gyrus (DG) neurons at the trough of theta resulted in an increase in freezing relative to baseline when averaging across all stimulation epochs. In contrast, open-loop stimulation and peak-locked stimulation had weaker effects. Analysis of local field potentials showed that only the theta-locked stimulation facilitated coupling between theta and mid-gamma, indicating that this manipulation likely enhances the flow of activity from DG to CA1 via CA3 (as opposed to promoting transmission from entorhinal cortex to CA1). Previous results from mice, rats, and humans support the hypothesis that memory encoding and recall occur at distinct phases of theta. This work further strengthens the case for phase-specific segregation of memory-related functions and opens up a path toward more precise clinical interventions that take advantage of intrinsic theta rhythm.


      This study recognizes that, when artificially reactivating a context-specific memory, the brain's internal context matters. In contrast to previous attempts at optogenetically inducing recall, this work adds an additional layer of precision by synchronizing the light stimulus to the ongoing theta rhythm. This approach is more challenging, because, in addition to viral expression and bilateral optical fibers, it also requires a recording electrode and real-time signal processing. The results indicate that this additional effort is worth it, as it results in a more effective intervention.

      The findings on theta-gamma cross-frequency coupling suggest a possible mechanism underlying the observed behavioral effects: trough stimulation enhances DG to CA1 interactions via CA3. LFP recordings showed that stimulation increases the coupling between theta and mid-gamma (though not in all mice), and the percentage of freezing during reactivation is correlated with the gamma modulation index.


      Given the precision of the intervention being performed, one might expect to see a stronger behavioral impact. Instead, the overall effect is subtle, and quite variable across mice. Looking at individual data points, the biggest overall increase in freezing actually occurred in 2 mice during the 6 Hz stimulation condition. Furthermore, trough stimulation decreased freezing in 3 mice This is not a weakness in itself; rather, the weakness lies in the lack of an attempt to make sense of this variability. There are a number of factors that could explain these differences, such as viral expression levels, electrode/fiber placement, and behavior during baseline. There is of course a risk of over-interpreting results from a few mice, but there is also a chance that the results will appear more consistent after accounting for these additional sources of variation.

      While trough-locked optogenetic stimulation significantly increases freezing, the effects are much weaker than placing the mouse in the actual fear-conditioned context (average time freezing of 15% vs. 50%). The discussion would benefit from additional treatment of ways to further increase the specificity and effectiveness of artificial memory reactivation.

      Using an open-source platform (RTXI) for real-time signal processing is commendable; however, more work could be done to make it easier to adopt these methods and make them compatible with other tools. The RTXI plugin used for closed-loop stimulation should be fully documented and publicly available, to allow others to replicate these results.

    2. Reviewer #2 (Public Review):

      In this manuscript, Rahsepar et al. test the hypothesis that the precise timing of engram cell activation in relation to the phase of hippocampal theta oscillations plays a causal role in recall. This hypothesis is derived from theories (e.g. the SPEAR model) positing that the hippocampus segregates information for memory encoding and retrieval in time and that separation is organized across the many neurons and subregions of the hippocampus by theta oscillations. They test this hypothesis using stimulation of dentate gyrus neurons active during the encoding of fear memory. Using closed-loop stimulation that they developed, the authors stimulate these dentate engram cells at different phases of theta to measure freezing behavior to determine if the fear memory is recalled. They compare this stimulation to stimulation at the same average frequency regardless of theta phase, or at a constant 20Hz, in line with prior research, as control conditions. The authors use an elegant within animal design. They find that stimulating at the theta phase when CA3 inputs most strongly influence CA1 leads to significant increases in freezing (relative to baseline), while none of the other stimulation conditions have significant effects on freezing. They then show that this stimulation also causes increases in gamma modulation by theta, which is correlated with learning in prior work. However, the gamma that is theta-modulated appears to be medium gamma which is not associated with CA3 inputs to CA1. Overall, the study is well-designed and well-controlled. The stimulation effects at the "best" theta phase are modest but do appear different than the other conditions. It is unclear why the authors chose to stimulate in dentate and not CA3 as the SPEAR hypothesis centers around CA3 and EC inputs to CA1. Furthermore, I wonder if the freezing behavior itself confounds the detection of the theta phase. Finally, some of the statistical analyses require controlling for multiple comparisons.

    3. Reviewer #3 (Public Review):

      The paper by Rahsepar et al. employed a closed-loop optogenetic approach to stimulate mouse dentate gyrus (DG) 'engram cells' at different phases of the ongoing theta rhythm. While stimulation of DG engram cells in fear conditioning paradigms has been conducted several times before (with similar results to those presented here), the current approach constitutes a significant methodological improvement over typical 'open loop' designs. The authors first characterize the performance of their closed-loop theta phase prediction method and show that it outperforms constant frequency stimulation in achieving a theta phase-specific stimulation, albeit with some limitations. A prominent theory in the field has proposed that memory encoding and recall preferentially take place at the peak and trough of theta respectively. Based on this framework, the authors compared the behavioral and physiological effects of stimulating engram cells at either the theta peak or trough as well as with constant frequencies. They found that, as predicted by the theory, stimulation at the theta through was the most effective in inducing enhanced fear memory recall (measured as freezing during re-exposure to a neutral context). Finally, the authors examined theta-gamma hippocampal LFP dynamics to provide physiological support for the observed behavioral differences of the different stimulation patterns.

      Overall, this work illustrates an interesting methodological development that will be of relevance for future studies conducting manipulations of engram cells and provides additional experimental support for an influential theory in the memory field. Experiments are well conducted and the results presented support the main interpretation of the authors, but several aspects of the interpretation and discussion of the work need to be improved. Likewise, several aspects of data analysis and interpretation, in particular in reference to hippocampal oscillations and regional differences need to be improved.

    1. Reviewer #1 (Public Review):

      In this article, Prassad and colleagues describe a new mechanism involved in the elimination of misspecified/mislocated cells in the wing imaginal disc. This study follows a previous study from the same group (Bilmeier et al. Curr Biol 2016) which showed that a large panel of genetic backgrounds changing locally cell fate can trigger aberrant sorting of the misspecified cells triggered by the increased of contractility at clone interfaces. This process was suggested to directly participate to clone elimination below a certain clone size. However, the mechanism involved in apoptosis induction was not really studied per se. Here, they use similar genetic backgrounds and showed that JNK activation occurs specifically at the interface of the misspecified clones on both side (inside and outside the clone) hence leading to a local increase of cell death both in the WT and misspecified cells. This local activation of cell death participates to clone elimination, although the authors also delineate an alternative mechanism of death induction in the center of the clone that may correlate with the local buckling and the deformation. Importantly, this mechanism seems quite specific of these misspecified backgrounds and is unrelated to other more classical cell competition scenarios which trigger the elimination of Minute mutant (affecting ribosomes) or based on differential levels of Myc.<br /> The model proposed is interesting and clearly delineate a distinctive feature of this quality control mechanism which triggers local JNK activation. It is based on solid genetic evidences and use a large panel of genetic backgrounds and careful quantifications. The demonstration is overall very convincing. Moreover, these results provide a novel perspective for the field of cell competition and quality control mechanism which has been dominated by the concept of absolute fitness, which is not at all required in this context (where both WT or altered cells can be eliminated provided they are in minority in the tissue).

      Admittedly, the unicity and novelty is bit tuned down by former studies showing similar patterns of JNK activity upon local distortion of morphogens (so called morphogenetic apoptosis, Adachi-Yamada and O'Connor Dev Biol 2002), or the pattern of JNK activation observed near polarity mutant clones (Ohsawa et al, Dev Cell 2011) suggesting that this bilateral JNK activation might not be completely unique to these contexts. But non of these studies characterised such large range of genetic backgrounds and this study clearly provide new mechanistic insights.

      It is important to note that at this stage, it is not clear whether there is any link between the sorting behaviour and the activation of JNK (they could be both activated by unknown upstream factors), while the terminology "interfacial contractility" used to define this type of clone elimination may convey the idea that this is the most upstream factor in the process. Also further quantifications may be required to see to which extend JNK activation is indeed restricted to cell directly contacting clone border and also to support the final proposed model suggesting that the number of contact could influence the levels of JNK (actually alternative models could also explain why smaller clones get eliminated). Finally, while the JNK levels clearly influence death in the clone, further experiments may be required to test how the line of JNK activation in WT cells contribute to their death and their elimination similar to mispecified cells, specially in the context where the majority of tissue is covered by mispecified clones.

    2. Reviewer #2 (Public Review):

      Prasad et al investigate mechanisms of interface contractility that occur at borders between cells of different specification states. Cells of different specification states typically sort out, minimizing interface contact, in association with increased junctional contractility that can be visualized by phalloidin labeling. Here, Prasad et al show, for multiple different examples of specification and/or signaling states, that bilateral activation of JnK flanks these interfaces, which are associated with elevated rates of Jnk-dependent apoptosis. Blocking Jnk activity does not seem to affect phalloidin labeling, however, placing interphase contractility upstream or parallel to Jnk activity and apoptosis. Interestingly, activated Ras[V12] is an exceptional case where interphase contractility and bilateral Jnk activation occur without elevated apoptosis. Indeed, RasV12 can suppress apoptosis associates with interfaces between other distinct cell types. Prasad et al suggest that this property of Ras[V12] activated cells may underlie their oncogenic potential in mammals. These are potentially interesting observations that address what happens when cell of disparate signaling and/or specification states are opposed. In principle, they could be of interest both to developmental biologists, from the perspective of correction of developmental errors, and to cancer biologists, from the perspective of eliminating precancerous cells.

      It is not clear how much advance is represented over the prior description of 'morphogenetic apoptosis', in which bilateral Jnk activity was also an integral part (Adachi-Yamada and O'Connor, Devl Biol vol251 pp74-90 2002). There is little new mechanistic insight provided here. As such, the observations seem preliminary and to represent only a limited advance.

    3. Reviewer #3 (Public Review):

      Elimination of aberrant cells from epithelial tissues is important for normal tissue physiology. Here the authors study a specific type of cell elimination that is dedicated to the removal of miss-specified cells. This type of elimination is dependent on interface contractility. The authors now identified an important role for JNK signaling, which is activated at this interface, where contractility is highest.

      Strength: The authors use a large variety of cell specification mutants and different drivers to manipulate cell specification. Together, this shows that the observed phenotypes are of a general nature and not dependent on single signaling pathways.<br /> Weakness: Quantitative characterization of much of the data is missing. Only single representative images are shown for many of the experiments. The manuscript would strengthen massively when these images are supported with a quantitative measurement. For example (but not limited to), TRE-GFP in correctly vs mis-specified clones in Figure 2K-L, TRE-GFP intensity in Figure 3, clonal analysis in Figure 5.

      Type of elimination:<br /> The authors describe a very distinct and specific phenotype of smooth rounded clones with high contractility. It is obvious that this is, on a phenotypic scale, different from other types of cell elimination, such as live extrusion and cell-cell competition. Throughout the manuscript the authors emphasize that the underlying nature of interface contractility is different to cell competition. Because cell competition "responds to a clearly defined fitness gradient between two neighbouring cells, which ensures that always the aberrant loser cell dies, independent of spatial context." And "linking apoptosis to a fixed loser genotype". However, this only holds true for the classical types of cell competition (e.g. Minute), while many examples of cell competition have been reported where elimination of cells is not set in stone, but also highly context dependent. For example, HRasV12 expressing cells are eliminated from epithelia in mice on a normal diet, while a high fat diet prevents their elimination (Sasaki et al, Cell Reports 2018). Without the experimental support that relative differences in cell specification do not cause a difference in cellular fitness it is hard to grasp the conceptual difference. Instead, the concept reported by the authors is better described as a variety of cell competition.

      Clone size<br /> The authors claim that remove aberrant cells by interface contractility is dependent on clone size and only occurs when aberrant cells are the minority compared to the surrounding tissue. Currently, there is no data in the manuscript that supports this claim. The only analysis of tissues containing a majority of miss-specified cells (Figures 2I-2J) shows a bilateral activation of JNK, similar to a minority of miss-specified cells. To support the claim that the phenotype is size dependent further analysis of clone size in relation to apoptosis and JNK activation is essential.

      JNK and cell autonomous regulation:<br /> The authors validate that expression of TRE-GFP is dependent on JNK signaling, through over-expression of a dominant negative variant of the JNK kinase (BSKDN) in clones of miss-specified cells (ey or tkv). This experiment nicely shows that activation of JNK in surrounding WT cells is not altered. This furthermore illustrates that JNK signaling in the miss-specified cells is not needed for activation of JNK in their neighbors. However, this does not support the conclusion that JNK is activated in a cell autonomous fashion in either of these populations. The interaction of the two cell types can still cause signaling, but through inhibition of one of the kinases within the pathway, this just does not lead to downstream activation of TRE-GFP. In fact, one could argue that the expression of TRE-GFP is not cell-autonomous, because tkvCA clones that are not mis-specified (within dad4-LacZ regions) do not show induction of TRE-GFP (Fig 2L). The only way to untangle cell autonomous vs non-autonomous effects is through manipulation of upstream communication between the different cell populations. Such experiments, for example manipulation of contractility, are likely beyond the scope of this study. Therefore, I would suggest rephrasing this paragraph.

      Apoptosis:<br /> A large part of the manuscript is dedicated to the characterization of elimination of miss-specified cells through apoptosis. This process is important for maintenance of tissue integrity and a crucial part of the manuscript. Some conclusions are not fully supported by the data represented in the current form of this manuscript;<br /> The authors claim that fkh- and ey-expressing cells are not eliminated when apoptosis is blocked by expression of p35. This is based on analysis of apical vs basal clone count (Figure 1T). This analysis reflects a combination of induction efficiency and clone retention. Therefore, information on the cellular behavior within clones is lacking and only provides information on survival of cells when complete clones are eliminated. The conclusion should be supported by additional analysis on clone size and total clone area, ideally based on cell number. In addition, statistical analysis of conditions with and without expression of p35 should be included.<br /> Furthermore, the analysis of apoptosis at clonal interfaces does not support the conclusion that "many, but not all apoptotic events occur at interfaces". Overall, there is increased apoptosis within clones compared to wild-type tissue. However, the rates of apoptosis are higher (ey, Fig S5B) or similar (fkh and tkvCA, Fig 5B-C) in clonal cells compared to clonal interface cells. The authors should revise these statements or provide more compelling analysis.

    1. Joint Public Review:

      Hepatitis E virus (HEV) causes over 20 million infections per year. The open reading frame 1 (ORF1) is responsible for genome replication, however very little is known about the structure and functions of several of the components. The author use a diverse a diverse number of techniques (molecular virology, structure prediction using AlphaFold, site directed mutagenesis and biochemistry) to probe ORF1 activity. The work is thorough, well prepared, and discusses the strength and weakness of the structural information. Interestingly, AlphaFold prediction of the papain-like cysteine protease domain did not identify a classic papain-like fold. Lastly, the authors demonstrate the necessity of six conserved cysteines within the putative PCP domain.

      The presence and necessity of proteolysis for genome replication or cleavage of other host factors still remains an uncharacterized problem, which is beyond the scope of this manuscript. My only concern relates to the presence of a zinc ion in ORF1.<br /> The authors use extensive triplet alanine scanning to test for virus replication capacity and in some cases see gains above WT (Figure 3). Do these patterns match natural variation observed in comparisons of HEV sequences un any way?

      Overall, the study presents an intriguing hypothesis for HEV ORF1 function not involving protease processing as assumed by early bioinformatic analysis. The alternate hypothesis of metal ion coordination is supported by increasingly sophisticated structural modeling tools and related experiments. However, a lack of direct evidence leaves, as the authors note, alternate hypotheses such as disulfide bond coordination or protease functions that occur intramolecularly within ORF1.

      The study will likely have an impact on the field, especially if evidence builds in the future directly supporting the mechanism proposed. HEV is an impactful pathogenic virus that is relatively underappreciated. In addition to a major revision in HEV biology, the idea that many proteins initially annotated with canonical functions might instead have different mechanisms is also of high interest beyond the field of virology.

    1. Reviewer #1 (Public Review):

      Of course, many of the most important aspects of feeding happen post-ingestion. As digested food moves through the intestines specialized epithelial cells (called Enterochromaffin Cells or EECs) sense and respond to the constituent chemicals. The function of EECs initiates physiological responses to facilitate nutrient absorption, protect from toxins and encourage proper waste removal. EECs are sparse and heterogenous and release a variety of transmitters and diffusible signaling molecules that signal to peripheral neurons and the brain. Their collective activity slows or speeds gut transit and promotes feelings of satiety or malaise. The current work by Liberles and colleagues seeks to provide deeper insight into the function of EECs. They build on previous work by further categorizing these cells by their unique gene expression signatures. The work utilizes single-cell transcriptomic analyses and intersectional approaches to define and genetically manipulate subsets of EECs. A key aspect of the study is behavioral assays used to investigate how direct stimulation of EEC subtypes influences key aspects of feeding, specifically gut transit, ingestion, and food preference.

      The work has several strengths. A new mouse line (Villin-flp) is developed and used intersectionally with Cre mouse lines to manipulate different subsets of epithelial cells. The authors characterize these compound mouse strains and how the labeled cells map onto transcriptomic class. These data are reasonably comprehensive and show the exclusion of marker expression from the central nervous system, important controls. The chemogenetic activation strategy is an elegant way to probe the consequences of EEC stimulation by Gq coupled GPCR signalling. The gut transit experiments show clear effects.

      The weakness is it remains unclear whether stimulation of the DREADD receptor outside the intestinal EECs really has consequences (e.g. in the tongue), the behaviors tested are somewhat limited, the responses to CNO administration variable between animals, and the effect sizes are small.

      Overall, this is an interesting study and provides useful tools for the field.

    2. Reviewer #2 (Public Review):

      Enteroendocrine cells (EEC) line the gut and prior evidence suggest that they are primary sensors of gut contents. In turn, these cells release transmitters that regulate gut function, including gut motility, enzyme secretion, and gut permeability. More recent studies have also found synaptic connections between EEC and neural sensory fibers that connect the gut to the brain, implicating this pathway in taste learning. Thus, EEC signals can be integrated with sensory signals originating in more distal areas of the alimentary canal.

      EECs express a variety of receptors and transmitters that are hypothesized to contribute to the diversity of sensing and motor functions. In this report, Hayashi et al develop a novel transgenic mouse that permits manipulation of EEC subtypes via intersectional methods. Using this approach, they identify differential roles for EEC subtypes in controlling gut motility and taste learning.


      • The authors supplement existing single-cell RNA sequencing of the proximal intestine.<br /> • A Vil1-2a-Flp mouse was generated, which exhibits highly selective expression in the gut epithelium. This mouse line can be used to manipulate EEC subtypes when bred with other Cre driver lines and double conditional (Flp/Cre) mice.<br /> • Using the above tool, different EEC subtypes were histologically characterized along the alimentary canal. Additionally, other tissues were examined, including the brain, pancreas, and lungs to demonstrate the gut specificity of their approach. The intersectional approach yield sparse recombination in the pancreas, therefore the authors included controls in their gut motility and feeding studies to account for this.<br /> • In probing the function of distinct EECs, it was found that Cck(cholecystokinin) and Gcg (GLP-1) expressing EECs slow down gut motility, whereas Tac1 (substance P) and Pet1(serotonin) expressing cells increase motility.<br /> • Food intake studies revealed several subpopulations that decrease feeding (Pet1, Npy1r, Cck, Gcg).<br /> • A conditioned flavor preference assay suggests that some of the above EEC subtypes (Pet1, Tac1, Npy1r, Gcg) decrease feeding in part through conditioned flavor avoidance.

    3. Reviewer #3 (Public Review):

      This manuscript describes a villin-2a-Flp-based intersectional strategy for selectively targeting EEC in the intestine and uses it to examine the function of subsets. The approach for targeting select subsets of enteroendocrine cells described here will be important for neuroscientists, endocrinologists, microbiologists, and other scientists studying nutritional biology. Here single-cell sequencing is used, primarily, to confirm what was already known about EEC classes at a transcriptomic level. The intersectional approach described here has the potential to provide broad access to EECs. However, from the relatively limited characterization of targeted EEC cells, it appears that the genes that have been combined with the villin driver largely fail to selectively target transcriptomically defined cell types. Thus, at present, this manuscript fails to convincingly target transcriptome-defined enteroendocrine cell types, and conclusions on gut motility, feeding behavior, and flavor avoidance are overstated.

      Some aspects of the study are compelling including the use of villin drivers as a means to restrict recombination to the epithelium containing EECs. The single-cell data (although not unique to this study) proved a basis for a better understanding of EECs and also their developmental specification. The charcoal-based gut motility assay appears valuable (although the results are perhaps not surprising given what was already known). In addition, some of the care taken characterizing extra-EEC expression is commendable. However, the manuscript is difficult to read with important details scattered in different figures and text (e.g., the characterization of expression patterns of the various lines). Moreover, whereas some things like the genetic makeup of the lines are always specified in full (excruciating) details, the expression patterns of the various lines are often casually dealt with e.g., describing separate targeting of L and I cells despite no evidence that this is actually being done. I would hope that the authors will address these issues and devote significant attention to making the paper more accessible to its readers.

    1. Reviewer #1 (Public Review):

      To explore possible functions of SA proteins in the absence of cohesin, authors use conditional AII-dependent proteins SA1 and SA2, after whose degradation they observe the phenotypes just indicated. 3D analysis shows that SA proteins cluster at specific regions. In addition, it is shown that SA proteins not only interact with CTCF after RAD21 degradation but with other F/YXF-motif containing proteins such as CHD6, MCM3 or HRNPUL2 as determined by ChIP. Mass spectrometry of proteins co-immunoprecipitated with SA1 reveals 136 interactor proteins that include a number of chromatin remodeling factors, transcription factors and RNA binding proteins including factors involved in RNA processing and modification, ribosome biogenesis and translation. After these results, authors perform CLIP to show that SA1 protein binds RNA in the absence of cohesin. Different analysis using RNH, mainly IF and IP and the S9.6 antibody, are used to conclude that SA1 binds to R-loop regions. The authors conclude that SA proteins are loaded to chromatin via NIPBL/mMAu complex at RNA:DNA hybrid regions. Further analyses suggest that SA proteins stabilize RNA via interaction with other RNA-binding proteins, some of which have been shown by other authors to be enriched at R loop-containing regions, a property that localizes to exon 32 in SA2. The manuscript provides a large amount of work that has been put together in a large collaboration to bring new roles for SA in RNA metabolism, even though this is not investigated.

    2. Reviewer #2 (Public Review):

      In the manuscript by Porter et al., the authors describe a putative role for the STAG proteins (SA1 and SA2), not as part of the cohesin complex, but in isolation and in particular at R-loops where they contribute to R-loop regulation, linking chromatin structure and cohesin loading.

      My major concern is rather general: " the role of SA1 and SA2 proteins (or cohesion subunits) its only highlighted upon acute depletion of RAD21 (cohesin subunit that holds together the complex)". I am not sure that this context is recapitulated in living cells. I.e is there a particular phase of the cell cycle where RAD21 is acutely depleted or targeted for specific degradation? How do we know that we are not looking at remnants of a complex (cohesin) that has been partially targeted by IIA mediated degradation? Is the RNA binding of SA1 an SA2 CTCF-independent (as CTCF encompasses an RNA binding domain?).

      Is there any proof that in untreated cells (ie before depletion) SA1 AND SA2 are chromatin bound independently of Rad21 (and /or SMC1-3)? Overall, it is a nice manuscript, but I am not sure whether the IAA-dependent degradation of a single subunit of pentameric complex is the right tool to assess whether other subunits of the same complex work independently.

    3. Reviewer #3 (Public Review):

      Porter, Li et al. investigate the roles of SA1 and SA2 in cohesin loading, and as well as roles that are independent of the cohesin ring. Using co-IP and imaging approaches, they show that both SA1 and SA2 interact with CTCF and they use auxin-induced degradation of Rad21 to show that this is only partially dependent on cohesin. The authors next use IP followed by mass spectrometry to identify additional SA binding partners, which include many RNA binding proteins including factors involved in RNA modification, export, splicing, and translation. Unlike the interaction with CTCF, these interactions are enhanced in cohesin depletion conditions. In fact, CLIP experiments show that SA binds RNA directly, in an R-loop-dependent manner. This co-localisation of SA with R-loops is confirmed by STORM.

      To address whether SA proteins are involved in cohesin loading, the authors measure chromatin-bound cohesin levels after auxin washoff in the presence and absence of NIPBL and SA. They find that SA knockdown has a comparable impact on cohesin binding to chromatin compared to NIPBL knockdown, and that combining the knockdowns reduces cohesin loading further. This newly synthesised cohesin co-localises with R-loop domains by STORM, and this localisation is sensitive to RNAse H. The authors propose that SA promotes cohesin loading at R-loops, and that SA1 is the main contributor to this. Finally, they provide evidence that differential usage of a conserved exon between SA1 and SA2 may be responsible for differences between SA1 and SA2 in this system, as SA2 with this exon included has higher RBP binding and is more enriched at R-loops.

      This paper provides convincing evidence that SA proteins associate with R-loops and various RNA-binding proteins, suggesting that they may have a cohesin-independent role related to RNA processing or R-loops specifically. Additionally, the paper provides evidence for a NIPBL-independent role of SA proteins at cohesin loading, which may occur at R-loops. These results will be of broad interest in relation to chromatin organisation and the role of SA proteins/cohesin in cancer.

      Overall, the experiments are thorough and well-controlled, including some nice validations such as the use of siRNA-mediated cohesin depletion and a different cell line to confirm the SA-CTCF interactions. In many cases STORM imaging is used to provide complementary evidence to western blots / IP experiments.

      However, one weakness is that imaging approaches can only address co-localisation. Although the vast majority of cohesin complexes will be bound to DNA, imaging approaches cannot distinguish between chromatin-bound and unbound nuclear proteins. For example, although cohesin co-localises with R-loops and SA after auxin washoff, and this is dependent on R-loops, it is not possible to tell from imaging whether this cohesin is chromatin bound and whether this is bound to specific genomic loci that contain R-loops or just associated with them in 3D space. Therefore it would be preferable to have a clearer distinction in terminology depending on whether the evidence discussed can demonstrate chromatin binding (e.g. chromatin fractionation experiments), or just co-localisation.

    1. Reviewer #1 (Public Review):

      This manuscript by Koropouli et al. is a much-needed study that provides novel mechanistic insight of how signaling receptors can be targeted to distinct subcellular domains or membrane locations that, in part, confer their functional specificity. It is well-established that members of the class 3 secreted semaphorins guidance cues can bind to the receptors the neuropilins (Nrp1 and Nrp2) to elicit numerous cellular processes important for circuit assembly. Previously, it was demonstrated that Sema3F signaling with Nrp2 and its co-receptor Plexin-A3 is required for the removal of excess excitatory synaptic spines on the apical dendrite of layer V cortical neurons, while the closely related member Sema3A signaling with Nrp1/Plexin-A4 promotes the elaboration of the basal dendritic arbor on the same neuron. The question is then how do the two different signaling pathways convey such precise and opposite cellular function of eliminating spines and promoting dendritic elaboration in distinct subcellular compartments of the same neuron? While some hints were provided that the Nrp2 receptor is localized to the apical dendrite and Nrp1 is distributed widely along all dendrites on the same cortical neuron in vitro, this has not been shown in vivo and the mechanism of such targeted subcellular localization is not known. In the current study, the authors used biochemical, cellular, and molecular assays in combination with mouse genetics and live-cell imaging to demonstrate that the post-translational modification of S-palmitoylation dictates the proper subcellular localization and trafficking of Nrp2, but not Nrp1, and is required for Sema3F-dependent pruning of spines on the apical dendrites of layer V cortical neurons. The following are the strength and novel findings of this study.

      1. This study confirms previous findings and adds new information by mapping the specific locations of the cysteine amino acid residues to the transmembrane/juxtamembrane region of neuropilin receptors for palmitoylation, which confers the subcellular localization specificity for Nrp2 but not Nrp1, in cortical neurons and non-neuronal cells.<br /> 2. The study also found that select cysteine residues on Nrp2 are palmitoylated by the palmitoyltransferase DHHC15, and palmitoylation of these sites are required for the homo-oligomerization of the Nrp2 receptor but not for the association with the co-receptor Plexin-A3.<br /> 3. The authors demonstrated that Sema3F signaling itself seems to enhance the level of Nrp2 palmitoylation in some sort of positive feedback loop. It would be interesting for future experiments to determine how Sema3F signaling promotes this palmitoylation.

      Although most of the key claims are supported by data presented in the paper, clarification of the following concerns would further strengthen the overall conclusion of the study.

      1. While some of the qualitative micrograph images are very convincingly in illustrating the drastic difference in Nrp2 versus Nrp1 expression patterns/cell-surface localization, such as Fig. 1A and 1D, many of the quantitative analyses have a low n number and/or low sample size, with only 2 replicate experiments or only 2 brains/animals per genotype analyzed. To increase the rigor of this study, the authors should add a few more replicates to the experiments with low n numbers.<br /> 2. The substitutions of C878, C885, and C887 to serines caused an ~80%, ~50%, and ~60% reduction, respectively, in Nrp2 palmitoylation compared to WT neuroblastoma-2a cells (as show in Fig. 2D and 2E). However, when mutating all three of these cysteine sites (the TCS plasmid), there is only ~80% total reduction in Nrp2 palmitoylation (Fig. 2F and 2G), just about equal to the C878S substitution alone. One would expect that the reduction in palmitoylation to be more severe with the TCS plasmid, but might this be due to the low n number in quantifications shown in Fig. 2E and 2G. It would add substantially to support the specificity of these cysteine residues' function if the single C878 was demonstrated to be required for either subcellular localization of Nrp2 leading to the rescue of the dendritic spine phenotype in Nrp2-/- primary neurons or in an in utero experiment.

    2. Reviewer #2 (Public Review):

      The study of Koropouli is a tour the force investigation of the Semaphorins receptor, Neuropilin-2, modification by Palmitoylation. The work consists of biochemical, cellular and in vivo experiments and overall underscore an interesting layer of regulation of axonal guidance receptors membrane localization and function by lipid modification.

    3. Reviewer #3 (Public Review):

      Although initially discovered as axon guidance molecules in the nervous system, Semaphorins, signaling through their receptors the Neuropilins and Plexins, regulate a variety of cell-cell signaling events in a variety of cell types. In addition, cells often express multiple Semas and receptors. Thus, one important question that has yet to be adequately understood about these important signaling proteins is: how does specificity of function arise from a ubiquitously expressed signaling family?

      This study addresses that important question by investigating the role of cysteine palmitoylation on the localization and function of the Neuropilin-2 (Nrp-2) receptor. It was already known that Sema3F signaling through a complex of Nrp-2 and Plexin-A3 regulates pruning of dendritic spines in cortical neurons while Sema3A signals through Nrp-1/PlexA4 to regulate dendritic arborization. The major finding of this study which is well-supported by the data is that palmitoylation of Nrp-2 regulates its cell surface clustering and dendritic spine pruning activity in cortical neurons. Interestingly, palmitoylation of Nrp-1 at homologous residue does not appear to regulate its localization or known neuronal function.

      A clear strength of this manuscript is the many techniques that are utilized to examine the question: this study represents a tour de force of biochemical, molecular, genetic, pharmacological and cell biological assays performed both in vitro and in vivo. The authors carefully dissect the function of distinct palmitoylated cysteine residues on Nrp-2 localization and function, concluding that palmitoylation of juxtamembrane cysteines predominates over C-terminal palmityolyation for the Nrp-2 dependent processes assayed in this study. The authors also demonstrate that a specific palmityl transferase (DHHC15) acts on Nrp-2 but not Nrp-1 and is required for Nrp-2 clustering and dendritic spine pruning. These findings are important because they demonstrate one mechanism by which different signaling pathways, even from a related family of proteins, can achieve signaling specificity in the cell.

      A minor weakness of the paper is that one would like to see a connection between palmitoylation-dependent cell membrane clustering of Nrp-2 on the cell surface and Nrp-2 regulation of dendritic spine pruning. Although the two phenotypes frequently correlate in the data presented, there are a few notable exceptions: e.g. Nrp-2TCS forms larger clusters in cortical neurons while Nrp-2FullCS is diffuse on the cell surface; both mutants affect spine pruning. In the future, it would also be interesting to know if increased clustering of Nrp-2 was observed at spines that were eliminated, for example. Nonetheless this manuscript represents an important advance in our understanding of synaptic pruning and cellular mechanisms that constrain protein surface localization and signaling pathways.

  2. Jan 2023
    1. Reviewer #1 (Public Review):

      The manuscript by Curtis et al. reports the interaction between CaMKII and alpha-actinin-2. The authors found that the interaction was elevated after NMDA receptor activation in dendritic spines. In addition, this study reveals NMDA receptor binding to CaMKII facilitates alpha-actinin-2 access to the CaMKII regulatory segment, indicating that the NMDA receptor is involved in this interaction. The authors identified the EF1-4 motifs mediated this interaction, and overexpression of this motif inhibited structural LTP. Moreover, biochemical measurements of affinities from various combination of protein fragments including autoinhibited CaMKII 1-315, regulatory segments of CaMKII, and the EF-hand motif reveals that autoinhibited CaMKII has limited access to alpha-actinin-2. The authors also solved the structure of the interaction, supporting their finding in neurons at the molecular level. The authors claim that the interaction between CaMKII and alpha-actinin-2 is essential for structural LTP through cooperative action by the NMDA receptor and actin cytoskeleton.

      Overall, the experiments are well-designed and the results are largely convincing and well-interpreted. But some aspects of the experiments need to be clarified.

      1. Time resolution of the interaction analysis appears to be poor, as calcium elevation in a dendritic spine would be at milli-second order. What is the time window to interact alpha-actinin-2 with CaMKII during NMDA receptor activation or LTP?<br /> 2. The authors analyzed the binding of CaMKII and alpha-actinin-2 with partial fragments. It remains to be unknown whether CaMKII can form a protein complex with GluN2B and alpha-actinin-2 in a single CaMKII protomer.<br /> 3. Besides synaptic localization, the effect of the interaction on the enzymatic activity of CaMKII is not known.<br /> 4. Although the authors quantify the effect of the EF-hand disruptor by measuring numbers of the dendritic spine by its shape, the specificity of the EF-hand disruptor needs to be clarified.

    2. Reviewer #2 (Public Review):

      Gold and his colleagues first ectopically expressed aACTN2 constructs with various deletions and determine the spatial proximity to CaMKII by PLA. Chemical LTP induced by brief glycine application in hippocampal cultures strongly augmented the PLA puncta density in spines (postsynaptic sites). This interaction specifically depended on the 4 EF hands near the C-terminus of aACTN. At the same time expression of the 4 EF hands (plus the C-terminal PDZ ligand) impaired the formation of larger mushroom spines under unstimulated conditions and the increase in mushroom spines seen after chemLTP when compared to non-transfected conditions or transection of the EF hands with a point mutation (L854R) that disrupted binding to CaMKII.

      To further define the interaction between aACTN and CaMKII the authors then solved a crystal structure formed by the aACTN EF3/4 and regulatory segment of CaMKII. This structure confirmed the role of L854 in the interaction. It also explained earlier results that phosphorylation of threonine in position 306 but not of threonine 305 of the CaMKII regulatory domain impaired aACTN binding as T306 but not T305 is engaged in critical interactions. This contrasts with Ca/CaM binding to CaMKII, which engages both threonines and is blocked by the phosphorylation of either residue. Consistently, earlier structures of Ca/CaM with the CaMKII regulatory domains showed respective differences to the new aACTN-CaMKII structure.

      Additional analysis of these data indicated that the association of the regulatory domain with the kinase domain occludes access to aACTN EF3/4. This is an important finding because it implies that only active CaMKII like T286 autophosphorylated CaMKII or bound to GluN2B would be able to effectively interact with aACTN in intact cells.

      Finally, and remarkably, binding was augmented by a protein fragment of the GluN2B C-terminus that contains the binding site for CaMKII even when Ca/CaM was still present. This result suggests that with GluN2B present aACTN can bind to CaMKII even though in the absence of GluN2B Ca/CaM occludes this binding. This finding opens up new research directions.

    3. Reviewer #3 (Public Review):

      This manuscript builds upon prior work showing that alpha-actinin-2 binds to the regulatory domain of the major postsynaptic protein kinase, CaMKII. The authors report the structure of a complex between the relevant domain in alpha-actinin-2 and a peptide based on the CaMKII regulatory domain. Data are presented indicating that the interaction of the NMDA receptor GluN2B subunit with the CaMKII catalytic domain stabilizes the complex with alpha-actinin-2. Furthermore, the authors present proximity ligation assay (PLA) data obtained in cultured neurons demonstrating that NMDA receptor activation strongly enhances the colocalization of CaMKII with alpha-actinin-2. Data obtained using mutated proteins indicate that this co-localization is mediated by the interaction characterized structurally.


      Significant strengths of this work are:<br /> 1. The high-quality structures of the complex that are reported.<br /> 2. Integration of these findings with the much better-studied complex of CaMKII and GluN2B.<br /> 3. The convincing PLA analyses show that NMDA receptor activation increases CaMKII colocalization with alpha-actinin-2.<br /> 4. The careful comparisons of data from these new studies with data reported in previous publications.


      Despite the significant strengths of the work, there are some gaps/weaknesses.<br /> 1. Although there is abundant published evidence that activated CaMKII colocalizes with NMDA receptors, the evidence for the involvement of GluN2B in the CaMKII-alpha-actinin-2 complex in neurons is lacking.<br /> 2. The evidence supporting a role for the EF1 and EF2 domains of alpha-actinin-2 in binding to CaMKII is not very convincing.<br /> 3. CaMKII autophosphorylation at multiple sites plays an important role in regulating the subcellular localization of CaMKII, but the role of autophosphorylation is not explored here.

      Taken to together the manuscript describes novel data that provide a significant extension to prior work, and the data convincingly, but perhaps only partially, support an interesting proposed model for the control of CaMKII targeting in spines.

      This more sophisticated delineation of the mechanisms underlying CaMKII targeting synapses will be of interest to the broader field of investigators studying the molecular basis for the regulation of excitatory synaptic transmission, learning, and memory.

    1. Reviewer #1 (Public Review):

      The study provides mechanistic insight into molecular events occurring at the onset of differentiation mediated by the kinase PASK. Specifically, the work focuses on the multiple steps that converge on post-translational modifications of PASK and its translocation to the nucleus during myogenesis. The authors present evidence that glutamine-fueled, CPB/EP300-mediated acetylation of PASK is required for its nuclear translocation. This allows (nuclear) PASK to interact with Wdr5 and consequently disrupt its association with the anaphase-promoting complex/cyclosome and inhibit Pax7 transcription, marking the onset of muscle differentiation. The conclusions are supported by an analysis of the effects of glutamine modulation on differentiation and maintenance of stemness in primary muscle stem cells; PASK localization in myoblasts and primary muscle stem cells as well as detailed biochemistry with modified forms of PASK to interrogate molecular interactions. C2C12 myoblast cells and primary muscle stem cells are cellular systems employed in the study with observations confirmed in cells derived from mice with genetic ablation of PASK. The study provides molecular detail on events linking glutamine metabolism to the transcriptional control of lineage differentiation, through the regulation of PASK. The analysis of these events in other systems would be of value to understanding their broader applicability.

    2. Reviewer #2 (Public Review):

      In this paper, Xiao et al. suggest that PASK is a driver for stem cell differentiation by translocating from the cytosol to the nucleus. This phenomenon is dependent on the acetylation of PASK mediated by CBP/EP300, which is driven by glutamine metabolism. Furthermore, this study showed that PASK interferes/weakens the Wdr5-APC/C interaction, where PASK interacts with Wdr5, resulting in repression of Pax7, leading to stem cell differentiation.

      There exist huge interest in maintaining adult stem cells and ES cells in their pluripotent form and the work painstakingly perform several experiments to present that PASK is a good target to achieve that goal.

      However, the work on the paper relies mostly on data from C2C12 cells as adult muscle stem cell models, in vivo experimental data, and primary myoblasts from mice. Using these models makes the story contextual in muscle stem cells. Authors have not tried to extrapolate similar claims in other adult stem cell models. This severely restricts the claim to muscle stem cells even though PASK is required for the onset of embryonic and adult stem cell differentiation in general. Their work could be much strengthened if it is also tried on mesenchymal stem cells as these cells are also as metabolically active as muscle cells.

    3. Reviewer #3 (Public Review):

      This manuscript entitled "PASK relays metabolic signals to mitotic Wdr5-APC/C complex to drive exit from self-renewal" by Xiao et al presents an interesting story on the role of PASK in the control of muscle stem cell fate by controlling the decision between self-renewal and differentiation. While the biochemistry presented is fairly compelling, the experiments revolving around the myogenic cells are lacking in quality and data.

      Major concerns:

      1. The isolation method used by this group to isolate muscle stem cells is inappropriate for the experiments used and may contribute to the misinterpretation of some of the results. It is simply a preplating method that results in a very heterogenous cell population in terms of cell type, comprised of numerous fibroblasts. While preplating can be used to isolate muscle stem cells and culture them as myoblasts, it takes days of growth and multiple rounds of passaging that are not used in this paper in order to get a more pure population of myogenic cells. This would also explain the high number of Pax7 negative cells in their primary myoblast experiments (~50% in some conditions) as they are most likely fibroblasts, which the authors could show by staining for fibroblast markers. The increase in Pax7 cells in certain conditions could also simply be due to the loss of contaminating cell types due to the treatment. Every single experiment that was performed on myoblasts must be redone using a more appropriate cell isolation method (i.e. FACS) or by culturing these isolated cells for a much longer period of time to eventually get a more pure cell population. As it stands, none of the data from the primary myoblast experiments are trustworthy.<br /> 2. The authors possess a genetic mouse model where PASK is knocked out. However, the mouse model is never described and the paper that is referenced also does not describe it. Please detail your mouse model.<br /> 3. The majority of experiments are performed on C2C12 cells. While C2C12s are adequate for biochemistry and proof of concepts, when it comes to biological significance primary myoblasts should be used. While the authors try to explain this use by claiming that primary myoblasts undergo precocious differentiation that can be avoided by using an appropriate growth media (F10, 20% FBS, 1% P/S, 5ng/mL of bFGF).<br /> 4. The authors possess a genetic mouse model, yet performed RNA-Seq on C2C12 myoblasts that were either untreated or treated with a PASK inhibitor. It would be much more informative and valuable to sequence the primary myoblasts from WT and PASK KO mice, thereby providing a more biologically relevant model.<br /> 5. The KO mouse model is rarely used and the cells isolated from it would be very useful in determining the biological role of PASK in muscle cells. The authors should isolate WT and KO cells and perform basic muscle functional experiments such as EDU incorporation for proliferation, and fusion index for differentiation to see whether the loss of PASK has an effect on these cells.<br /> 6. The authors never look at quiescent muscle stem cells and early activated muscle stem cells in terms of PASK protein expression and dynamics. The authors should isolate EDL myofibers and stain for PASK and PAX7 at 0, 24, 48, and 72-hour post isolation. This would allow the authors to quantify the changes in PASK expression and cell localization, as well as confirm the number of muscle stem cells in WT and KO mice, during quiescence and during the process of muscle stem cell activation, proliferation, and differentiation in a near in vivo context.<br /> 7. Contrary to their claim, MyoD is not a stemness/self-renewal gene.<br /> 8. The authors state that PASK is necessary for exit from self-renewal and establishment of a progenitor population but this is a vast overstatement. In the genetic KO mouse model, the mice are able to regenerate their muscle after injury, therefore PASK cannot be a necessary protein for the formation of progenitor cells.<br /> 9. In numerous figure panels, the y-axis represents the # of cells, rather than a percentage or ratio. This is uninformative as the number of cells will never be the same between conditions and experiments. These panels need to be replaced with a more appropriate y-axis.

    1. Reviewer #1 (Public Review):

      The authors of this study used SMART-seq to study differentiating B cells. Then they performed extensive in silico analyses to validate that a subset of the cells mimicked human antibody-secreting cells. For example, they compared gene expression profile of each cluster in B cell developmental trajectory (Figs 1, 2), investigated gene enrichment in ASC-like cluster (Fig 3), adopted independent dataset (Fig 3), and compared gene expression signatures of their cells to those of GC ASCs (Fig 4). Overall, the results from these analyses are convincing and valuable, but still do not seem to be a big leap from their Unger 2021 paper and therefore making this study preliminary.

      The methodology that they established should be described more clearly so that it can be shared with the research community. For example, they say cells how many donors were recruited for this experiment? are there differences in efficiency in B cell differentiation by individual?

      Also, it would be important to assay for antibodies in the culture media. How would you suggest to improve the culture system to be used to model diseases?

      At the beginning the largest contributing factor for cell culstering was cell cycle. But B cell differentiation may also influence to cell cycle regulation. Rather than normalize its effect, can authors analyze effect of cell cycle in B cell differentiation? For example, identify sub-clusters shown in supple Fig 1g.

    2. Reviewer #2 (Public Review):

      In this work, Verstegen and colleagues try to delineate human B cell differentiation trajectories by using in vitro differentiation culture of human naive B cells. The authors adopted a protocol of B cell stimulation with CD40L-expressing fibroblasts and IL-4/IL-21, and cultured B cells were analyzed by single-cell transcriptome analysis. Five distinct clusters were identified with features of memory B cells, germinal center-like B cells, ASCs, pre-ASCs, or post-GC B cells. This work provides a precise description of gene expression profiles of activated B cell populations and some insight into the pathways of effector B cell differentiation. This work will be a solid basis for human B cell study using in vitro culture of target B cell populations, providing an excellent experimental protocol.

    1. Reviewer #1 (Public Review):

      Doostani et al. present work in which they use fMRI to explore the role of normalization in V1, LO, PFs, EBA, and PPA. The goal of the manuscript is to provide experimental evidence of divisive normalization of neural responses in the human brain. The manuscript is well written and clear in its intentions; however, it is not comprehensive and limited in its interpretation. The manuscript is limited to two simple figures that support its concussions. There is no report of behavior, so there is no way to know whether participants followed instructions. This is important as the study focuses on object-based attention and the analysis depends on the task manipulation. The manuscript does not show any clear progression towards the conclusions and this makes it difficult to assess its scientific quality and the claims that it makes.

      Strengths:<br /> The intentions of the paper are clear and the design of the experiment itself is simple to follow. The paper presents some evidence for normalization in V1, LO, PFs, EBA, and PPA. The presented study has laid the foundation for a piece of work that could have importance for the field once it is fleshed out.

      Weakness:<br /> The paper claims that it provides compelling evidence for normalization in the human brain. Very broadly, the presented data support this conclusion; for the most part, the normalization model is better than the weighted sum model and a weighted average model. However, the paper is limited in how it works its way up to this conclusion. There is no interpretation of how the data should look based on expectations, just how it does look, and how/why the normalization model is most similar to the data. The paper shows a bias in focusing on visualization of the 'best' data/areas that support the conclusions whereas the data that are not as clear are minimized, yet the conclusions seem to lump all the areas in together and any nuanced differences are not recognized. It is surprising that the manuscript does not present illustrative examples of BOLD series from voxel responses across conditions given that it is stated that that it is modeling responses to single voxels; these responses need to be provided for the readers to get some sense of data quality. There are also issues regarding the statistics; the statistics in the paper are not explicitly stated, and from what information is provided (multiple t-tests?), they seem to be incorrect. Last, but not least, there is no report of behavior, so it is not possible to assess the success of the attentional manipulation.

    2. Reviewer #2 (Public Review):

      My main concern is in regards to the interpretation of these results has to do with the sparseness of data available to fit with the models. The authors pit two linear models against a nonlinear (normalization) model. The predictions for weighted average and summed models are both linear models doomed to poorly match the fMRI data, particularly in contrast to the nonlinear model. So, while I appreciate the verification that responses to multiple stimuli don't add up or average each other, the model comparisons seem less interesting in this light. This is particularly salient of an issue because the model testing endeavor seems rather unconstrained. A 'true' test of the model would likely need a whole range of contrasts tested for one (or both) of the stimuli, Otherwise, as it stands we simply have a parameter (sigma) that instantly gives more wiggle room than the other models. It would be fairer to pit this normalization model against other nonlinear models. Indeed, this has been already been done in previous work by Kendrick Kay, Jon Winawer and Serge Dumoulin's groups. So far, may concern above has only been in regards to the "unattended" data. But the same issue of course extends to the attended conditions. I think the authors need to either acknowledge the limits of this approach to testing the model or introduce some other frameworks.

    3. Reviewer #3 (Public Review):

      In this paper, the authors model brain responses for visual objects and the effect of attention on these brain responses. The authors compare three models that have been studied in the literature to account for the effect of attention on brain responses to multiple stimuli: a normalization model, a weighted average model, and a weighted sum model.

      The authors presented human volunteers with images of houses and bodies, presented in isolation or together, and measured fMRI brain activity. The authors fit the fMRI data to the predictions of these three models, and argue that the normalization model best accounts for the data.

      The strengths of this study include a relatively large number of participants (N=19), and data collected in a variety of different visual brain regions. The blocked design paradigm and the large number of fMRI runs enhance the quality of the dataset.

      Regarding the interpretation of the findings, there are a few points that should be considered: 1) The different models that are being studied have different numbers of free parameters. The normalization model has the highest number of free parameters, and it turns out to fit the data the best. Thus, the main finding could be due to the larger number of parameters in the model. The more parameters a model has, the higher "capacity" it has to potentially fit a dataset. 2) In the abstract, the authors claim that the normalization model best fits the data. However, on closer inspection, this does not appear to be the case systematically in all conditions, but rather more so in the attended conditions. In some of the other conditions, the weighted average model also appears to provide a reasonable fit, suggesting that the normalization model may be particularly relevant to modeling the effects of attention. 3) In the primary results, the data are collapsed across five different conditions (isolated/attended for preferred and null stimuli), making it difficult to determine how each model fares in each condition. It would be helpful to provide data separately for the different conditions.

    1. Reviewer #1 (Public Review):

      The article by Mann et al. describes a knockin (KI) mouse model of mitofusin 2- related lipodystrophy, in mice carrying MFN2 R707W. The mice recapitulate some but not all aspects of the human phenotype, as summarized in Table 2. The phenotypic characterization is extensive and is generally well done. There was an adipose-specific alteration of mitochondrial morphology, accompanied by activation of the integrated stress response and reduced adipokine secretion. These findings are consistent with the human phenotype. The alteration in fat distribution that is present in humans with this mutation was not observed, and the mice did not have the insulin resistance seen in humans. The transcriptome analyses revealed a reduced epithelial-mesenchymal transition (EMT) in the KI mice, suggesting possible involvement of TGF-beta related pathways. There was also upregulation of the mTorc signaling pathway, suggesting that a possible therapeutic approach in humans may involve the mTORC1 inhibitor sirolimus. The reason for the largely adipose -specific effect of the mutation remains unexplained. As well, the hypothesis that changes in EMT pathways reflect altered activity of TGF-beta pathways must remain somewhat speculative at this point. Notwithstanding these weaknesses, the manuscript provides an important advance in understanding this lipodystrophy (and potentially other lipodystrophies), and the model that has been generated will enable further studies to further characterize the pathophysiology.

    2. Reviewer #2 (Public Review):

      This study generated a valuable preclinical model of patients with Mfn2-related lipodistrophy (R707W). Such a mouse model enables the understanding the pathogenic mechanism causing this lipodistrophy and testing specific therapeutic approaches for these patients.

      The strengths are the thorough phenotypic characterization of the mice and the clear decrease in circulating leptin and adiponectin levels in the absence of changes in fat mass observed in Mfn2 R707W/R707W mice. This partially recapitulates one of the key phenotypes of human patients with these mutations.

      The major weakness is the conclusion that the integrated stress response is activated in white adipose tissue is not supported by the data and the phenotype. The ISR caused by primary insults to mitochondria was defined as a response that decreases the translation of mitochondrial proteins, thus decreasing mitochondrial respiratory function via ATF4 without engaging ATF5 (Quiros et al., JCB 2016). In addition, the increase in ATF4 caused by phosphorylation of eif2alpha is in ATF4 translation and translocation to the nucleus, not in ATF4 transcription. It is a possibility that it is a selective increase in ER stress that is responsible for defective leptin secretion, as Mfn2 R707W/R707W adipose tissue shows no mitigation of mitochondrial function as expected from ATF4-ISR activation.

    3. Reviewer #3 (Public Review):

      Mann and colleagues have generated a knock-in mouse model carrying a recently identified mutation in the Mfn2 gene that leads to a syndrome of severe upper body adipose overgrowth in humans (Mfn2R707W). The goal was to gain a better mechanistic understanding on how this mutation leads to such a dramatic phenotype in humans. The authors consistently demonstrate how the knock-in mutation leads to abnormalities in mitochondrial shape, mtDNA content, as well as in the abundance of some mitochondrial proteins, most notably in brown adipose tissue. The authors detect some stress response signatures, which could explain the decreased leptin and adiponectin levels observed in the knockin mice.

      The authors have to be praised for their effort in trying to provide mechanistic insights to such a rare condition. This work constitutes a real tour de force in the characterization of Mfn2R707W mice. The path, however, was full of surprises. On one side, the knockin mouse model fails to recapitulate multiple aspects of the human syndrome. This is, of course, beyond the control of the researchers, but somehow tells us that there are some elements missing in our understanding of the effects of this Mfn2 mutation at the cellular level (not just organismal), and on why it impacts so much adipose tissues. A second layer of complexity is that the authors find an interesting connection between Mfn2R707W, the integrated stress response and a severe decrease in the expression of leptin and adiponectin. However, whether these elements have any causal role in the human syndrome or in the phenotypes observed in the mice, remains an open question.

    1. Reviewer #1 (Public Review):

      In this study, Barthe et al. developed an approach to selectively activate beta-adrenergic receptors in the sarcolemma of ventricular myocytes. The approach involved the linking of a 5Kd PEG chain to the beat agonist isoprenaline. This prevents the agonist from entering transverse tubules. Using this approach, the authors find that activation of beta-adrenergic receptors in the surface sarcolemma of ventricular myocytes leads to lower cytosolic cAMP levels but longer-lasting effects on EC coupling than when TT receptors were activated.

      Strengths of the study:<br /> 1) The PEG-ISO, size exclusion approach is very interesting and useful.<br /> 2) The observation that activation of beta-adrenergic receptors in the surface sarcolemma of ventricular myocytes leads to lower cytosolic cAMP levels, but longer-lasting effects on EC coupling than when TT receptors were activated is interesting.<br /> 3) The observation that beta-adrenergic receptors in the TT lead to stronger nuclear activation of nuclear cAMP/PKA signaling is interesting.

      Weaknesses of the study:<br /> 1) There seems to be a paucity of mechanistic insights into the study.<br /> 2) It is unclear what would be the ideal control for these experiments. Would the addition of the PEG chain, by itself, alter the binding of and activation of beta-adrenergic receptors regardless of their location?<br /> 3) The novelty of the findings is unclear, as other studies have suggested differential effects of beta-adrenergic receptors in membrane compartments.

      Impact on the field:<br /> 1) PEG-ISO may become a useful strategy to selectively activate surface sarcolemmal beta-adrenergic receptors.

    2. Reviewer #2 (Public Review):

      Barthé et al. present a manuscript examining membrane-domain specific signaling by βAR stimulation in cardiomyocytes. Specifically, the authors seek to use a size exclusion approach using PEGylated-isoproterenol to allow only surface sarcolemmal βAR receptor stimulation without T-tubule βAR stimulation. This innovative approach was advanced using confocal microscopy to determine the accessibility of the PEGylated substrates to the T-tubule network. The authors show comparable responses of L-type Ca channels, Ca transients, and contraction using equipotent doses of PEG-Iso and Iso, but differences in nuclear and cytoplasmic cAMP responses based on FRET reporters.

      Strengths<br /> 1. The size exclusion strategy using PEGylation technology is well rationalized and well supported by the physicochemical characterization of PEGylated Iso. This represents a novel strategy to decipher cardiomyocyte cell surface signaling from T-tubule network signaling resulting from the stimulation of β-adrenergic receptors. This approach can be used to study the compartmentalization of various signaling pathways in cardiomyocytes as well as in other cell types that exhibit complex cytoarchitecture. The authors use multiple cAMP FRET sensors as well as assay a number of relevant physiological cellular responses to assess the effect of Iso vs. PEGylated Iso which are informative.

      Weaknesses<br /> 1. The authors' evidence that PEG-FITC does not penetrate the TT network is not convincing as presented in Figure 1. A single confocal image from one cell showing a lack of fluorescence (Figure 1A) could be due to an outlier cell or lack of penetration to more central regions of the cell where images are taken from. More convincing would be a confocal Z-scan series comparing PEG-FITC and FITC in ARVM. Some form of quantification of T-tubule network density from multiple cells would provide even more robust evidence, similar to the many studies that have done this characterization in models of dilated cardiomyopathy showing a loss of TT network. This exclusion of PEG-FITC provides the critical foundation for the paper and it is somewhat unanticipated given the large dimensions of the t-tubules relative PEG-Iso, so strong data here are particularly important.

      2. The conclusion on line 160 that 'the maximal efficacy of PEG-Iso was significantly lower by 30% than that of Iso,' may be overstated. What approach was used to conclude significantly differently as this implies a statistical comparison? Were the concentration-response curves fit to determine maximal responses? In the examples given, the responses are continuing to increase at the highest concentrations tested, so it is difficult to simply compare the responses to the highest doses tested.

      3. For experiments using adenovirus delivery of FRET-based sensor, the culture of ARVM is required which may impact the biology. Such culture is known to result in changes in cell structure and physiology with loss of the TT network over time. It is essential for the authors to demonstrate that under the conditions of their FRET experiments, the cells continue to exhibit a robust TT network.

      4. As pointed out by the authors, the interpretation of OSM/TTM adrenergic receptor functions in this study is limited by the fact that the relative contributions of β-adrenergic receptor subtypes had not been assessed. This particularly complicates the interpretation of their results in that the authors demonstrate in Figure 2 that PEGylation increases the Ki for Iso for β1 receptors by 700-fold whereas the increase for β2 receptors is about 200-fold. Thus, the relative contribution of β1 and β2 receptors to a 'comparable' dose of Iso and PEGylated Iso will potentially be different. Could that difference in relative β1/β2 receptors be the cause of the different 'efficacy of nuclear and cytoplasmic' cAMP changes between the two tested ligands in Figure 8 and supplemental Figure 3? This would fundamentally alter the conclusions of the paper.

      5. The equipotent doses of Iso and PEG-Iso were initially defined based on their ability to elevate global [cAMP]i. The authors then further demonstrated that such equipotent doses of Iso and PEG-Iso also had equal effects on ICa,L amplitude, Ca2+ transient parameters, and cellular contractility (shortening), presumably because they raised global [cAMP]i to the same levels. These findings seem to defy the importance of nanodomain organization and local [cAMP]i in the regulation of LTCCs, Ca2+ cycling proteins, and contractile machinery. The authors argued that "Since OSM contributes to ~60% of total cell membrane in ARVMs, either β-ARs and ACs are more concentrated in OSM than TTM, or they are in large excess over what is needed to activate PKA phosphorylation of proteins involved in EC coupling. Also, cAMP produced at OSM must diffuse rapidly in the cytosol in order to activate PKA phosphorylation of substrates located deep inside the cell, such as LTCCs in TTM" (lines 336-341). Although this argument may be valid at high concentrations of Iso and PEG-Iso when PKA activation is saturated, it also implies that discrepancy could be detectable at lower (non-saturating) doses of Iso and PEG-Iso. Thus, additional experiments using lower Iso and PEG-Iso doses are required to support this notion.

      6. The size excluded compartment for PEG-Iso proposed by the authors is the TT network, but this ignores other forms of sarcolemmal nanodomains such as caveolae, which include β2 receptors and AC, and may exhibit similar if not great sensitivity to the size exclusion approaches pioneered by the authors.

    3. Reviewer #3 (Public Review):

      The manuscript by Barthe et al compares the effects derived from the application of isoprenaline (Iso) or isoprenaline covalently linked to PEG (PEG-Iso) on adult rat ventricular myocytes (ARVM). Iso is a well-characterized β-AR agonist and the authors work under the assumption that PEGylation of Iso prevents it from accessing the T-tubules. Therefore, due to its larger size, PEG-Iso is only able to activate β-ARs located on the outer surface membrane (OSM), and any additional effect observed by Iso stimulation is attributed to the activation of β-ARs located in T-tubules. First, the authors determined that the affinity of PEG-Iso for β-ARs is about 100 times lower than the one of Iso. Then, they analyze the effects of Iso (10 nM) and PEG-Iso (1 µM) on calcium channel currents, contractility, calcium transients, and cytosolic and nuclear PKA activity. They only found a stronger effect of Iso on nuclear pKA activity. Therefore they conclude that, while OSM β-ARs stimulation mainly results in positive inotropy and lusitropy, T-tubules ARs stimulation mainly results in increased nuclear pKA activity.

      Overall the manuscript is well written and the findings are biologically important from the perspective of understanding the mechanism of β-AR stimulation as well as in assigning the functional contribution of β-ARs in the OSM and in the T-tubules. However, the major conclusion is not strongly supported by the data. The interpretation of the results is all based on the assumption that PEG-Iso is excluded by the T-tubules, but no experiment presented here rigorously demonstrates this.

      1. The only indication that PEG-Iso may be excluded by the T-tubules is one confocal image in which FITC or PEG-FITC were applied on ARVM. No experiment has been performed to assess if PEG-Iso is indeed not able to enter the T-tubules.<br /> The treatment of ARVM with neuraminidase made the T-tubules accessible to PEG-FITC. If the authors could demonstrate that neuraminidase treatment followed by PEG-Iso would result in similar nuclear pKA activity as Iso, this would strengthen their conclusion.<br /> 2. The fact that PEG-Iso treatment resulted in a lower increase of intracellular cAMP (Figure 3) could also be due to the activation of a smaller fraction of β-ARs, independent of their localization.

    1. Reviewer #1 (Public Review):

      In this manuscript, Braet et al provide a rigorous analysis of SARS-CoV-2 spike protein dynamics using hydrogen/deuterium exchange mass spectrometry. Their findings reveal an interesting increase in the dynamics of the N-terminal domain that progressed with the emergence of new variants. In addition, the authors also observe an increase in the stabilization of the spike trimeric core, which they identify originates from the early D614G mutation.

      Overall this is a timely and interesting exploration of spike protein dynamics, which have so far remained largely unexplored in the literature.<br /> What I find a bit missing in this manuscript is a link between how the identified changes in protein dynamics lead to increased viral fitness. While there are some possibilities listed in the discussion, I think these should be elaborated upon further. In addition, it should also be discussed how understanding the changes in the spike protein dynamics could have implications for the development of small molecule inhibitors for the virus.

    2. Reviewer #2 (Public Review):

      The study systematically looks at dynamic differences across variants longitudinally and the authors appropriately only limit their analyses to peptides that are conserved across the different variants.

      There are some concerns listed below, particularly related to the ensemble heterogeneity that is reported and need considerable revision.

      1) The authors explain that cold-temperature treatment of the S trimer ectodomain constructs has been shown to lead to instability and heterogeneity. They also show this with a comparison of untreated vs. 3-hour 37 C treated samples. I'm confused as to why "During automated HDXMS experiments protein samples were stored at 0 degrees". Will this not cause issues in protein heterogeneity, where the longer the protein sits at 0 C the more potential heterogeneity there will be, and thus greatly confound the analysis?

      2) The authors presume that the bimodal spectra that are observed reflect EX1 kinetics, however, there can be multiple reasons for an apparent bimodal distribution in the spectra. I agree that some of the spectra indicate that more than a single species is present, but what the two populations represent is murky. In Figure 2D, the apparent size of the highly deuterated population gets larger going from the 60 sec to the 600-sec spectra, as expected for an EX1 transition. However, in Figure 3D the WT highly deuterated population gets smaller going from the 60-sec to the 600-sec spectra. Were bimodal examples observed beyond those shown in Figure 2?

      3) How were the spectra that appeared broadened analyzed? There is no description of this in the methods, and the only data shown for this is in table 1. The left/right percentages are reported without any description of how they were obtained. Are these solely from a single spectrum? The most alarming issue is that Table 1B reports 9.4% for the right population of the 988-998 peptide, but the corresponding spectra in Figure 3D doesn't seem to have any highly deuterated population at all.

      4) The authors state on page 12: "Replicate analysis of stabilized S trimers with incubation at 4C prior to deuterium exchange (see methods) showed a time-dependent reversal of stabilization as reported previously (Costello et al., 2022), most evident at the same peptides." Is this data shown anywhere? If not then it should be included somewhere, possibly in table 1 as I would expect the cold treatment to offset the left/right population sizes.

      5) The authors state that peptide 899-913 'exhibits a slow conformational interconversion (time scale ~ 15-30 min)'. Where did this estimated rate come from? From the data shown and the limited number of time points, I don't think there is sufficient sampling of this conformational transition to really narrow down the exact timescale, especially since the ratio of left/right populations is so dependent on the pre-treatment of the sample prior to deuterium exchange. (See 1st comment)

      6) The woods plots presented in the Supporting information: (Figures 2-S4, 2-S5, 3-S4, 4-S2, 5-S2, 6-S2) are not conventional Woods plots. Normally the plots would indicate a global threshold for what is deemed to be significant based on the overall error in the dataset. From what I gather the authors used error within an individual peptide to establish significance for each specific peptide, which would be okay, but the authors don't describe the number of replicates or how the p-value was calculated. I would strongly recommend that the authors instead rely on a hybrid significance testing approach, as described recently: (PMID 31099554). What's really alarming with the current approach is that several of the Woods plots shown have data points found to be significantly different that are right at zero on the y-axis.

      7) Table 1: The summary of the peptides with observed bimodal behavior should include data from the replicates, particularly for assessment of how consistent the left/right population sizes are across replicates. Instead of just a percentage, the table should report an average and the standard deviation from the replicate measurements. Furthermore, the table should also include peptides that are overlapping with those presented. Based on Figure 2-figure supplement 1, there are at least two other peptides that cover the 899-913 region. These additional peptides should show a similar trend with bimodal profiles and will be important for showing how reproducible the apparent EX1 kinetics are in the dataset.<br /> All available replicates and overlapping peptides should be analyzed to ensure that these percentages reported are consistent across the data. It is also odd that the authors choose to use the 3+ charge state of the WT, but the 2+ for the D614G mutant. If both charge states were present, then both of them should be analyzed to ensure the population distributions are consistent within different charge states.

      8) The method for calculating p-values used to assess the significance of a difference in observed deuterium uptake is not described. The manuscript mentions technical replicates, but no specific information as to how many replicates were collected for each time point. These details should be included as they are also part of the summary table that is recommended for the publication of HDX data.

    3. Reviewer #3 (Public Review):

      The authors use hydrogen-deuterium exchange mass spectrometry (HDXMS) to assess the dynamics of several relevant mutant forms of SARS-CoV 2 Spike protein including the most recent Omicron variant. The Spike protein is heavily glycosylated and is a trimer so is a very difficult protein to study by HDXMS. The authors confirm the glycosylation sites, which can't be covered by the HDXMS experiment, yet they still manage to cover nearly 50% of the sequence revealing many interesting changes in dynamics in the prevalent circulating mutant forms. The beautiful HDXMS data reveal consistent trends as SARS-CoV2 mutates to survive including stabilization of the stalk and increased dynamics of the N-terminal domain where ACE2 receptor binding occurs. The authors incubate the protein at 37C and discover additional stabilization of the trimer occurs under these conditions explaining a lot of conflicting data in the literature done at different temperatures. These results have profound implications for the development of small molecule inhibitors of the Spike protein-ACE2 interaction.

    1. Reviewer #1 (Public Review):

      This is an exceptional paper that investigates a 208.6 kb region of the Burkholderia thailandensis chromosome that had previously been thought to excise itself and form extrachromosomal circles. Through a series of elegant experiments , the authors conclusively show that (i) the 208.6 kb region in fact forms tandem duplications, (ii) the region can switch between duplicated and non-duplicated forms via RecA-mediated homologous recombination, and (iii) duplication provides a selective advantage in biofilms. The data are of uniformly high quality and the conclusions are fully supported by the data. The significance of the work is high because it identifies a novel form of phase variation in bacteria that represents a bet-hedging strategy to facilitate growth in diverse environments.

    2. Reviewer #2 (Public Review):

      This beautiful study identifies a genetic mechanism controlling colony morphology differences in Burkholderia thailandensis. There is a large region of the genome which can be duplicated or triplicated in a RecA-dependent recombination process, leading to phenotypic changes. In addition to colony morphology differences in cells with one, two, or three copies of the region, other phenotypes like biofilm formation are impacted. This appears to be an unstable genetic change since some of the colony types can interconvert to others after restreaking. The authors are commended for the development of elegant genetic approaches to study and carefully prove the existence of the copy number variation of this genomic region. These approaches will be of great use to the field in studying copy number variation in bacteria far beyond Burkholderia or colony morphology/biofilm formation. Bacteriology has for decades focused on average measurements of a culture, and this study helps usher the field to a new future where we appreciate and measure the behaviors of individual populations of cells within the same culture.

    3. Reviewer #3 (Public Review):

      This paper shows that RecA-mediated recombination between two insertion sequence elements can drive the duplication of a large (~200 kb) region that leads to a growth advantage in biofilms, but a disadvantage during planktonic growth. The experiments presented are incisive and definitive. While IS elements are more commonly implicated in gene inactivation, this paper reveals that they can provide a benefit by driving a reversible genome modification in the form of a large-scale duplication. The paper should appeal to readers interested in mechanisms of genome evolution, phase variation, biofilms, and bacterial pathogenesis. The final model is convincing and also lays the foundation for future studies aimed at identifying which gene(s) in the duplicated region are ultimately responsible for the biofilm growth benefit. The paper also serves to correct this lab's prior interpretation of related data in which they concluded that the genomic region being investigated excised and circularized. They very nicely lay out what led them to conclude this previously and how their new data led to a revised model, as well as many additional, important new insights. To be clear, there were no issues with the prior data, just the interpretation/model. So in my view, this is exactly how science should unfold - new data can and should lead to revised models. I applaud the authors for laying this trajectory out in such a straightforward, open manner.

    1. Reviewer #1 (Public Review):

      This work aims to understand whether MSCs support the resistance in tumor cells upon CAR T cell treatment and whether the expression of STC1 in MSCs contributes to those changes. Overall, the in vivo data is interesting. However, the mechanistic understandings are correlated and based on many assumptions. Furthermore, the differences in Treg changes presented in Figure 2 are not convincing. It is also not clear the underlying mechanisms by which the presence of MSCs leads to these changes.

      Major points:

      1. How STC1 controls changes in MSCs' ability for hampering CAR T cell-mediated anti-tumor responses is unclear.

      2. Is ROS important? It is not tested directly.

      3. The changes in CD8 and Treg are not convincing. Moreover, it is not tested how these changes can be elicited by the presence of MSCs.

    2. Reviewer #2 (Public Review):

      Zhang et al. addressed an intriguing question - whether the presence of mesenchymal stem cells (MSCs) could influence the efficacy of CAR-T therapy. After observing that CAR-T cytotoxicity was strongly inhibited by MSCs by modulating certain correlated immune response pathways, the authors sought to uncover the underlying mechanisms by examining the interaction between MSCs and macrophage, immune escaping mechanisms, and oxidative stress. Notably, the authors discovered that a single gene, STC1, played a major role in reversing the suppression when it was knocked down/out. Although more research is necessary to clarify the signaling pathways, the data presented by the authors were generally well-supported and convincing.

      Major points:

      1. STC-1 is expressed and secreted by many human cancer cells. This should be discussed in the introduction or discussion with more inter-related background info on both its regulation in cancer cells and secretion pattern into TME. It is important because you state that the STC-1 secreted by MSC has such strong functions, then how about those produced and secreted by cancer cells? Are those also stimulated by macrophages or other components in TME? Do they have possible functions in helping cancer cell to escape the immune surveillance mechanisms?

      2. In Figure 4B, using a single marker of IL-1β to show the immune suppressive capability of MSC in vivo is not sufficient, staining for CD4+ and CD8+ should also be included to demonstrate whether MSC could modulate T cell compositions, which can give more direct evidence about MSC's impacts on CAR-T cell.

      3. One of the major risks associated with CAR-T therapy is an excessive immune response that causes cytokine release syndrome. MSCs have been used in clinics as a way to suppress immune response including post-CAR-T. What does the author think about using MSC with STC-1 knockout? Can it still help reduce toxicity while maintaining CAR-T efficacy? This might be a potential application.

      4. There was a recent study published in Cancer Cell (Lin et al. Stanniocalcin 1 is a phagocytosis checkpoint driving tumor immune resistance. 2021), and they also reported that STC1 negatively correlates with immunotherapy efficacy and patient survival. It should be cited, and in fact, it provided support to the authors' present study with completely different experimental settings.

    1. Reviewer #1 (Public Review):

      This theoretical (computational modelling) study explores a mechanism that may underlie beta (13-30Hz) oscillations in the primate motor cortex. The authors conjecture that traveling beta oscillation bursts emerge following dephasing of intracortical dynamics by extracortical inputs. This is a well written and illustrated manuscript that addressed issues that are both of fundamental and translational importance. Unfortunately, existing work in the field is not well considered and related to the present work. The rationale of the model network follows closely the description in Sherman et al (2016). The relation (difference/advance) to this published and available model needs to be explicitly made clear. Does the Sherman model lack emerging physiological features that the new proposed model exhibits? The authors may also note the stability analysis in: Yaqian Chenet et al., "Emergence of Beta Oscillations of a Resonance Model for Parkinson's Disease", Neural Plasticity, vol. 2020, https://doi.org/10.1155/2020/8824760

      The model-based analysis of the traveling nature of the beta frequency bursts appears to be the most original component of the manuscript. Unfortunately, this is also the least worked out component. The phase velocity analysis is limited by the small number (10 x 10) of modeled (and experimentally recorded) sites and this needs to be acknowledged. How much of the phase velocities are due to unsynchronized random fluctuations? At least an analysis of shuffled LFPs needs to be performed. How were border effects treated in the model and which are they? Is there a relationship between the localizations of the non-global external input and the starting sites of the traveling waves?

      In summary, this work could benefit from a widening of its scope to eventually inspire new experimental research questions. While the model is constructed well, there is insufficient evidence to conclude that the presented model advances over another published model (e.g. Sherman et al., 2016).

    2. Reviewer #2 (Public Review):

      Kang et. al., model the cortical dynamics, specifically distributions of beta burst durations and proportion of different kind of spatial waves using a firing rate model with local E-I connections and long range and distance dependent excitatory connections. The model also predicts that the observed cortical activity may be a result of non stationary external input (correlated at short time scales) and a combination of two sources of input, global and local.

      Overall, the manuscript is very clear, concise and well written. The modeling work is comprehensive and makes interesting and testable predictions about the mechanism of beta bursts and waves in the cortical activity. There are just a few minor typos and curiosities if they can be addressed by the model. Notwithstanding, the study is a valuable contribution towards developing data driven firing rate.

      1) The model beautifully reproduces the proportion of different kind of waves that can be seen in the data (Fig 3), however the manuscript does not comment on when would a planar/random wave appear for a given set of parameters (eg. fixed v_ext, tau_ext, c) from the mechanistic point of view. If these spatio-temporal activities are functional in nature, their occurrence is unlikely to be just stochastic and a strong computational model like this one would be a perfect substrate to ask this question. Is it possible to characterize what aspects of the global/local input fluctuations or interaction of input fluctuations with the network lead to a specific kind of spatio-temporal activity, even if just empirically ? Do different waves appear in the same trial simulation or does the same wave type persist over the whole trial? If former, are the transition probabilities between the different wave types uniform, i.e probability of a planar wave to transit into a synchronized wave equal to the probability of a random wave into synchronized wave?

      2) Denker et al 2018, also reports a strong relationship between the spatial wave category, beta burst amplitude, the beta burst duration and the velocity (Fig 6E - Denker et. al), eg synchronized waves are fastest with the highest beta amplitude and duration. Was this also observed in the model ?

    3. Reviewer #3 (Public Review):

      In this manuscript, the authors consider a rate model with recurrently connections excitatory-inhibitory (E-I) modules coupled by distance-dependent excitatory connections. The rate-based formulation with adaptive threshold has been previously shown to agree well with simulations of spiking neurons, and simplifies both analytical analysis and simulations of the model. The cycles of beta oscillations are driven by fluctuating external inputs, and traveling waves emerge from the dephasing by external inputs. The authors constrain the parameters of external inputs so that the model reproduces the power spectral density of LFPs, the correlation of LFPs from different channels and the velocity of propagation of traveling waves. They propose that external inputs are a combination of spatially homogeneous inputs and more localized ones. A very interesting finding is that wave propagation speed is on the order of 30 cm/s in their model which is consistent with the data but does not depend on propagation delays across E-I modules which may suggest that propagation speed is not a consequence of unmylenated axons as has been suggested by others. Overall, the analysis looks solid, and we found no inconsistency in their mathematical analysis. However, we think that the authors should discuss more thoroughly how their modeling assumptions affect their result, especially because they use a simple rate-based model for both theory and simulations, and a very simplified proxy for the LFPs.

      The authors introduce anisotropy in the connectivity to explain the findings of Rubino et al. (2006), showing that motor cortical traveling waves propagate preferentially along a specific axis. They introduce anisotropy in the connectivity by imposing that the long range excitatory connections be twice as long along a given axis, and they observe waves propagating along the orthogonal axis, where the connectivity is shorter range. Referring specifically to the direction of propagation found by Rubino et al, could the authors argue why we should expect longer range connections along the orthogonal axis? In fact, Gatter and Powell (1978, Brain) documented a preponderance of horizontal axons in layers 2/3 and 5 of motor cortex in non-human primates that were more spatially extensive along the rostro-caudal dimension as compared with the medio-lateral dimension, and Rubino et al. (2006) showed the dominant propagation direction was along the rostro-caudal axis. This is inconsistent with the modeling work presented in the current manuscript.

      The clarity and significance of the work would greatly improve if the authors discussed more thoroughly how their modeling assumptions affect their result. In particular, the prediction that external inputs are a combination of local and global ones relies on fitting the model to the correlation between LFPs at distant channels. The authors note that when the model parameter c=1, LFPs from distant channels are much more correlated than in the data, and thus have to include the presence of local inputs. We wonder whether the strong correlation between distant LFPs would be lower in a more biologically realistic model, for example a spiking model with sparse connectivity and a spiking external population, where all connections are distant dependent. While the analysis of such a model is beyond the scope of the present work, it would be helpful if the authors discussed if their prediction on the structure of external inputs would still hold in a more realistic model.

    1. Reviewer #1 (Public Review):

      Luckey et al. used a sophisticated, multimodal approach to test the hypothesis that engaging LC-hippocampal pathways promote behavioral tagging processes in humans. To activate this mechanism in a causal manner, they apply transcutaneous electrical stimulation of the greater occipital nerve (NITESGON), a relatively novel and non-invasive technique for stimulating brainstem pathways linked to arousal-related neuromodulation. To test the behavioral tagging hypothesis, they use a variety of indirect methods, including pharmacology, EEG, fMRI, saliva assays, and eye-tracking to measure LC-related activity, hippocampal activity/connectivity, and potential dopamine states/release. At the behavioral level, they demonstrate that NITESGON stimulation during or after learning benefits long-term but not immediate associative memory. These long-term memory improvements were related to increased gamma power in the MTL. In another set of experiments, they show that NITESGON during associative learning promotes associative learning on a subsequent unrelated (object-location) or highly overlapping (paired word associates) task. Consistent with prior VNS and other NITESGON studies, they show robust evidence that this intervention leads to significant increases in salivary alpha-amylase, a putative marker of central noradrenergic activity. This increase in sAA was also correlated with long-term associative memory across several experiments using paired word associates. Using fMRI, they demonstrate resting-state increases in local hippocampal, LC, and VTA low-frequency fluctuations as well as increased rs-FC between the LC and hippocampus during and after stimulation. Finally, they show that NISTESGON does not enhance long-term associative memory in individuals taking a dopamine antagonist medication, implicating a potential dopamine mechanism in these stimulation-induced memory effects.

      This paper is impressive in scope and takes advantage of both causal and indirect methods to cross-validate their results. Behavioral tagging is a relatively nascent area of research in humans, and this paper provides compelling evidence for the role of noradrenergic activity (whether related to behavioral tagging or more general arousal-related consolidation processes) in facilitating memory encoding and consolidation. Beyond basic science research, these findings also have important clinical implications. In recent years, there has been intense interest in studying the LC's role in promoting healthy cognitive function and its involvement in AD-related neuropathology. The LC is one of the earliest sites of tau pathology and thereby represents an important target for clinical intervention in early AD. The current study advances our understanding of a non-invasive technique that may be used to bolster learning in both healthy populations and potentially in older individuals with AD.

      The key claims of the manuscript are generally well supported by the data. However, while the large number of studies is a significant virtue of this paper, it is also - at times - a potential weakness. There are many measures and pieces to this puzzle to assemble. While the multimodal approach is admirable and rigorous, the fit between some of these pieces is sometimes overstated. The correlational nature of the data helps cross-validate some of the predictions about the LC mechanisms involved in behavioral tagging. But the most compelling test of this hypothesis would be to link the LC/hipp/VTA fMRI data - arguably the most direct outcome measure in this study - to long-term memory performance and the other neurophysiological measures (e.g., sAA, blink rate, etc.). Many of the results are compelling but they are often observed in parallel studies. Thus, interpreting them as engaging a common mechanism is tenuous. This important shortcoming notwithstanding, there is still a strong replication in other findings (e.g., sAA-memory correlations) across experiments that lend support to some of the hypotheses.

      A related issue is that the reliability of these indirect measures of noradrenergic signaling and dopaminergic receptors, including salivary alpha-amylase and spontaneous eyeblink rate, is oversold. While this stimulation technique elicits parallel increases in many of the neurophysiological and behavioral measures, these patterns might not reflect the engagement of a shared underlying mechanism. It's an especially big stretch to interpret the eyeblink effects as relating to LC-DA, which cannot be verified using the current methods. In addition, the spatial resolution of the neuroimaging data is poorly suited for testing predictions about such a small brain structure. This represents a potential weakness of the paper, as the large smoothing kernel in the fMRI data may capture the contributions of other brainstem nuclei and regions activated by NITESGON. It is also worth noting that many of the individual differences findings are confounded by group clustering effects. That is, the between-group effects belie whether the same linear relationships exist in the sham and stimulation groups individually. This necessitates additional correlation analyses within groups to verify that stimulation doesn't decorrelate the relationship between physiological measures and performance.

      While the behavioral tagging predictions are intriguing and supported by some findings in the literature, they may not be entirely appropriate for this study. In short, I'm not fully convinced these data satisfy all assumptions of BT (see Dunsmoor et al., 2022 for an overview). Behavioral tagging is thought to be a process that stabilizes weak learning. While it's very difficult to operationalize the "strength" of a memory representation, I'm not sure if the current paired-associates paradigm yields weak learning. Participants have multiple opportunities to learn the memoranda, which casts some doubt as to whether these are weak memory representations. This possibility is supported by the generally high memory performance (~80% on average) during the immediate test and even accurate recall after 7 days.

      Behavioral tagging also does not make any explicit predictions about interference effects. Much of this theory centers upon the idea that arousing learning events lead to memory enhancements/benefits; but it does not speak directly as to whether these events confer protection from memory interference (and there was no baseline condition in Dunsmoor et al., 2015 to test any predictions regarding reduced retroactive interference for CS+ stimuli, for example). I find the protective effects of stimulation in Experiment 4 very interesting, and they speak to the importance of this technique as a memory intervention. However, I think this is an example of the authors relying too heavily on a behavioral tagging framework when these could simply reflect arousal-related (Nielson et al., 1996; 2014) and/or noradrenergic-related (e.g., McGaugh, 2013) consolidation benefits more broadly. In summary, I think it would strengthen the paper to walk back claims related to behavioral tagging specifically and address the possibility of alternative (but related) mechanisms.

      To summarize, the results of this study are very interesting and the project is very ambitious. There is much therapeutic potential for NITESGON to improve memory and this study represents an important advance towards achieving that goal. The work would primarily be improved by not relying on too many assumptions or inferences, and being more agnostic with respect to certain mechanisms (e.g., whether this is behavioral tagging or general consolidation mechanisms).

    2. Reviewer #2 (Public Review):

      Luckey et al. investigated the mechanisms by which non-invasive transcutaneous electrical stimulation of the greater occipital nerve (NITESGON) enhances long-term memory. They find that NITESGON applied during or after a word-association task enhances memory recall at a retrieval test 7 days later but not at an immediate test, suggesting NITESGON's memory-enhancing effect involves the consolidation process. They show that NITESGON applied during a second spatial memory task not only enhances later recall for that task, but also for an initial word-association memory task unpaired with stimulation administered before the second task. This highlights NITESGON's ability to retroactively strengthen memories and provides further evidence for behavioral tagging. Furthermore, the authors perform a series of in-depth experiments to examine the mechanisms by which NITESGON enhances memory consolidation. They show that NITESGON increases salivary a-amylase levels, a marker of endogenous noradrenergic activity, and spontaneous eye blink levels, a proxy for dopamine levels, both in support of locus coeruleus involvement. Resting-state fMRI results further suggest NITESGON induces increased communication between the locus coeruleus and hippocampus, suggesting a circuit-based mechanism by which NITESGON enhances memory consolidation. Interestingly, the data also indicate that NITESGON's memory-enhancing effect is not sleep-dependent but is dopamine-receptor-dependent.

      The conclusions of this paper are mostly well supported by the data, however, some of the key mechanistic findings lack the appropriate controls required for the authors' claims.

      Strengths<br /> 1) The manuscript is written in an easy-to-read manner with clarity for each of the individual experiments conducted.<br /> 2) The authors provide convincing evidence that NITESGON targets the memory consolidation process and enhances long-term but not short-term memory. This provides a unique non-invasive method for enhancing memory and has an important potential impact on neurocognitive disorders.<br /> 3) The manuscript provides convincing evidence that NITESGON increases LC-hippocampus connectivity as well as MTL gamma power, providing a circuit-based mechanism by which stimulation enhances memory.

      Weaknesses (major)<br /> 1) Adding control groups (sham stimulation) to Experiment 5 and Experiment 8 would be needed to increase confidence that NITESGON's memory-enhancing effects do not depend on sleep but do depend on dopamine receptor activity.<br /> 2) Task order in the interference study in Experiment 4 was randomized during the first visit for task training as well as during the memory test, however, the word-association and spatial navigation tasks used in Experiments 3 and 4 were not counterbalanced during training or memory testing. Thus, the authors cannot rule out the possibility of order effects.<br /> 3) It is unclear how Experiment 3 and Experiment 4 differ. Percent of words recalled is the measure of memory performance, however, there is not a clear measure of interference in Experiment 4 (i.e. words recalled during Memory task II that were from Memory task I).<br /> 4) In Experiment 5 the learning and test phases for the two sleep groups were conducted at different times of day (sleep group: training at 8pm and testing the next morning at 8am, sleep deprivation group: training at 8am and testing at 8pm) which introduces the possibility of circadian effects between the two groups. Additionally, the memory test occurred at the 12h point for this experiment instead of the 7-day point. Therefore, the authors' conclusions are not addressed by this experiment, and it remains unclear whether the 7-day long-term memory effects of NITESGON are sleep-dependent.

      Weaknesses (minor)<br /> 1) Salivary amylase is being used as a proxy of noradrenergic activity, however, salivary amylase levels increase with stress as well, which impacts memory performance. It would be helpful if the authors addressed this and whether they measured other physiological indicators of stress/sympathetic nervous system activation.<br /> 2) Insufficient details of how the blinding experiment was conducted make it difficult to determine whether participants had awareness or subjective responses during the NITESGON stimulation. Adding physiological indicators of heart rate, skin conductance, and respiration would provide a better indicator of a sympathetic nervous system response. Additionally, a series of randomized stimulation and sham trials delivered to the participant would provide a more objective measure of the detectability of the stimulation.<br /> 3) It would be appreciated if the authors could speak to the possible role of the amygdala in the memory-enhancing effects of NITESGON, as this region is a well-known modulator of many types of memory consolidation and is implicated in noradrenergic-related memory enhancement.

    1. Reviewer #1 (Public Review):

      Ras is the first discovered oncogene and KRAS is the most frequently mutated isoform. Recent studies led to the development of mutation specific inhibitors, especially against the KRASG12C mutant. However, unfortunately the patients treated with Adagrasib or others develop resistance due to further gain of function mutations and amplification of KRASG12C allele apart from mutations in the downstream signaling components. One of the oldest approaches to target Rho GTPases like RAS is to compete with the nucleotide binding of RAS and it has for a long time remained difficult owing to the picomolar affinity for GTD/GDP. Gray and colleagues in 2014 tried to overcome these issues by employing GDP derivatives that can undergo covalent reaction with disease specific mutations but Muller etal reported in their previous work (Sci.reports 2017) that the issue with these derivatives was with the loss of reversible affinities for beta modified derivatives for RAS of atleast 10000 fold compared to GDP and GTP. Here the authors present novel GDP derivatives different from Gray and colleagues and demonstrate that they could lock KRASG13C covalently, another important mutant of KRAS in an inactive form with a multiple set of biochemical, structural and cellular assays.

      However, the issue is a lack of evidence to demonstrate "target engagement" in cells and these derivatives need to be developed further as they cannot pass through cell membranes. The complete covalent modification of the compound is achieved at very high pH. Also its not clear if addition of edaGDP would disrupt KRASG13C and effector interaction directly.

    2. Reviewer #2 (Public Review):

      The authors have demonstrated a covalent strategy to target the oncogenic K-Ras(G13C) mutation, which is found in about 3,000 cancer patients in the US each year. G13C is a major contributor to G13 mutations, the next hotspot mutation after codon 12. Moreover, there is no approved therapy for G13 mutations and no published inhibitors of any KRAS G13 mutant proteins, making this a particularly important contribution to the rapidly expanding repertoire of RAS inhibitors. A striking difference in comparison to G12 mutations, mutations occurring at Codon 13 exhibit impaired pM-nucleotide binding affinity of K-Ras. This weaker nucleotide affinity offered the authors the opportunity to develop a nucleotide based inhibitor of a RAS protein. With the high nucleophilicity of cysteine mutation, G13C the authors set out to target this mutant oncogene.

      The authors developed several covalent molecules derived from GDP/GTP, the natural substrate of K-Ras's nucleotide binding pocket, interestingly, not through the oligophosphate chain (explored by Gray and co-workers in an earlier report) but the 2,3-diol of the ribose. This turned out to be a judicious choice for targeting G13C because of the closer proximity to the 2',3' rather than the phosphates. Previous work by Gray et. al. used the phosphate attachment point for the electrophile but this compromised binding affinity overall-whereas the relatively tolerant modifications at 2',3' led to higher affinity electrophilic ligands. This change led to much tighter binders and effective covalent modifiers through C13. With two co-crystal structures resolved, the authors unambiguously showed the covalent cross-linking between artificial G-nucleotides and K-Ras(G13C).

      It is not surprising that one of the major limitations of these GDP-based competitive ligands suffer from permeability issues. GDP or GTP analogs made in this study were not permeable through plasma membrane. The authors nicely worked around these limitations by delivering the fully modified proteins to the cells and measured cell signaling effects. Through electroporation the authors demonstrated the covalent adduct to be able to inhibit downstream signaling by compare introduction of K-Ras WT or K-Ras(G13C) or K-Ras(G13C) covalent adduct.

      A number of very intriguing aspects of the covalent adduct were noted which should guide others in the field, including that the adduct with eda-GTP could get hydrolysed to eda-GDP after the covalent modification of the protein--furthermore GAP stimulation of this adduct still occurred. By use of a non-hydrolyzable form of GTP (CP) this could be prevented and could be a very useful method for preventing hydrolysis after introduction in cells--an application Goody and coworkers applied to a previous covalent base adduct.

      Overall, the manuscript addresses an important problem relating to whether covalent small molecules can engage K-Ras(G13C) and provided two timely co-crystal structures for future research and development.

    3. Reviewer #3 (Public Review):

      Ras mutations are found in almost 25 percent of cancer patients. It has been difficult to directly target Ras proteins due to the lack of druggable pockets on the surface of the protein and the extremely high binding affinity of nucleotides to Ras proteins. Recently a mutant specific irreversible drug that targets the mutation G12C has been FDA approved. This drug binds to a shallow pocket on the surface of Ras and attacks the G12C mutation irreversibly. Another approach is to compete with the nucleotides bound to Ras. An attempt to generate nucleotide competitors that can take advantage of the G12C mutant has been proposed. Nevertheless, these published competitors had much lower affinities compared to endogenous nucleotides which would hinder the covalent modification in the presence of other nucleotides.

      To overcome this, the authors propose to introduce a warhead in the ribose ring. Indeed, this modification did not affect the reversible binding affinity of these nucleotides to Ras wild type, in comparison to GDP and GTP. This finding represents a new opportunity to target G13C ras by competing with the nucleotides in cells. The authors support their claims with the appropriate in vitro experiments. Nevertheless, these experiments were performed at non physiologically high pH (9.5) and those compounds were not able to cross the cellular membrane. Thus, it is too early to draw conclusions regarding the appropriateness of the approach and whether it will prove successful in cells or if it will have medical application.

    1. Reviewer #1 (Public Review):

      In this work, Aggad et al. focused on the multi-folded membrane structure (termed meisosomes) located between the apical extracellular matrix and the epidermal cells of the C. elegans. The authors performed detailed analysis on the morphology and 3D distribution of the meisosomes at different developmental stages of the C. elegans skin. They also investigated factors affecting the biogenesis and reorganization of the meisosomes, as well as the involvement of meisosomes in cuticle synthesis and maintenance. The meisosomes are particularly intriguing membrane structures connecting the epidermis to the extracellular matrix, which potentially have vital functions but were given very little attention before this study. Therefore, the work presented by Aggad et al. is rich in novelty and may greatly benefit the related fields if the main conclusions stand. However, the authors' claims are not very well-supported by the data due to improper use of reporters and mutants, as well as some flaws in experimental design.

      1. One major problem with this manuscript is the investigation about meisosome functions. Instead of generating knockdown animals or mutants that directly and specifically disrupt meisosome structures, the authors used several cuticular collagen mutants, which harbor multiple complex cuticular and epidermal defects. Therefore, the main conclusions drawn from the analysis using collagen mutants, such as "meisosomes may play an important role in attaching the cuticle to the underlying epidermal cell" or "furrow collagens are required for stiffness potentially as they are essential for the presence of normal meisosomes" do not stand well. In fact, it is not surprising that the collagen mutants display a detached cuticle, because the extracellular domains of MUP-4 and MUA-3 (the transmembrane receptors of apical hemidesmosomes that are primarily responsible for tethering the epidermis to the cuticle) both contain vWFA collagen-binding domain (Hong et al., JCB 2001; Bersher et al., JCB 2001). Hence loss of certain collagens in the cuticle directly affects cuticle-epidermis attachment due to defective ligand-receptor interactions is a much more plausible explanation. Likewise, it is more resonable to propose that lack of certain collagens in the cuticle directly affects cuticle stiffness, rather than working indirectly through epidermal meisosomes. In a word, this study did not answer the long-standing question since the 1980s: what are the primary functions of the apical membrane stacks (AKA meisosomes) in the C. elegans epidermis?

      2. Another problem with this manuscript is the representation of meisosome structures by VHA-5::GFP reporter alone from Figure 3 to Figure 7. The authors claim that VHA-5::GFP is a meisosome-specific marker, but only provided indirect and superficial evidence to support this claim: 1) VHA-5::GFP signal is distributed in the same general epidermal area as the majority of meisosomes (so are many other membrane organelles in the C. elegans epidermis);2. VHA-5::GFP does not co-localize with fluorescent markers for MVB, recycling endosomes and autophagolysosomes. By claiming this, the authors made a huge assumption that the overexpressed VHA-5::GFP fusion protein can only possibly associate with four types of organelles (meisosomes, MVB, recycling endosomes and autophagolysosomes) but not any other known or to-be-identified subcellular structures. In addition, a previous study did report that VHA-5 is localized in several other places besides the apical membrane stacks (Liegeois et al., JCB 2006). In a word, there is no solid, direct evidence showing that VHA-5::GFP can specifically represent meisosomes and faithfully visualize meisosome morphology in the C. elegans epidermis. There are also no alternative approaches for meisosome morphological analysis to back up the results obtained from VHA-5::GFP reporter. Therefore, most of the data from Figure 3-7 can only be interpreted as the influence of various factors on the distribution patterns of VHA-5::GFP, not just meisosomes.

    2. Reviewer #2 (Public Review):

      The manuscript by Aggad et al., describes an interesting folded structure that links the epidermis to the cuticle in C. elegans. They analyzed the structure by TEM and tomography and found groups of parallel folds in both L4 and adult animals. They show VHA-5 localizes to this structure and have used VHA-5::GFP transgenic reporter to investigate differently cuticle furrow-related genes by RNAi. It is an important step to describe the character of this structure, which the authors named "meisosomes". However, the structure has been reported and well defined as "apical membrane stacks" in previous studies and reviewed by a few articles (Liegeois et al., 2006, Hyenne et al., 2015, Chisholm and Xu, 2012, Cohen and Sundaram 2020). It is very confusing that the authors want to change the name of this structure.

      The major problem of this paper is that there is not much new information. It is already known that these stacks exist, VHA-5 localizes to the stacks, cuticle damage induces AMPs, "furrowless" dpy mutants result in complete disorganization of the epidermis, defective cuticle structure causes abnormalities via gene expression, etc. The function of these stacks remains unknown. Another issue is the transgenic reporter of VHA-5::GFP, which is not endogenously expressed, and its puncta intensity only reflects the protein distribution but not the stack structure.

    3. Reviewer #3 (Public Review):

      This study by Aggad, Pujol, and colleagues provides some exciting new insights into a largely overlooked organelle/structure present in C. elegans epidermial cells, the "meiosome". Although noted by several previous researchers, this folded-membrane structure was never fully characterized. In particular, the authors provide an important and thorough characterization of meiosome morphology during development. The authors also provide data suggesting that meiosomes may function to provide attachment points between the epidermis and overlying cuticle, although this portion was less clear cut. In addition, the authors show that certain cuticle collagens can affect the morphology and position of meiosomes in addition to the formation of molting-associated actin cables. Some of these latter results, which suggest an 'outside-in' type of patterning regulation, run counter to certain previous models.

      The major strengths of the paper are the novelty of describing a 'new organelle' and the thoroughness and clarity of the morphological analysis. The various EM studies were particularly well done and likely required a good deal of technical development, which may be of use to others in the field. One clear weakness is that it's not currently clear if the reported cuticle detachment defect is due to altered meiosomes, to the altered cuticle composition, or perhaps both, and thus the exact function(s) of meiosomes is left open. Other concerns include the use of extrachromosomally expressed VHA-5::GFP as a meiosome-specific marker. Although this could certainly be the case, it wasn't proven.

    1. Reviewer #1 (Public Review):

      In this manuscript, Wang et al provide a pathway required for the production and degradation of exophers - large neuronal extrusions proposed to discard toxic cargo. Exophers were fairly recently described by this group and have now been observed in mammalian neurons, suggesting a broad importance in neuronal health. How exophers were disposed of by surrounding tissues was not known. Here, the authors identify a pathway required for exopher degradation into small debris (starry night), and intriguingly, genes proposed to be required in the degrading cells (hypodermis) for exopher production in neurons.

      Strengths of the manuscript include significant new insights into a problem that had not been investigated in mechanistic detail, and the combined use of genetics and cell biology to sort genes into pathways involved in exopher production and degradation. Several differences are found between exopher and cell corpse disposal, highlighting the importance of the study. The findings should be of interest to a broad audience.

    2. Reviewer #2 (Public Review):

      Wu Yang et al. investigated how exophers (large vesicles released from neuronal somas) are degraded. They find that the hypodermal skin cells surrounding the neuron break up the exophers into smaller vesicles that are eventually phagocytosed. The neuronal exophers accumulate early phagosomal markers such as F-actin and PIP2, and blocking actin assembly suppressed the formation of smaller vesicles and the clearance of neuronal exophers. They show the smaller vesicles are labeled with various markers for maturing phagosomes, and inhibiting phagosome maturation blocked the breakdown of exophers in to smaller vesicles. Interestingly, they discover that GTPase ARF-6, effector SEC-10/Exocyst, and the phagocytic receptor CED-1 in the hypodermis are required for efficient production of exophers by neurons.

      Strength<br /> The study clearly demonstrates that exophers are eliminated via hypodermal cell-mediated phagocytosis. Exophers are broken down into smaller vesicles that accumulate phagocytic markers, and inhibiting this process shows that exophers are not resolved. The paper does a thorough examination of various markers and mutants to demonstrate this process.

      The hypodermal cells not only engulf these small vesicles, but they also play a role in the formation of exophers. Exopher production is reduced when ARF-6, SEC-10, or CED-1 are knocked down in the hypodermis. This is intriguing because phagocytosis is a critical step in the final elimination of cells, but in this unique situation, it appears that the neuron fails to extrude the exopher without phagocytes.


      Non-professional phagocytes engulfing cell corpses and many other types of cellular debris (e.g. degenerating axons) have been shown in multiple systems and the observations here are not surprising. Many of the markers used in the study are well-established phagocytic markers and do not bring forward a new technological advance.

      What's interesting is that the breakdown of exophers into smaller vesicles and eventual clearance follows a different sequence of events than macrophages. Exophers appear to undergo phagosomal fission before interacting with lysosomes. This would be difficult to appreciate by a general reader.

      While the paper has strengths, it appears that the message is not clear. The title suggests that the reader will learn about how ARF-6 and CED-1 control exopher extrusion. Although this observation is intriguing and maybe the main point of the paper, there does not appear to be a substantial amount of data to support this claim. The only data to back this up is in the final figure and the majority of the paper is focused on how hypodermal cells phagocytose exophers.

      To show exopher secretion is dependent on the hypodermal cells-

      1. Could authors induce exopher production through other means? And test any involvement of CED-1? For example, authors note exopher production increases under stress conditions including expression of mutant Huntingtin protein. It would be intriguing if loss of CED-1 would be sufficient to block or reduce exopher production in that context and would highlight an exciting role for phagocytic cell types.<br /> 2. It is not clear if the CED-1 localization to the exopher is due to CED-1 expression during phagocytosis or is it involved in the extrusion. Perhaps the basal level of CED-1 is important for the extrusion but the strong expression is important for recognition of the exopher.<br /> 3. While the data with ttr-52 and anoh-1 alleles is compelling, do we know that exophers actually expose PS? Especially since at a certain point, the exopher is still attached to the neuronal soma. Is PS still exposed by exopher in CED-1 background?<br /> 4. What is the fate of a neuron that is unable to produce exophers? Could one look at lifespan of ALMR neuron in CED-1, ARF-6 or Sec-10 allele (potentially with specificity to hypodermis)?

    3. Reviewer #3 (Public Review):

      In this paper, the authors examine the fate of exophers ejected from C. elegans neurons overexpressing a presumably aggregated mCherry protein. They show that exophers are taken up by adjacent hypodermal cells, split into smaller fragments, and eventually degraded by lysosome fusion. They identify a number of small GTPases and accessory components, as well as the phagocytic receptor (CED-1) and the likely eat-me signal (phosphatidylserine).

      The manuscript follows up on previous exopher work from some members of the current collaboration, and provides a detailed analysis of exopher fate, that will likely be useful for understanding similar events in other settings. The studies are well done, the images and data are convincing, and the interpretations are generally appropriate.

    1. Reviewer #1 (Public Review):

      This study investigated the roles of sams-1 and sams-4, two enzymes that generate the major methyl donor SAM, in heat stress response and the associated molecular changes. The authors provided evidence that loss of sams-1 resulted in enhanced resistance to heat stress, whereas loss of sams-4 resulted in heightened sensitivity to heat stress. The authors further showed that whereas the basal level of the histone modification H3K4me3 in intestinal nuclei was substantially reduced in sams-1 loss-of-function mutants, H3K4me3 level greatly increased upon heat stress, and this increase depended on sams-4. Additional RNA-seq results revealed largely distinct heat stress-induced RNA expression changes in the sams-1 mutant and sams-4 knockdown worms. The authors further profiled genomic locations of H3K4me3 in sams-1 mutant and sams-4 knockdown worms. Unfortunately, the lack of sufficient technical detail made it difficult to evaluate the H3K4me3 profiling data.

      The paper provided several conceptual advances:<br /> - Uncovering interesting and opposing heat stress phenotype associated with the loss of two related SAM synthases. Thus, even though both SAMS-1 and SAMS-4 produce SAM, the source of SAM production appears to have distinct consequences on the organismal heat stress response.<br /> - Demonstration that SAMS-4 appeared able to compensate for the loss of SAMS-1 upon heat shock, resulting in restoration of the histone mark H3K4me3 in intestinal cells.<br /> - Revealing largely different gene expression changes upon heat shock in animals lacking sams-1 or sams-4. Thus, the gene expression profiles corroborated the differential heat stress response.

      This paper describes one of the first adaptations of CUT&TAG in C. elegans, which can be of high impact on the field. Unfortunately, the lack of experimental detail made it difficult to evaluate the quality of the CUT&TAG data and the consequent interpretations.

      Overall, the paper reported a number of interesting findings that will be of substantial interest to the field. However, the paper in its current form has substantial shortcomings, particularly related to the difficulty in evaluating the validity of H3K4me3 profiling data. The paper would also benefit from further discussion that attempts to reconcile some of the inconsistent results.

    2. Reviewer #2 (Public Review):

      In this manuscript titled "S-adenosylmethionine synthases specify distinct H3K4me3 populations and gene expression patterns during heat stress", the authors Godbole et al investigated how C. elegans SAM synthases, SAMS-1 and SAMS-4, affected gene expression, H3K4 trimethylation (H3K4me3), and the survival under heat stress. They found in this study that SAMS-4 was required for survival during heat shock. They reasoned that SAM supplied by SAMS-4 but not SAMS-1 might be responsible for generating H3K4me3 under heat shock and claimed that the two SAM synthases differentially affected histone methylation and thus gene expression in the heat shock response. This study suggested a stress-responsive mechanism by which the specific isozyme of SAM synthetase provided a specific pool of cellular SAM for H3K4me3. Overall, this study is interesting but descriptive. Lacking necessary controls and mechanistic details weakened the significance of this work.

      Strengths: Very interesting survival phenotypes in the loss of different SAM synthetases; technical success in CUT&tag in C. elegans.

      Weaknesses: No clear conclusion can be drawn about whether and how SAM synthetases affect H3K4me3.

    3. Reviewer #3 (Public Review):

      The manuscript " S-adenosylmethionine synthases specify distinct H3K4me3 populations and gene expression patterns during heat stress " by Godbole et al proposes a novel mechanism by which different S-adenosylmethionine (SAM) synthase enzymes exhibit specificity towards target sequences, thereby providing a layer of control over H3K4 trimethylation (H3K4me3) in Caenorhabditis elegans. The authors detail an extensive investigation of the function of two C. elegans SAM synthase enzymes, SAMS-1 and SAMS-4. They provide evidence that mutation or knockdown of these two enzymes affected gene expression of distinct gene sets and that loss of these enzymes has opposite effects on survival under heat stress. These differential effects are linked to differential effects on histone modification H3K4me3 of specific target gene sets. It is unclear from this work how exactly this specificity may be achieved and some of the data regarding the role of other components of the methylation machinery are somewhat superficial and confusing. Nevertheless, the study suggested a novel mechanism by which H3K4me3 of specific gene sets may be controlled and this mechanism is novel and potentially important.

    1. Reviewer #1 (Public Review):

      Castelán-Sánchez et al. analyzed SARS-CoV-2 genomes from Mexico collected between February 2020 and November 2021. This period spans three major spikes in daily COVID-19 cases in Mexico and the rise of three distinct variants of concern (VOCs; B.1.1.7, P.1., and B.1.617.2). The authors perform careful phylogenetic analyses of these three VOCs, as well as two other lineages that rose to substantial frequency in Mexico, focusing on identifying periods of cryptic transmission (before the lineage was first detected) and introductions to and from the neighboring United States. The figures are well presented and described, and the results add to our understanding of SARS-CoV-2 in Mexico. However, I have some concerns and questions about sampling that could affect the results and conclusions:

      1) The authors do not provide any details on the distribution of samples across the various Mexican States, making it hard to evaluate several key conclusions. Although this information is provided in Supplementary Data 2, it is not presented in a way that enables the reader to evaluate if lineages were truly predominant in certain regions of the country, or if these results are attributable purely to sampling bias. Specifically, each lineage is said to be dominant in a particular state or region, but it was not clear to me if sampling across states was even at all time points. For example, the authors state that most B.1.1.7 genome sampling is from the state of Chihuahua, but it is not clear if this was due to more sequenced samples from that region during the time that B.1.1.7 was circulating, or if the effects of B.1.1.7 were truly differential across the country. The authors do mention sequencing biases several times but need to be more specific about the nature of this bias and how it could affect their conclusions.

      2) It is surprising to see in this manuscript that the B.1.1.7 lineage did not rise above 25% prevalence in the data presented, despite its rapid rise in prevalence in many other parts of the world. This calls into question if the presented frequencies of each lineage are truly representative of what was circulating in Mexico at the time, especially since the coordinated sampling and surveillance program across Mexico did not start until May 2021.

    2. Reviewer #2 (Public Review):

      The authors use a series of subsampling methods based on phylogenetic placement and geographic setting, informed by human movement data to control for differences in sampling of SARS-CoV-2 genomes across countries. Of note, the authors show that 2 variants likely arose in Mexico and spread via multiple introductions globally, while other variant waves were driven by repeat introductions into Mexico from elsewhere. Finally, they use human mobility data to assess the impact of movement on transmission within Mexico.

      Overall, the study is well done and provides nice data on an under-studied country. The authors take a thoughtful approach to subsampling and provide a very thorough analysis. Because of the care given to subsampling and the great challenge that proper subsampling represents for the field of phylodynamics, the paper would benefit from a more thorough exploration of how their migration-informed subsampling procedure impacts their results. This would not only help strengthen the findings of the paper but would likely provide a useful reference for others doing similar studies. Additionally, I would suggest the authors provide a bit more discussion of this subsampling approach and how it may be useful to others in the discussion section of the paper.

    1. Reviewer #1 (Public Review):

      Animal colour evolution is hard to study because colour variation is extremely complex. Colours can vary from dark to light, in their level of saturation, in their hue, and on top of that different parts of the body can have different colours as well, as can males and females. The consequence of this is that the colour phenotype of a species is highly dimensional, making statistical analyses challenging.

      Herein the authors explore how colour complexity and island versus mainland dwelling affect the rates of colour evolution in a colourful clade of birds: the kingfishers. Island-dwelling has been shown before to lead to less complex colour patterns and darker coloration in birds across the world, and the authors hypothesise that lower plumage complexity should lead to lower evolutionary rates. In this paper, the authors explore a variety of different and novel statistical approaches in detail to establish the mechanism behind these associations.

      There are three main findings: (1) rates of colour evolution are higher for species that have more complex colour phenotypes (e.g. multiple different colour patches), (2) rates of colour evolution are higher on island kingfishers, but (3) this is not because island kingfishers have a higher level of plumage complexity than their mainland counterparts.

      I think that the application of these multivariate methods to the study of colour evolution and the results could pave the way for new studies on colour evolution.

      I do, however, have a set of suggestions that should hopefully improve the robustness of results and clarity of the paper as detailed below:

      1) The two main hypotheses tested linking plumage complexity and island-dwelling to rates of colour evolution seem rather disjointed in the introduction. This section should integrate these two aspects better justifying why you are testing them in the same paper. In my opinion, the main topic of the paper is colour evolution, not island-mainland comparisons. I would suggest starting with colours and the challenges associated with the study of colour evolution and then introducing other relevant aspects.

      2) Title: the title refers to both complex plumage and island-dwelling, but the potential effects of complexity should apply regardless of being an island or mainland-dwelling species, am I right? Consider dropping the reference to islands in the title.

      3) The results encompass a large variety of statistical results some closely related to the main hypothesis (eg island/mainland differences) tested and others that seem more tangential (differences between body parts, sexes). Moreover, quite a few different approaches are used. I think that it would be good to be a bit more selective and concentrate the paper on the main hypotheses, in particular, because many results are not mentioned or discussed again outside the Results section.

      4) Related to the previous section, the variety of analytical approaches used is a bit bewildering and for the reader, it is unclear why different options were used in different sections. Again, streamlining would be highly desirable, and given the novel nature of the analytical approach (as far as I know, many analytical approaches are applied for the first time to study colour evolution) it would be good to properly explain them to the reader, highlighting their strengths and weaknesses.

      5) The Results section contains quite a bit of discussion (and methods) despite there being a separate Discussion section. I suggest either separating them better or joining them completely.

      6) The main analyses of colour evolutionary rates only include chromatic aspects of colour variation. Why was achromatic variation (i.e. light to dark variation) not included in the analyses? I think that such variation is an important part of the perceived colour (e.g. depending on their lightness the same spectral shape could be perceived as yellow or green, black or grey or white). I realize that this omission is not uncommon and I have done so myself in the past, but I think that in this case, it is highly relevant to include it in the analyses (also because previous work suggests that island birds are darker than their mainland counterparts). This should be possible, as achromatic variation may be estimated using double cone quantum catches (Siddiqi et al., 2004) and the appropriate noise-to-signal ratios (Olsson et al., 2018). Adding one extra dimension per plumage patch should not pose substantial computational difficulties, I think.

      7) The methods need to be much better explained. Currently, some methods are explained in the main text and some in the methods section. All methods should be explained in detail in the methods section and I suggest that it would be better to use a more traditional manuscript structure with Methods before Results (IMRaD), to avoid repetition (provided this is allowed by the journal). Whenever relevant the authors need to explain the choice of alternative approaches. Many functions used have different arguments that affect the outcome of the analyses, these need to be properly explained and justified. In general, most readers will not check the R script, and the methods should be understandable to readers that are not familiar with R. This is particularly important because I think that the methodological approach used will be one of the main attractions of the manuscript, and other researchers should be able to implement it on their own data with ease. Judging from the R script, there are quite a few analyses that were not reported in the manuscript (e.g. multivariate evolutionary rates being higher in forest species). This should be fixed/clarified.

    2. Reviewer #2 (Public Review):

      In "Complex plumages spur rapid color diversification in island kingfishers (Aves: Alcedinidae)", Eliason et al. link intraspecific plumage complexity with interspecific rates of plumage evolution. They demonstrate a correlation here and link this with the distinction between island and mainland taxa to create a compelling manuscript of general interest on drivers of phenotypic divergence and convergence in different settings.

      This will be a fantastic contribution to the literature on the evolution of plumage color and pattern and to our understanding of phenotypic divergence between mainland and island taxa. A few key revisions can help it get there. This paper needs to get, fairly quickly, up to a point where the difference between plumage complexity and color divergence is defined carefully. That should include hammering home that one is an intraspecific measure, while one is an interspecific measure. It took me three reads of the paper to be able to say this with confidence. Leading with that point will greatly improve the paper if that point gets forgotten then the premise of the paper feels very circular.

      Also importantly, somewhere early on a hypothesized causal pathway by which insularity, plumage complexity, and color divergence interact needs to be laid out. The analyses that currently follow are good ones, and not wrong, but it's challenging to assess whether they are the right ones to run because I'm not following the authors' reasoning very well here. I think it's possible a more holistic analysis could be done here, but I'll refrain from any such suggestions until I better get what the authors are trying to link.

      We also need something near the top that tells us a bit more about the biogeography of kingfishers. Are kingfisher species always allopatric? I know the answer is no, but not all readers will. What I know less well though is whether your insular species are usually allopatric. I suspect the answer is yes, but I don't actually know.

      In short, how do the authors think allopatry/sympatry/opportunity for competition link to mainland vs. island link to plumage complexity? And rates of color evolution? Make this clear upfront.

    3. Reviewer #3 (Public Review):

      In this article, the authors examined color evolution in the kingfishers, a group of birds that have achieved a spectacular diversity of colors and color patterns as they have diverged across the continents and island chains of the globe. Like many other avian taxa, kingfishers on islands often exhibit color patterns distinct from their close relatives. The authors focus here on putting this informally recognized pattern of evolutionary change to a formal test, asking if plumage color diversity and evolutionary rate are elevated on islands. They also explore whether a notable characteristic of some kingfishers - their simultaneous use of many of the coloration mechanisms available in birds - contributes to the evolutionary lability of their color patterns.

      The authors have previously explored how when color varies in birds it is not just in dimensions of color, but also in the distribution of those colors in patches on the body. Summarizing this variation is challenging, and there are statistical obstacles to comparing it in a holistic manner. In this study, the authors use an exceptional set of analyses to study color in total as a multivariate trait. These are the major strengths of the paper. The authors' efforts are somewhat less convincing when they pursue a univariate model fitting on a small number of principal components, but these analyses are not central to the study. And as with all studies using ancestral state reconstruction to test hypotheses, it's an important tool and one that contributes to this study's effectiveness, but we should acknowledge some level of uncertainty with its results.

      The authors report two important relationships in this study. They provide convincing evidence that rates of color evolution are elevated in island kingfishers, without convergence towards a particular island phenotype. They also describe a relationship between the complexity of plumage patterns and the rate at which they evolve, which has fundamental implications for our understanding of the tempo of trait evolution.

      Islands make up a tiny portion of the earth's surface but are home to a seemingly disproportionate amount of life's diversity. This paper makes an important contribution to our understanding of how this diversity is generated, by showing that the evolutionary rate is elevated on islands for traits relevant to mate choice and recognition. The authors find that "plumage complexity, rather than uniformity, provides more phenotypic traits for natural selection to act upon". Given the number of different coloration mechanisms they express, the kingfishers are a unique group in which to study this issue, so I look forward to reading and hearing more from the authors on this issue in the future.

    1. Reviewer #1 (Public Review):

      In this manuscript, Huang et al., assess cognitive flexibility in rats trained on an animal model of anorexia nervosa known as activity-based anorexia (ABA). For the first time, they do this in a way that is fully automated and free from experimenter interference, as apparently experimenter interference can affect both the development of ABA as well as the effect on behaviour. They show that animals that are more cognitively flexible (i.e. animals that had received reversal training) were better able to resist weight loss upon exposure to ABA, whereas animals exposed to ABA first show poorer cognitive flexibility (reversal performance).

      Strengths:<br /> - The development of a fully-automated, experimenter-free behavioural assessment paradigm that is capable of identifying individual rats and therefore tracking their performance.<br /> - The bidirectional nature of the study - i.e. the fact that animals were tested for cognitive flexibility both before and after exposure to ABA, so that direction of causality could be established.<br /> - The analyses are rigorous and the sample sizes sufficient.<br /> - The use of touchscreens increases the translational potential of the findings.

      Weaknesses<br /> - Some descriptions of methods and results are confusing or insufficiently detailed.<br /> - It seems to me that performance on the pairwise discrimination task cannot be directly (statistically) compared to performance on reversal (as in Figure 4E), as these are tapping into fundamentally different cognitive processes (discrimination versus reversal learning). I think comparing groups on each assessment is valid, however.<br /> - Not necessarily a 'weakness' but I would have loved to see some assessment of the alterations in neural mechanisms underlying these effects, and/or some different behavioural assessments in addition to those used here. In particular, the authors mention in the discussion that this manipulation can affect cholinergic functioning in the dorsal striatum We (Bradfield et al., Neuron, 2013) and a number of others have now demonstrated that cholinergic dysfunction in the dorsomedial striatum impairs a different kind of reversal learning that based on alterations in outcome identity and thus relies on a different cognitive process (i.e. 'state' rather than 'reward' prediction error). It would be interesting perhaps in the future to see if the ABA manipulation also alters performance on this alternative 'cognitive flexibility' task.

      Nevertheless, I certainly think the manuscript provides a solid appraisal of cognitive flexibility using more traditional tasks, and that the authors have achieved their aims. I think the work here will be of importance, certainly to other researchers using the ABA model, but perhaps also of translational importance in the future, as the causal relationship between ABA and cognitive inflexibility is near impossible to establish using human studies, but here evidence points strongly towards this being the case.

    2. Reviewer #2 (Public Review):

      Huang and colleagues present data from experiments assessing the role of cognitive inflexibility in the vulnerability to weight loss in the activity-based anorexia paradigm in rats. The experiments employ a novel in-home cage touchscreen system. The home cage touch screen system allows reduced testing time and increased throughput compared with the more widely used systems resulting in the ability to assess ABA following testing cognitive flexibility in relatively young female rats. The data demonstrate that, contrary to expectations, cognitive inflexibility does not predispose to greater ABA weight loss, but instead, rats that performed better in the reversal learning task lost more weight in the ABA paradigm. Prior ABA exposure resulted in poorer learning of the task and reversal. An additional experiment demonstrated that rats that had been trained in reversal learning resisted weight loss in the ABA paradigm. The findings are important and are clearly presented. They have implications for anorexia nervosa both in terms of potentially identifying those at risk also in understanding the high rates of relapse.

    3. Reviewer #3 (Public Review):

      Activity-based anorexia (ABA), which combines access to a running wheel and restricted access to food, is a most common paradigm used to study anorexic behavior in rodents. And yet, the field has been plagued by persistent questions about its validity as a model of anorexia nervosa (AN) in humans. This group's previous studies supported the idea that the ABA paradigm captures cognitive inflexibility seen in AN. Here they describe a fully automated touchscreen cognitive testing system for rats that makes it possible to ask whether cognitive inflexibility predisposes individuals to severe weight loss in the ABA paradigm. They observed that cognitive inflexibility was predictive of resistance to weight loss in the ABA, the opposite of what was predicted. They also reported reciprocal effects of ABA and cognitive testing on subsequent performance in the other paradigm. Prior exposure to the ABA decreased subsequent cognitive performance, while prior exposure to the cognitive task promoted resistance to the ABA. Based on these findings, the authors argue that the ABA model can be used to identify novel therapeutic targets for AN.

      The strength of this manuscript is primarily as a methods paper describing a novel automated cognitive behavioral testing system that obviates the need for experimentalist handling and single housing, which can interfere with behavioral testing, and accelerate learning on the task. Together, these features make it feasible to perform longitudinal studies to ask whether cognitive performance is predictive of behavior in a second paradigm during adolescence, a peak period of vulnerability for many psychiatric disorders. The authors also used machine learning tools to identify specific behaviors during the cognitive task that predicted later susceptibility to the ABA paradigm. While the benefits of this system are clear, the rigor and reproducibility of experiments using this paradigm would be enhanced if the authors provided clear guidelines about which parameters and analyses are most useful. In their absence, the large amount of data generated can promote p-hacking.

      The authors use their automated behavioral testing paradigm to ask whether cognitive inflexibility is a cause or consequence of susceptibility to ABA, an issue that cannot be addressed in AN. They provide compelling evidence that there are reciprocal effects of the two behavioral paradigms, but do not perform the controls needed to evaluate the significance of these observations. For example, the learning task involves sucrose consumption and food restriction, conditions that can independently affect susceptibility to the ABA. Similarly, the ABA paradigm involves exercise and restricted access to food, which can both affect learning.

      In the Discussion, the authors hypothesize that the ABA paradigm produces cognitive inflexibility and argue that uncovering the underlying mechanism can be used to identify new therapeutic targets for AN. The rationale for their claim of translational relevance is undermined by the fact that the biggest effect of the ABA paradigm is seen in the pair discrimination task, and not reversal learning. This pattern does not fit clinical observations in AN.

      In summary, the significance of this manuscript lies in the development of a new system to test cognitive function in rats that can be combined with other paradigms to explore questions of causality. While the authors clearly demonstrate that cognitive flexibility does not promote susceptibility to ABA, the experiments presented do not provide a compelling case that their model captures important features of the pathophysiology of AN.

    1. Reviewer #1 (Public Review):

      This article describes simultaneous surface recordings with a transparent electrode array and two-photon calcium imaging in the mouse cortex. The study shows that spiking activity recorded by surface electrodes or imaged layer 2/3 activity is decoupled. Moreover, simulations indicate that this decoupling may be due to a dominance of L1 projecting axons (input to the cortex) in surface spiking activity.

      This is a rigorous study capitalizing on the new Windansee surface recording device, which provides extremely useful evidence that surface electrodes may not be able to capture information processed in the cortical layers. Recordings and simulations seem adequately performed. The indication that axons contribute significantly to multiunit activity is extremely important for the interpretation of multiunit activity in surface recordings. Here the claim is limited to surface recording, and one wonders to which extent this conclusion would transpose to recordings made with penetration electrodes.

    2. Reviewer #2 (Public Review):

      The manuscript describes a novel transparent electrode array and demonstrates its combination with two-photon calcium imaging in mouse neocortex. Using a computational model, the authors propose that surface multi-unit activity mainly reflects L1 axonal activity and they find a small population of L2/3 neurons that correlates with this activity. While the multi-modal approach with the innovative device in our view is interesting and potentially useful, we have several technical and scientific concerns that should be addressed by the authors.

      Strengths:<br /> We find the general scope of this manuscript, to establish a hybrid electrophysiological and optical approach for studying neocortical activity, very interesting and relevant. The authors provide a compelling use case for combined ECoG and two-photon imaging. While extracellular action potentials have been recorded from the cortical surface, the underlying source is unknown and the device and techniques introduced by the authors are appropriate to address this question. The introduced device can be implanted chronically and has good long-term stability, providing longitudinal optical and electrical recordings from the cortex. The authors perform recordings in awake, head-fixed animals which provides the opportunity to relate ECoG and single-cell data to the animal's behavioral state. The combination of empirical data and biophysical modelling is a powerful means by which to answer such questions.

      Weaknesses:<br /> The central claim of the paper relies heavily on the computational model and the physiological data could be more completely analyzed. Based on a sample of 136 L2/3 neurons the authors find a small proportion (13%) that correlates with the ECoG MUA (eMUA). Based on this, they use a model to show that ECoG MUA likely reflects axonal spikes. They then posit that these layer 2/3 neurons are tightly correlated to the layer 1 input. The presentation of their data and the specifics of their model makes it difficult to assess the validity of this claim. They do not sufficiently discuss possible confounds in the data, caveats of their model, or alternative explanations of the observed low proportion of L2/3 neurons that correlate with the ECoG MUA.

      Most relevantly, the authors do not measure single units with their ECoG. The eMUA is a complex mixture of many neuronal sources, and interpretation is therefore difficult. They relate the calcium transients of small populations of single L2/3 neurons with the aggregate measure of population activity reflected in eMUA. It is possible that the eMUA reflects population activity in the local circuit and might therefore have a low correlation with individual single units. Critically, there is no information on the sensitivity of calcium recordings. Do the imaging data detect single action potentials, or are they biased to bursts of more than 1 AP?

      The analysis pipeline and values used for computing the correlation coefficients are counterintuitive. The fluorescence data are first interpolated from 15 Hz to 4 kHz and then both eMUA and imaging data are effectively down-sampled to 2 Hz. A single correlation coefficient is then estimated for each neuron, regardless of behavioral state, even though the authors themselves show that the activity of single neurons and the ECoG signal depend on the state of the animal.

      There is also insufficient information on the weight of the implant and its effect on mouse behavior. How does the movement of implanted and non-implanted mice differ? Must mice be singly housed? Finally, the modeling parameters are highly specific, using independently driving spikes, while the activity of neurons can be highly correlated. Likewise, the contribution of tangentially oriented axons that could relate to long-range connections conveying information related to the animal's motion or level of arousal is not considered. The manuscript would benefit from further analysis of the physiological data, consideration of alternative explanations and forthright discussion of limitations and caveats of their device and approach.

    3. Reviewer #3 (Public Review):

      The authors have developed a new form of transparent surface multielectrode integrated into an imaging window, enabling simultaneous recording of electrical activity at the surface of the cortex combined with two-photon imaging through the window and electrode. The authors characterise the electrical signals and use simulations to argue that they reflect the activity of axons in layer 1. This is then correlated with calcium imaging signals from layer 2/3 pyramidal cells. A subset of these displayed strong correlations with the layer 1 activity.

      The raw electrical recordings appear to be contaminated by large movement artefacts. The authors attempt to decompose the signal into neuronal activity and artefact. The independent component analysis (ICA) employed yields plausible results. However, there is no definitive validation of this procedure.

      The simulations strongly suggest that only layer 1 axons will generate significant neuronal signals at the surface, but the authors have not attempted to reconstruct the multiunit activity in the simulations, which could provide additional assurance for their interpretation.

      A small fraction of pyramidal cells has activity strongly correlated with the signal at the surface electrode. However, the authors have not examined whether the distance from neuron to the electrode influences the strength of correlation. It remains possible that the differential correlation reflects a distance effect rather than the existence of two populations.

    1. Reviewer #1 (Public Review):

      This is an interesting manuscript that highlights the potential for 'clogging' of import channels by mutant proteins to promote mitochondrial dysfunction in disease. One of the challenges with this study is deconvoluting potential loss-of-function phenotypes associated with reductions in ANT1/AAC2 from gain-of-toxicity phenotypes linked to import clogging. This was addressed primarily in yeast, showing that phenotypes associated with overexpression of mutants (e.g., reduced growth on glucose media). The experiment showing that yeast AAC2 clogs import was also convincing including both in vitro and in vivo characterization, although it isn't clear why the proteomic experiments were performed with acute expression of A128P instead of the 'superclogger' double mutant. The extension of this work to mammalian cells and then mice is also admirable. However, the quality of characterization does begin to decline when moving into mammalian models. For example, there is no clear evidence that observed phenotypes can be attributed to gain of toxicity instead of loss of function in mammalian cells and mice. There are similarities to yeast, but this needs to be better defined in my opinion. Lastly, I have questions related to the mouse model, such as how do these phenotypes compare to KO animals and why were homozygous mice used in some situations and heterozygous mice used in others.

      Overall, this manuscript is interesting, as it describes a mechanism whereby mutant proteins can lead to import deficiencies in the context of disease. The strengths primarily reside with the yeast work, where the demonstration of import clogging and the functional implications of this clogging are best defined. The transition to mammalian cells and mice is admirable as well, but doesn't reach the same level of characterization, leaving open the possibility that the observed effects could be attributed (at least in part) to loss of function of ANT1.

    2. Reviewer #2 (Public Review):

      Mitochondrial dysfunction is now widely recognized as an underlying cause of many human diseases. In many cases, however, very little is known about the molecular etiology of mitochondrial disorders. In this comprehensive study Coyne et al. describe a mechanism by which dominant pathogenic variants of adenine nucleotide translocase Aac2p/ANT1 impair mitochondrial protein import pathway leading to cytotoxicity and mitochondrial dysfunction. By elucidating the fate of this protein in yeast, human cell culture, and murine models, the authors showed that mutant Aac2p variants accumulate the outer membrane translocase TOM complex jamming up mitochondrial protein import and affecting TIM22-mediated carrier import pathway, thus causing proteostatic stress. Furthermore, they showed that the i-AAA protease Yme1p and not the ubiquitin-proteasome system is responsible for proteolytic removal of the mutant Aac2p variants. Finally, the demonstrated that mitochondrial protein import clogging caused by the ANT1 A114P, A123D variant causes severe dominant neurodegenerative phenotype in mice, which resembles neuromuscular disease manifestations in humans. The authors propose this as a candidate pathological mechanism in ANT1-linked human disorders and by extension, to other diseases arising from defects in mitochondrial protein import.

      Overall, this is a well-designed and thoroughly executed study that reports on a novel aspect of ANT1 associated dysfunction and provides mechanistic insights into the pathological mechanisms at play.

    3. Reviewer #3 (Public Review):

      Dominant pathogenic variants of the Aac2/Ant1 ATP transporter cause disease by an unknown mechanism. In this manuscript the authors aim to reveal how these gain of function mutants impair cellular and mitochondrial health. To characterize the phenotype of Aac2 mutants in yeast, the authors use a series of single and double Aac2 mutations, within the 2nd and 3rd transmembrane domains that are associated with human diseases. Aac2A128P,A137D mutant, which caused high toxicity and damaged the mitochondrial DNA was selected for further analysis. This mutant was not imported efficiently into mitochondria and exhibited an increased association with TOM, suggesting that it clogs the TOM translocase. As a result, expression of Aac2A128P,A137D led to impaired import of other mitochondrial proteins. Several findings suggested that the single mutant Aac2A128P impaired mitochondrial import in a similar manner: 1. mass spec analysis revealed its increased association with cytosolic chaperones, TOM and TIM22 subunits, 2. Aac2A128P overexpression led to global mitochondrial protein import deficiency, demonstrated by HSP60 precursor accumulation and activation of stress responses (transcription of chaperons, proteosome induction, and CIS1).<br /> Parallel mutants of human Ant1 (AntA114P and Ant1A114P,A123D) were ectopically expressed in HeLa cells. The mutants were demonstrated to clog TOM and cause a global defect in mitochondrial protein import. This was confirmed in tissues from Ant1A114P,A123D/+ knock-in mice. The Ant1A114P,A123D/+ mice exhibited decreased maximal mitochondrial respiration in muscles. Examination of the skeletal muscle myofiber diameter and COX and SDH activity revealed that Ant1A114P,A123D expression in heterozygous mice acts dominantly and causes a myopathic phenotype and in some case neurodegeneration.

      Major strengths -

      The ability of proteins to clog TOM and sequentially disrupt protein import into mitochondria was demonstrated in recent years. However, till now this was achieved using chemicals, artificial cloggers and overexpression of mitochondrial proteins. This study reveals, for the first time, that disease associated variants of native mitochondrial proteins can clog the entry into the organelle. Thus, this work demonstrates that TOM clogging is a physiological relevant phenomenon that is involved in human diseases.

      The manuscript is well-written and the experiments are well-designed, presenting convincing data that mostly support the conclusions. The methods used are well-establish and suitable techniques that are often used in the field. This work took advantage of 3 different biological systems/model organism, yeast, cell culture, and mice tissues, to validate the results, show conservation, and exploit the strengths of each system.

      Overall, this study is impactful, greatly contributes to the field and should be of interest to the general scientific community. The work sheds light of the mechanisms by which Ant1 pathogenic mutants impact cellular health and provides evidence for the involvement of translocases clogging and impaired protein import in human diseases. The gain of function Aac2/Ant1 mutants will provide a new and powerful tool for future studies of mitochondrial quality control and repair mechanisms.

      Major weaknesses -

      1. The evidence for clogging of mitochondrial translocases and for general defect in protein import are solid. However, there are not enough evidence to conclude that all phenotype seen in mice and yeast are directly connected to clogging.

      2. This work implies that Aac2/Ant1 variants can clogg TOM, TIM22, or both. Clogging of TIM22 is novel and interesting but is not fully discussed in the manuscript, as well as the possibility that clogging of different translocases can result in different defects.

    1. Reviewer #1 (Public Review):

      This manuscript presents a comparison between models that may explain psychophysical performance in sensory integration tasks, where a subject essentially has to count stimulus samples and make a motor report about the final count.

      The work has many technical strengths:

      - The problem of model mimicry is clearly articulated.

      - The work shows that the use of discrete sample stimulus (DSS) is key for being able to disambiguate multiple candidate mechanisms that could possibly underlie the observed behavioral data.

      - The authors use rigorous model comparison and analysis techniques, some (like the integration maps) newly developed for the current application.

      - The model comparison involves both qualitative and qualitative contrasts between alternative models.

      - Consistent results are obtained with several data sets involving humans, monkeys, and rats.

      - The results provide insight into why the simpler alternative models (the snapshot and extrema detection models) fail.

      No glaring weaknesses were found in this manuscript. However, there are some limitations that are worth noting, to put things into context:

      - The results are consistent with what has become a well-known principle of operation of sensory-motor circuits, namely, that they are highly effective at integrating sensory evidence over time. Thus, the results are not particularly surprising.

      - The results are valuable in that they specifically refute two mechanisms that had been recently proposed as potential alternatives to the more standard temporal integration. To some, these alternative mechanisms may have seemed somewhat far-fetched to begin with, as they would lead to suboptimal performance in general. Nevertheless, settling the question was important.

      - Temporal integration and accumulation of evidence have been the focus of many computational studies in systems neuroscience. Although these are certainly important functions, sensory-guided choices require the deployment and coordination of numerous sensory, motor, and cognitive mechanisms, of which integration is just one.

      Overall, this is a valuable study that has important theoretical implications in the field of computational neuroscience. It presents a compelling case that temporal integration is a common capability of sensory-motor circuits and that it explains a variety of behavioral data sets much better than two simpler, alternative mechanisms.

    2. Reviewer #2 (Public Review):

      The authors' goal is to uncover the most likely method used by mammals to make choices based on a time-limited stream of noisy incoming sensory data. To achieve this, they analyze with great rigor several large datasets obtained from tightly controlled two-alternative forced choice behavioral experiments. The tight control of fluctuating incoming sensory input over a large number of trials allows the authors to extract the influence of different components of that input on the behavioral choice. The conditional analysis, showing the impact of early information on the importance of later information, or vice versa, is an excellent new technique.

      They compare three models and find one based on a form of weighted integration of evidence across time is very strongly favored compared to models in which only short segments of the sensory input are used, or the most extreme fluctuations of the sensory input generate a response. Overall, the results clearly do indicate that the integration-like family of models outperforms the other families. The authors succeed well in giving a fair comparison of the different families of models, allowing multiple parameters to be optimized to test different versions of each model.

      It should be said that the integration model is a strange type of integration, as the weight of incoming evidence depends on the time at which it arrives-by a factor of 4 in one animal (Fig. 2)-and with an over-weighting of evidence in the middle of the sequence in one case, while the more expected effects of primacy and recency (over-weighting of early or late evidence) in another. It would be nice to see more discussion of how these differences might arise across animals, what it may say about the neural circuit performing such unbalanced integration, and how suboptimal such differential weighting of evidence is. This is important, as in some discussions integration is contrasted with state transitions, which are akin to integration over a barrier, and not necessarily ruled out by the models compared here.

    1. Reviewer #1 (Public Review):

      Jordan and Keller investigated the possibility that sensorimotor prediction error (mismatch between expected and actual inputs) triggers locus coeruleus (LC) activation, which in turn drives plasticity of cortical neurons that detect the mismatch (e.g. layer 2/3 neurons in V1), thus updating the internal presentation (expected) to match more the sensory input. Using genetic tools to selectively label LC neurons in mice and in vivo imaging of LC axonal calcium responses in the V1 and motor cortex in awake mice in virtual reality training, they showed that LC axons responded selectively to a mismatch between the visual input and locomotion. The greater the mismatch (the faster the locomotion in relation to the visual input), the larger the LC response. This seemed to be a global response as LC responses were indistinguishable between sensory and motor cortical areas. They further showed that LC drove learning (updating the internal model) despite that LC optical stimulation failed to alter acute cellular responses. Responses in the visual cortex increased with locomotion, and this was suppressed following LC phasic stimulation during visuomotor coupled training (closed loop). In the last section, they showed that artificial optogenetic stimulation of LC permitted plasticity over minutes, which would normally take days in non-stimulated mice trained in the visuomotor coupling mode. These data enhance our understanding of LC functionality in vivo and support the framework that LC acts as a prediction error detector and supervises cortical plasticity to update internal representations.

      The experiments are well-designed and carefully conducted. The conclusions of this work are in general well supported by the data. There are a couple of points that need to be addressed or tested.

      1) It is unclear how LC phasic stimulation used in this study gates cortical plasticity without altering cellular responses (at least at the calcium imaging level). As the authors mentioned that Polack et al 2013 showed a significant effect of NE blockers in membrane potential and firing rate in V1 layer2/3 neurons during locomotion, it would be useful to test the effect of LC silencing (coupled to mismatch training) on both cellular response and cortical plasticity or applying NE antagonists in V1 in addition to LC optical stimulation. The latter experiment will also address which neuromodulator mediates plasticity, given that LC could co-release other modulators such as dopamine (Takeuchi et al. 2016 and Kempadoo et al. 2016). LC silencing experiment would establish a causal effect more convincingly than the activation experiment.

      2) The cortical responses to NE often exhibit an inverted U-curve, with higher or lower doses of NE showing more inhibitory effects. It is unclear how responses induced by optical LC stimulation compare or interact with the physiological activation of the LC during the mismatch. Since the authors only used one frequency stimulation pattern, some discussion or additional tests with a frequency range would be helpful.

    2. Reviewer #2 (Public Review):

      The work presented by Jordan and Keller aims at understanding the role of noradrenergic neuromodulation in the cortex of mice exploring a visual virtual environment. The authors hypothesized that norepinephrine released by Locus Coeruleus (LC) neurons in cortical circuits gates the plasticity of internal models following visuomotor prediction errors. To test this hypothesis, they devised clever experiments that allowed them to manipulate visual flow with respect to locomotion to create prediction errors in visuomotor coupling and measure the related signals in LC axons innervating the cortex using two-photon calcium imaging. They observed calcium responses proportional to absolute prediction errors that were non-specifically broadcast across the dorsal cortex. To understand how these signals contribute to computations performed by V1 neurons in layers 2/3, the authors activated LC noradrenergic inputs using optogenetic stimulations while imaging calcium responses in cortical neurons. Although LC activation had little impact on evoked activity related to visuomotor prediction errors, the authors observed changes in the effect of locomotion on visually evoked activity after repeated LC axons activation that were absent in control mice. Using a clever paradigm where the locomotion modulation index was measured in the same neurons before and after optogenetic manipulations, they confirmed that this plasticity depended on the density of LC axons activated, the visual flow associated with running, and the concurrent visuomotor coupling during LC activation. Based on similar locomotion modulation index dependency on speed observed in mice that develop only with visuomotor experience in the virtual environment, the authors concluded that changes in locomotion modulation index are the result of experience-dependent plasticity occurring at a much faster rate during LC axons optogenetic stimulations.

      The study provides very compelling data on a timely and fascinating topic in neuroscience. The authors carefully designed experiments and corresponding controls to exclude any confounding factors in the interpretation of neuronal activity in LC axons and cortical neurons. The quality of the data and the rigor of the analysis are important strengths of the study. I believe this study will have an important contribution to the field of system neuroscience by shedding new light on the role of a key neuromodulator. The results provide strong support for the claims of the study. However, I also believe that some results could have been strengthened by providing additional analyses and experimental controls. These points are discussed below.

      Calcium signals in LC axons tend to respond with pupil dilation, air puffs, and locomotion as the authors reported. A more quantitative analysis such as a GLM model could help understand the relative contribution (and temporal relationship) of these variables in explaining calcium signals. This could also help compare signals obtained in the sensory and motor cortical domains. Indeed, the comparison in Figure 2 seems a bit incomplete since only "posterior versus anterior" comparisons have been performed and not within-group comparisons. I believe it is hard to properly assess differences or similarities between calcium signal amplitude measured in different mice and cranial windows as they are subject to important variability (caused by different levels of viral expression for instance). The authors should at the very least provide a full statistical comparison between/within groups through a GLM model that would provide a more systematic quantification.

      Previous studies using stimulations of the locus coeruleus or local iontophoresis of norepinephrine in sensory cortices have shown robust responses modulations (see McBurney-Lin et al., 2019, https://doi.org/10.1016/j.neubiorev.2019.06.009 for a review). The weak modulations observed in this study seem at odds with these reports. Given that the density of ChrimsonR-expressing axons varies across mice and that there are no direct measurements of their activation (besides pupil dilation), it is difficult to appreciate how they impact the local network. How does the density of ChrimsonR-expressing axons compare to the actual density of LC axons in V1? The authors could further discuss this point.

      In the analysis performed in Figure 3, it seems that red light stimulations used to drive ChrimsonR also have an indirect impact on V1 neurons through the retina. Indeed, figure 3D shows a similar response profile for ChrimsonR and control with calcium signals increasing at laser onset (ON response) and offset (OFF response). With that in mind, it is hard to interpret the results shown in Figure 3E-F without seeing the average calcium time course for Control mice. Are the responses following visual flow caused by LC activation or additional visual inputs? The authors should provide additional information to clarify this result.

      Some aspects of the described plasticity process remained unanswered. It is not clear over which time scale the locomotion modulation index changes and how many optogenetic stimulations are necessary or sufficient to saturate this index. Some of these questions could be addressed with the dataset of Figure 3 by measuring this index over different epochs of the imaging session (from early to late) to estimate the dynamics of the ongoing plasticity process (in comparison to control mice). Also, is there any behavioural consequence of plasticity/update of functional representation in V1? If plasticity gated by repeated LC activations reproduced visuomotor responses observed in mice that were exposed to visual stimulation only in the virtual environment, then I would expect to see a change in the locomotion behaviour (such as a change in speed distribution) as a result of the repeated LC stimulation. This would provide more compelling evidence for changes in internal models for visuomotor coupling in relation to its behavioural relevance. An experiment that could confirm the existence of the LC-gated learning process would be to change the gain of the visuomotor coupling and see if mice adapt faster with LC optogenetic activation compared to control mice with no ChrimsonR expression. Authors should discuss how they imagine the behavioural manifestation of this artificially-induced learning process in V1.

      Finally, control mice used as a comparison to mice expressing ChrimsonR in Figure 3 were not injected with a control viral vector expressing a fluorescent protein alone. Although it is unlikely that the procedure of injection could cause the results observed, it would have been a better control for the interpretation of the results.

    1. Reviewer #1 (Public Review):

      This is a quite nice work equipped with healthy scientific substance underpinned by a solid mathematical approach.

      The authors based on a PGG with the threshold; M (that ranges; 1 < M < N, where N indicates the game size), whether cooperation bringing fruit or not, in which, according to the commonly used parameterization, b and c mean the cooperation fruit and the cost for cooperation. As a kernel in their model, they presumed that an individual will lose his endowment (cooperation fruit in this context) with a probability r, which represents the risk level of collective failure (Eqs. (1 & 2)). Let alone, they presumed a well-mixed and infinite mother-population to ensure their analytical formulation and analysis, and to apply the replicator dynamics. Subsequently, they presumed the co-evolution of cooperation fraction; x, and risk level; r, by introducing another dynamical system for r, of which the general form is defined by Eq. (3).

      For a down-to-earth discussion, they presumed two types of concrete forms for non-linear function; U(x,r). Both types premise the so-called logistic type form; containing r*(1 - r). One is what-they-called Linear; Eq. (5). Another is Eq. (7), called Exponential. Up to here, all the modeling approach is well depicted and quite understandable.

      By exploring some numerical results backed by their theoretical ground, the authors got phase diagram (Figs. 3 and 5); whether a co-evolutionary destiny evaluated by (x,r) being absorbed by the dominance of unwilling (less cooperative) situation (say, D-dominant); (0,1), or by bi-stable equilibrium (either better state or D-dominant depending on an initial condition) along u (parameter appeared in the dynamical equation for r) and c/b (roughly speaking; it implies dilemma strength).

      The result seems interesting and conceivable. As a rough sketch, the two types of U(x,r) seem less different. But the higher absorbing point of (x,r) out of the two cases of bi-stable equilibria is mutually different (yellow region). The authors deliberately illustrated the time-series of properties and trajectory of (x,r) in some representative cases in Figs. 4 and 6.

      As a whole, I really evaluate this work as impressive.

    2. Reviewer #2 (Public Review):

      Liu, Chen and Szolnoki investigated the coupled dynamics of individual cooperation level and collective risk (i.e. the probability of future loss of all endowment). Their model encapsulates the assumption that not only does risk affect individual decision-making, but that there is also feedback between individual strategies, i.e. the level of individual contributions, and the level of risk. The authors investigate two main forms of this feedback, considering strategies linearly affecting the evolution of risk as well as non-linear (exponential) feedback. They mathematically analyze both these dynamical systems, identifying the fixed points, parametrized by the enhancement rate of defection u and the cost/benefit ratio of cooperation, and analyzing the stability of these points. The results of this systematic analysis show that, while the undesirable equilibrium state of full defection and high risk is always stable independent of the form of the feedback, the coevolutionary dynamics can exhibit a wide range of behaviors. In particular, depending on the initial conditions (frequency of cooperators), sustainable cooperation levels can be reached. This can happen by convergence to a stable fixed point with positive cooperation rates; additionally, the authors also prove that a Hopf bifurcation can take place in the system, such that a stable limit cycle with persistent oscillations in strategy and risk state can appear. Interestingly, the evolutionary outcomes do not depend significantly on the character of the feedback between strategy and risk. These theoretical results are supplemented by representative numerical examples, visualizing the phase plane and temporal dynamics of cooperation and risk for particular initial conditions and parameters.

      The main conclusions of the paper are fully supported by the results, as they are directly derived from the comprehensive mathematical analysis of the coevolutionary dynamics and do not rely on external data. Additionally, the stability analysis is clean and the comprehensive numerical examples deepen the reader's understanding. Another strength of the paper is the fact that the considered model is complex enough to be able to still represent somewhat realistic settings while being simple enough to rigorously analyze. One particularly interesting finding is the fact that the exact form of the risk feedback function or its speed does not play a very significant role in the outcome of the dynamics.

      The paper hence adds to the literature on the coevolution of environment and strategies in a productive way and will be of interest to various research communities in mathematical biology/ecology and decision-making.

    1. Reviewer #1 (Public Review):

      The role of increased temperature on immunity and homeostasis in cold-blooded vertebrates is an understudied yet important field. This work not only examines how immunity is impacted by fever, but also incorporates an infection model and examines resolution of the response. This work can serve as a model for other groups interested in the study of hyperthermia and immunity.

      Generally speaking, I agree with the authors' strategy and interpretations of the data.

      - In the Introduction, the authors chose to begin with how fever in endotherms impact the immune system. Considering that this work exclusively examines the response of a teleost (goldfish), the authors might consider flipping the way they present this work. After all, cold-blooded vertebrates rely on this response because of their basic physiology.

      - I thought the set up of the work in figure 1 was innovative and could provide an example of how to study such a problem.

      - Figure 2 was (to me) unexpected. One would not expect such tight response to hyperthermia and infection. This experiment in and of itself was quite interesting, and worth following up in future experiments (by the authors and other groups).

      - The other work, on the response to infection and the resolution of infection were unique to this paper, and (sorry to be repetitive) can be an example of how to devise such studies.

      - On the other hand, I am not sure this is a study of "fever." That implies how increased temperature impacts immunity and resolution in endotherms. Perhaps the authors could temper the comparisons between cold- and warm-blooded vertebrates regarding the response to hyperthermia.

    2. Reviewer #2 (Public Review):

      Fever is an ancient and conserve response to infection from invertebrates to humans. However, the functional benefits of engaging fever responses are not clear, especially when it comes to moderate fever responses where pathogen growth Is not impaired by temperature. This study aims to develop a natural in vivo fever model in fish that overcomes many of the technical challenges to investigate fever in mammals. In ectotherms, fever is manifested as a behavioral response by which animals move to warmer temperatures. By using this new developed in vivo behavioral ring, the present study reveals new functional roles for fever in vertebrates. Additionally, upon infection, sickness behavior did not only consist of fever, but two novel lethargic behaviors not previously described in fish. The experimental evidence is compelling and supports the authors' conclusions. The data presented strongly indicates that moderate fever levels are critical for fine tuning immune responses to pathogens. By triggering earlier but weaker antimicrobial defenses, moderate fever in teleosts results in controlled inflammation and improved wound healing. These exciting results reveal novel roles of fever as a way to minimize the collateral damage that inflammatory responses often cause to the host. This work advances our conceptual view of the evolutionary advantages that fever brings to host-pathogen interactions. The technological development of the annular temperature preference tank can now become the gold standard platform to investigate the consequences of fever during teleost infection.

    1. Reviewer #1 (Public Review):

      In this report, Yeung et al studied a mutation in Orai1 channels (L138F) that is associated with constitutive CRAC channel activity and tubular aggregate myopathy (TAM) in humans. They put forth a model whereby substitution with large amino acids at position L138 on TM2 or the neighboring T92 on TM1 causes a steric clash between TM1 and TM2 and elicits a highly Ca2+ selective current in the absence of STIM1, the ER Ca2+ sensor protein that is the physiological activator of Orai channels. The authors went on to study one typical biophysical property of Orai1-mediated CRAC channels which is the fast Ca2+-dependent inactivation (CDI), after the surprising finding of the presence of CDI in CRAC currents mediated by T92 and L138 Orai1 mutants in the absence of STIM1. The authors showed differences in CDI between WT and mutants when using weak vs strong buffers and through computation and experimentation, they show that the Orai1 mutants have enhanced cytosolic Ca2+ sensitivity, which could be normalized when STIM1 was present. The experiments are carefully conducted and the manuscript is clearly written. The study has significant novelty and impact.

    2. Reviewer #2 (Public Review):

      The manuscript "A human tubular aggregate myopathy mutation unmasks STIM1-independent rapid inactivation of Orai1 channels" describes the effects of a disease-related gating checkpoint at the TM1-TM2 interface. The authors suggest that the mutation of one of the two oppositely located positions T92 - L138 into a large amino acid leads to constitutive activity due to steric clash. Notably, the mutants also exhibit robust Ca2+ dependent inactivation (CDI) suggesting that this feature is intrinsic to the Orai1 channel, and not as previously thought a key process that is triggered by STIM1. Nevertheless, STIM1 is able to fine-tune Ca2+ selectivity and CDI.

      This study provides an extensive electrophysiological characterization of the tubular aggregate myopathy (TAM)-disease-related Orai1 L138F mutation and based on mutational studies provides compelling evidence that constitutive activity is caused by a steric clash between TM1/TM2 Orai helices. Additionally, yet unexpectedly, the constitutive Orai1 mutants exhibit CDI behavior which is thoroughly characterized by experiments using various intracellular Ca2+-buffering reagents. By this, it is proposed that the Orai1 T92W mutant shows increased sensitivity to intracellular Ca2+. This is further revealed in a sophisticated tow step protocol, which would profit from additional control experiments. The unusual behavior of the T92W Orai1 mutant is "corrected" to that of the Orai1 wild-type form by the presence of STIM1.

    3. Reviewer #3 (Public Review):

      In this paper, Yeung et al., use patch-clamp electrophysiology measurements combined with structural analyses and mutagenesis to compellingly reveal how the tubular aggregate myopathy (TAM)-associated Orai1 L138F mutation leads to the gain of CRAC channel function. They discover that L138F not only constitutively activates Orai1-composed channels but also enhances Ca2+-dependent inactivation (CDI). The authors find that the L138F gain of function occurs due to a steric clash with T92 from an adjacent subunit and that introduction of a bulky residue at the T92 position similarly activates CRAC channels and enhances CDI in the absence of STIM1. Nevertheless, co-expression of STIM1 with strongly activating T92W or L138F mutants regularized the CDI to wild-type levels. Collectively, the work represents an important conceptual advancement, exposing that STIM1 is not necessary for CDI and that Orai1 likely contains the Ca2+ sensor intrinsically for this phenomenon.

      Strengths:<br /> The authors use rigorous and careful electrophysiological measurements to probe how the TAM-related mutation (L138F) affects the biophysical properties of CRAC channels. The extensive and systematic mutagenesis (i.e. substitution to every possible amino acid at the T92 and L138 sites) coupled with these functional assessments reveal a steric clash between L138F and T92 and provide a complete picture of how any residue type at the so-called T92/L138 lever point may contribute to constitutive CRAC and CDI activity. The use of available high-resolution structural data to interpret functional data, rationalize the consequence of new mutations related to the mechanisms of L138F dysfunction, and generate new hypotheses is a strength of the research. Overall, the work provides a considerable conceptual advance in terms of understanding the molecular requirements for CRAC and CDI activity; in particular, the discovery that CDI can occur independently of STIM1 and the notion that Orai1 may contain an intrinsic Ca2+ sensor that regulates CDI are important steps forward for the field.

      Weaknesses:<br /> While the work provides a phenomenological advancement regarding CRAC channel regulation and pinpoints new important residues for function, some aspects of the study appear incomplete. It was shown that STIM1 can normalize the enhanced CDI caused by the T92W mutation, but it is not clear how this happens. Further, the authors propose a "push" - "pull" mechanism for the complementary roles L138 and H134 in channel regulation but do not provide any structural dynamics data to support this idea. The authors provide a mathematical explanation for chelator-specific differences in CDI observed for the T92W compared to WT Orai1 but do not show any fitted data to accompany and support the model. Finally, the authors show that a considerable portion of the CDI can be eliminated after a C-terminal Orai1 deletion (i.e. residues 267-301) and probe the idea that N-terminal W76, Y80, and R83 residues may contribute to the residual CDI effect; however, after W76E, Y80E, R83E mutations showed enhanced CDI (rather than suppressed) in the context of the T92W mutation, no further experiments were pursued to account for the residual CDI.

      Overall, the strengths far outweigh the weaknesses of this study, and the conclusions drawn based on the data are compelling. The work represents an important conceptual advancement as future studies can now steer towards identifying the STIM-independent Ca2+ sensor underlying the CDI of CRAC channels and revealing structural mechanisms by which Ca2+ sensing leads to pore closure.

    1. Reviewer #1 (Public Review):

      FLOWERING LOCUS C (FLC) is a key repressor of flowering in Arabidopsis thaliana. FLC expression creates a requirement for vernalization which is the acquisition of competence to flower after exposure to the prolonged cold of winter. Vernalization in Arabidopsis and other Brassicas results in the suppression of FLC expression.

      How exposure to winter cold initiates the vernalization process (i.e., the silencing of FLC) is not fully understood. It is known that cold exposure causes several long non-coding RNAs, including COOLAIR and COLDAIR, to be transcribed from FLC. this work shows that COOLAIR induction by cold results requires the binding of CRT/DRE-binding factors (CBFs) to their cognate recognition elements which reside at the 3' end of the FLC locus. The authors demonstrate this regulation in many ways including studying the effect on vernalization of knocking out all CBFs and also by showing that constitutive CBF expression causes COOLAIR levels to be elevated even without cold exposure. Intriguingly, plants with genetic alterations that eliminate COOLAIR expression (loss of CBF activity and FLC deletion mutants that eliminate COOLAIR expression) do not have a significant impairment in becoming vernalized.

      The work appears to be done properly and provides much important information about how this remarkable environmentally-induced epigenetic switch operates.

    2. Reviewer #2 (Public Review):

      Here the authors questioned the regulation and functional roles of anti-sense transcripts at the 3'end of an important flowering-time regulator FLC.

      The authors present compelling genetic, molecular biology, transgene, and biochemical data on the molecular details of how COOLAIR is induced by cold temperatures. They report that cold-induction of COOLAIR is mediated by C-repeat/dehydration-responsive elements (CRT/DREs) at the 3'-end of the FLC and relatively small deletions of the CRT/DREs prevent cold-induction of COOLAIR. They also report that long-term cold results in an increase in the expression of CRT/DRE BINDING FACTORs (CBFs) that bind to the CRT/DREs and result in the activation of genes containing CRT/DREs.

      Interestingly, in lines in which COOLAIR is not induced the vernalization proceeds normally with respect to flowering behavior and cold-mediated FLC chromatin changes, a result that is at odds with some publications but consistent with other reports.

      The major strength of this research is the comprehensive battery of relevant assays used to address their aim. Using ChIP they demonstrate CBF3 directly binds to the 3'end of FLC in vivo, and of less interest, but still very relevant, CBF3 binds to a CRT/DRE motif containing oligo-nucleotides in vitro using an EMSA. Using CRISPR-mediated genetic deletion of these sequences in vivo, they demonstrated that the downstream antisense transcripts are no longer transcribed. Interestingly, in these CRISPR mutants or genetic mutants of higher-order CBF mutants, the vernalisation response (chromatin modifications) is not impaired. They also show that CBF mRNA transcription occurs in at least two waves, an early peak, and over a prolonged cold period.

      While the CRISPR genetic motif mutants are relatively small, a few hundred base pairs, ideally they would have been smaller if only encompassing the CRT/DRE motif.

      The authors clearly achieved their aims and the presented results strongly support their conclusions. The compelling data clearly questions a widely held view in the vernalisation field. The presented methods can be widely transferable to a broader research community.

    3. Reviewer #3 (Public Review):

      The authors start by examining the COOLAIR promoter and identifying a CRT/DRE motif that is bound by the CBF transcription factor family that is involved in the short-term cold. This is confirmed by gel shift assays and chromatin immunoprecipitation. However, it should be noted that the gel shift assays are an in vitro assay and the chromatin immunoprecipitation is carried out with plants over-expressing CBF3-myc from the pSuper promoter and so do not necessarily reflect the native state. The authors then examine COOLAIR expression in lines over-expressing each of the three CBF proteins of Arabidopsis and found COOLAIR expression elevated in the warm in all three, but with small differences in the variants of COOLAIR that are expressed. Examination of the expression of COOLAIR after short-term cold shows that transcript abundance increases after 6 hours, this expression was not observed in the cbfs-1 where all three CBFs are knocked out. Taken together this provides good evidence that COOLAIR transcription is rapidly induced via CBFs on exposure to cold.

      The authors then go on to look at the roles of CBFs in longer-term cold. COOLAIR has previously been shown to increase during long-term cold (multiple weeks duration), so the question was whether this increase is CBF-dependent. The increase in COOLAIR abundance is similar to other CBF targets but does begin to decline with 40-day cold periods, presumably reflecting the shutdown of the FLC locus. The lack of COOLAIR expression in the cbfs-1 mutant is good evidence that increased COOLAIR expression is CBF-dependent. The authors also present evidence that CBFs are required for COOLAIR induction by the first seasonal frost, which is consistent with this being a short-term cold response.

      The authors then examine deletions of the COOLAIR promoter. In agreement with the hypothesis that CBFs regulate COOLAIR transcription via the CRT/DREs in the COOLAIR promoter, deletions that include the two elements do not show cold induction of COOLAIR, while one that contains them does. It should be noted that these deletions are relatively coarse so could include other elements than the CRT/DREs.

      The authors then use the finding that COOLAIR is not induced in the cbfs-1 mutant or in the deltaCOOLAIR1 and 3 lines to ask whether COOLAIR is required for the repression of FLC in the vernalization response. The data in Figures 6 and 7 show that these lines don't show different responses to vernalization treatment at the FLC expression, FLC chromatin modifications, or flowering time/leaf number to flowering. This supports the conclusion that the COOLAIR transcript does not play an essential role in the vernalization response.

      The Discussion is well-balanced and considers previous publications in this area and highlights differences with this study. The conservation of COOLAIR in other brassica species suggests that it does have a biological function, but the data here suggest it isn't an essential component of the vernalization response. Whether there is a function in more natural conditions where the temperature fluctuates in a diurnal manner during the vernalization period is a possibility that is considered. When the data presented here are taken with other publications, the precise biological role of COOLAIR remains enigmatic.

    1. Joint Public Review:

      In this study, the authors transcriptomically characterize TIL from head and neck cancers and associate their transcriptional programs with overall survival as a function of HPV positivity. Specifically, they study the impact of CDK4 inhibition on TIL from these tumors. They find an exhausted T cell subset that preferentially expresses CDK4. They then perform some in vitro studies to test the function of exhausted T cells and the impact of CDK4 inhibition on different TIL subsets from head and neck tumors. Understanding the functional impact of different cancer therapies on cells in the TME is of high importance and interest to the field.

      1. Line 215: The authors state that pairing TCRseq with RNAseq reflects the magnitude of TCR signaling. This is absolutely not the case. TCR sequencing does not reflect TCR signaling strength.<br /> 2. A lot of discussion around "activation" is presented, but there is no evidence to support which genes or gene programs are associated with "activation".<br /> 3. Line 249: It is unclear why the authors are indicating that TCRseq was used in pseudotime analysis. This type of analysis does not take TCRs into account but rather looks at the proportion of spliced mRNA of individual genes from the DGE data.<br /> 4. There is no way to know if the differences in proliferation and cell viability shown in Figs. 4a and b, respectively, are meaningful or not. Proper controls or replicates should be provided to fully understand if this difference is biologically meaningful. Likewise, what is the evidence that P-Tex cells are self-renewing rather than expanding?

    1. ‘Running on Emptiness – The Pathology of Civilisation’John Zerzan (2002) All religions have problems with ‘unbelievers’, but that response is insignificant compared to their visceral hatred of ‘apostates’.

      !- Book Review : Free Range Activist !- Title : ‘Running on Emptiness – The Pathology of Civilisation’ !- Author : John Zerzan (2002) !- Website : http://www.fraw.org.uk/blog/reviews/023/index.shtml

      • All religions have problems with ‘unbelievers’, but that response is insignificant compared to their visceral hatred of ‘apostates’.
    1. Reviewer #1 (Public Review):

      This study reports the results of a computational and EEG analysis of altruistic decision making. The authors intend to examine whether fundamentally different mechanisms operate to drive altruistic decision making in different contexts, which they here manipulate by examining choices in the realm of advantageous and disadvantageous inequality. The authors find that changes in self payoff are encoded in opposite manners in the two contexts, but that a similar evidence accumulation mechanism leading up to the time of response seems to operate equally in both. In addition, they find that individual differences in generosity are predicted more by differences in sensitivity to change in the other's payoff in the disadvantageous inequality condition, and by stronger phase coupling between sensors related to this delta-other signal and sensors related to the evidence accumulation signal.

      This study makes a valuable contribution by combining a sophisticated suite of modelling and neurophysiological analyses to shed light on the decision parameter adjustments that inform altruistic decisions in different contexts. The conclusions regarding those adjustments appear well supported by the data. One aspect that could be clarified is that there is an apparent discrepancy between the cross-condition bound adjustments identified by the modelling and the absence of any corresponding neural evidence accumulation signal amplitude difference.

      One of the stated overarching goals of this study is to determine whether the neural mechanisms and circuits for altruistic decisions are context-specific or general. The manuscript would benefit from greater clarity on this point, in particular defining what is meant by 'mechanisms' and what qualitative and quantitative criteria should be applied when identifying them as distinct versus common. As all decisions in this study are reported via the same manual actions it seems implausible that there would be no overlap at all in the circuits and mechanisms involved. In addition, the prior literature has demonstrated that even individual neurons can trace different computations depending on the circumstances. Therefore, it is necessary to clarify whether the authors are searching for context-dependence in the brain areas/signals that are recruited and/or in the computations that are performed within a brain area.

    2. Reviewer #2 (Public Review):

      Recent work in the neurosciences has suggested that decision making in most domains consists of computations at multiple stages. In value-based choices, initial evidence is perceived, categorized, and evaluated, then accumulated over time in a process that essentially compares the relative value of two options, until the accumulated evidence passes a threshold for choice. Although previous work has shown that this basic structure also applies to decisions in the domain of prosocial choice, it has remained unclear at what stage of this decision process variation in prosocial choices arises. The authors aimed to resolve this issue by using a combination of computational modeling and EEG, applied to a choice paradigm that evokes variation in altruistic behavior through two distinct routes: exogenous variation in the inequality context of a choice (i.e., advantageous vs. disadvantageous inequality), and endogenous variation as a function of individual differences in prosocial preferences.

      One of the strengths of this approach (particularly the use of EEG) over previous studies is that the authors can use the timing and nature of the EEG signals to disentangle both HOW preferences evolve, and WHEN differences evoked by context or individual preference emerge. This work very clearly shows that late-stage choice comparison processes, locked to the time of response (i.e., the evidence accumulation phase of a choice) are likely NOT where variations in altruistic choice arise. Instead, the evidence points to a set of distinct signals that occur time-locked to the onset of an option that enables participants to make a choice, which implies that the computations driving choice behavior likely occur at the perceptual and/or valuation stage. This is not wholly surprising, but is interesting and important to verify.

      A second potential strength of this approach is that the methodology allows the authors to determine whether the observed signals more strongly resemble encoding of the overall magnitude of outcomes to self and others, or instead are more related to signals sensitive to distributional values (i.e., inequality/fairness). The evidence here paints a quite intriguing, but somewhat mixed picture, in my view, and I think needs to come with more caveats than the authors currently acknowledge. The authors claim that their evidence supports the idea that people are making choices by considering inequality, rather than by computing outcomes for self or other directly. The lack of a consistently-signed association between EEG signals and either self or other outcome magnitude across contexts is not consistent with the idea that values are encoded in terms of self and other, which has sometimes been argued from fMRI data. However, I also do not think they are fully consistent with the authors' claims that they are observing signals related directly to fairness considerations either. Fairness/inequality, as typically defined by economic models of social preferences, involves computing the differences between self and other payoffs. The authors find ERP signals scaling with payoff changes for self but not other. Those signals do move in opposite directions in the two inequality contexts, which is why the authors interpret this as meaning that these ERP signals represent some calculation related to fairness. But there is no sensitivity of these signals to payoff change for the other, suggesting that these signals are not precisely driven by fairness as it is canonically conceived. Instead, it seems that these signals might reflect something about how people orient to self outcomes differently in the two contexts. This actually is an intriguing finding, but is somewhat difficult to interpret, since it is not wholly clear what these ERP signals represent (i.e., are they related to perception, valuation, attention, etc.?). Moreover, as the authors acknowledge in their discussion, the design of the study, with its presentation of a first option that determines the inequality context and a second option that determines the relative values of the options, means that it is difficult to know when and how one would expect to see raw self and other values as opposed to comparative value signals related to differences in self and other. Finally, the sensitivity (or lack thereof) of EEG to more subcortical signals means that it is not clear one is getting a whole picture of the computations driving choice. Thus, I think the conclusion that behavior is related to inequality processing rather than to a focus on self- and other-payoffs directly, while intriguing, needs to be tempered a bit.

      What also seems somewhat puzzling is that the behavioral and neural signals do not always seem fully consistent with one another, or with prior research. For example, behaviorally people seem to put more weight on others' payoff changes in the advantageous inequality context. And in other work (Morishima et al., 2012), it is behavioral variation in the advantageous context that correlates with neural (anatomical) variation. Yet here, there are no EEG signals that encode changes in other outcomes as a main effect in the advantageous context, and it is individual variation in encoding of others' payoffs in the DISADVANTAGEOUS context that relate to individual differences in equality-seeking in that context. Thus, it is actually in the context where one would expect people to be paying *less* attention to other outcomes (based on the modeling parameters) that neural signals seem to be *more* sensitive to those outcomes. This doesn't mean that these signals aren't interesting, but it does point to a need to more fully understand what they represent before coming to firm conclusions about what they actually mean, computationally and psychologically.

      Thus, I think this paper will likely have an impact on the field largely for the intriguing questions it raises about how people make altruistic choices rather than for providing definitive answers. This is an important contribution and researchers will, I expect, find this paper thought-provoking.

    1. Reviewer #1 (Public Review):

      In this study, the authors examine the function of Tomosyn, in dense core vesicle fusion using CRE-mediated deletion in neuronal cultures from mice expressing conditional alleles of tomosyn and tomosyn-2. Tomosyn is a large soluble SNARE protein, where earlier work in multiple species suggested that it functions as a competitive inhibitor of cognate SNARE interactions impairing fusion. The authors show that while loss of tomosyns did not affect dense core vesicle exocytosis, it reduced the expression of several key dense core cargos, including BDNF. Limited (if anything opposite) impact of tomosyn loss-of-function on intracellular vesicle trafficking or Golgi function.

      The authors concluded that tomosyns regulate neuropeptide and neurotrophin secretion by regulating dense core vesicle cargo production but not exocytosis.

    2. Reviewer #2 (Public Review):

      The authors provide here a very careful and thorough analysis of the effects of tomosyn elimination in neurons, in relation to dense-core vesicles. They find strong effects on vesicle generation (size, protein composition), but not on vesicle exocytosis, in spite of tomosyn's known interaction with the exocytosis SNAREs.

    3. Reviewer #3 (Public Review):

      Based on studies over the last two decades, tomosyns participate in processes as diverse as synaptic SNARE complex stability (Yu H et al., 2014), dendritic spine density (Saldate JJ et al., 2018), mossy fiber synaptic plasticity (Ben-Simon Y et al., 2015), inhibition of mast cell degranulation (Madera-Salcedo IK et al., 2018), insulin-stimulated GLUT4 exocytosis by adipocytes (Wang S, et al., 2020), and both basal and stimulated secretion by PC12 cells (Williams et al., 2011). In yeast, which lacks storage granules, two tomosyn orthologs control the formation of post-Golgi vesicles. The actions of tomosyn are cell-type specific and subject to regulation by phosphorylation and the ubiquitin-proteasome system (Saldate JJ et al., 2018; Williams et al., 2011; Madera-Salcedo IK et al., 2018). In beta-cells, the ability of tomosyn to decrease insulin secretion by binding syntaxin1A requires its SUMOylation (Ferdaoussi M, et al., 2017). The carefully designed and validated mouse line developed by the authors will facilitate detailed, mechanistic studies of the diverse, cell-type specific actions of tomosyns.

      Using cultures derived from the hippocampi of this new mouse strain, multiple differences were observed between two-week-old WT and DKO (double knockout of tomosyn-1 and -2) cultures. Analysis of dense core vesicle release by single neurons revealed no change in their exocytosis, but identified a decrease in levels of the dense core vesicle reporter, leading to the discovery of a decrease in levels of two endogenous dense core vesicle proteins, BDNF and IA-2. In contrast, levels of two lysosomal/endocytic markers were unaltered, demonstrating granule specificity.

      WT and DKO cultures were compared using mass spectrometry. Significant changes in the levels of 3% of the proteins were identified. Strikingly, levels of several additional dense core vesicle proteins were decreased in DKO cultures. In contrast, levels of multiple mitochondrial proteins were greatly increased in DKO cultures. In addition, significant increases in VGLUT2 (a marker of glutamatergic neurons) and in GAD67, GAT1, and GAT3 (GABAergic markers) confirmed the presence of widespread differences in hippocampal cultures that matured in the absence of tomosyns. Focusing on BDNF and other dense core vesicle proteins, qPCR studies revealed decreases in mRNA levels for a subset of dense core vesicle proteins.

      The use of multiple culture systems allowed the authors to employ different approaches, ranging from monitoring the release of single granules expressing a dense core vesicle reporter to quantifying the accelerated trafficking of a tagged cargo protein from the ER through the TGN and into DCVs in the absence of tomosyns. While no changes in synaptic complex formation were observed, both electron microscopy and analysis of single vesicles expressing a dense core vesicle reporter revealed a decrease in granule diameter.

      Weaknesses of methods and results. Within 8 h of plating, hippocampal cultures prepared from a single litter were transduced with a lentivirus encoding active or inactive mCherry-tagged Cre-recombinase, generating WT and DKO cultures; expression of Cre-recombinase was limited to neurons using the synapsin promoter. Cultures were generally examined after two weeks. Culture conditions were varied to allow comparison of dense core vesicle exocytosis by single neurons (a neuron on a glial microisland) or protein and mRNA levels in dense neuronal networks plated on coated plastic without a glial feeder layer in WT vs. DKO cultures. Whether cultures allowed to develop under these vastly different conditions respond to the absence of tomosyns in a different manner is unknown. No attempt was made to rescue any of the differences observed by expressing tomosyn in DKO neurons. Successful rescue experiments would alleviate concerns about the effects of developmental differences on the phenotypes observed.

      Immunocytochemical studies revealed an approximately two-fold drop in BDNF protein levels in the soma and neurites of DKO neurons. In contrast, BDNF, which was detectable in WT cultures using mass spectrometry, was not detectable using mass spectrometry to analyze DKO cultures. No explanation for this discrepancy between immunocytochemistry and mass spectrometry is offered. Despite the fact that neither BDNF secretion nor BDNF degradation was assessed, the authors state in their Abstract that "tomosyns regulate neuropeptide and neurotrophin secretion via control of DCV cargo production".

      The authors do not adequately refer to the rich literature discussing the many secretory pathways used by different cell types, referring only to synaptic vesicles and dense core vesicles. Golgi by-pass pathways are known to take membrane proteins to dendrites and tomosyns are known to play a role in the trafficking of GLUT4 from endocytic compartments to the plasma membrane. Soluble cargo proteins such as BDNF are released both constitutively and in response to stimuli. Cargo proteins (proinsulin, proANP, and growth hormone, for example) can drive the appearance of dense core vesicles.

      The mass spectrometry data presented in Fig. 3 are not well incorporated into the Discussion. KIF6, which plays a role in retrograde Golgi to ER traffic, is detectable in DKO cultures, but not in WT cultures and could contribute to the accelerated trafficking phenotype observed using RUSH. Coordinate control of the expression of dense core vesicle genes has been studied in a variety of systems, ranging from mammals to C. elegans to D. melanogaster. Levels of these gene products could have been assessed using existing mass spectrometric data or by additional qPCR studies. The diminished levels of dense core vesicle reporters observed in Fig.1 remain unexplained. Intracellular degradation and increased basal secretion, neither of which was assessed, could contribute to this observation.<br /> The authors did not take advantage of the structure/function studies used to dissect the roles of the beta-propeller and SNARE-domains of tomosyns. In yeast, loss of SR07/SR077, tomosyn orthologs which lack a SNARE-like domain, causes a defect in the exocytosis of post-Golgi vesicles and the accumulation of secretory vesicles with altered composition (Forsmark et al., 2011).

      Are claims and conclusions justified by data: The title of the manuscript, "SNARE protein tomosyn regulates dense core vesicle composition but not exocytosis in mammalian neurons" is misleading. The authors present no evidence that the SNARE-domain of tomosyn is necessary for its effects on dense core vesicle composition. The yeast orthologs of tomosyn, which lack a SNARE domain, affect post-Golgi vesicular trafficking via their beta-propeller domains. Hippocampal neurons are not representative of all "mammalian" neurons. In rat sympathetic neurons, tomosyn depletion results in a decrease in neurotransmitter release. A key conclusion is that tomosyns regulate neuropeptide and neurotrophin secretion by controlling cargo production, not cargo release - this conclusion is not supported by the data presented.

      Likely impact of work on the field: The mouse line developed for these studies will be of great use in mechanistic studies of the multiple roles of tomosyns. The authors identified a range of parameters that are altered in hippocampal neurons which develop in the absence of tomosyns. Additional mechanistic studies are needed to directly assess the manner in which the absence of tomosyns contributes to these changes.

    1. Reviewer #1 (Public Review):

      These findings for the first time provide a comprehensive multiscale assessment of the arrhythmogenic potential of elite exercise training.

      The authors trained canines using a treadmill over 16 weeks, and compared these animals (n=12) to sedentary animals (n=13). The authors found global evidence of electrophysiologic remodeling ECG indices and heart rate, as well as repolarization variability in trained animals relative to controls.

      The authors also demonstrate a range of effects of ventricular cardiomyocyte ion channels and fibrosis. Finally, using an induction protocol, the authors show enhanced risk for ventricular fibrillation as well as spontaneous arrhythmias in trained dogs.

      The authors conclude that structural and electrophysiologic remodeling of ventricles in elite trained athletes is associated with ventricular arrhythmogenesis.

      First, this is a difficult study to achieve given the logistical challenges of managing a large animal set up as utilized in this study. Further protocols that involve in vivo and subsequently in vitro studies of tissues from large animals are challenging to accomplish. Finally, the multimodal assessments undertaken in this study to achieve these comprehensive objectives are an additional strength.

      Weaknesses include the descriptive nature of the work and somewhat low level of rigor in presenting the observed data. The presentation of the data in the text could also be improved. Finally, some of the counterintuitive/conflicting findings e.g. enhanced HCN4 expression with reduced heart rate.

    2. Reviewer #2 (Public Review):

      In this manuscript, Polyák et al. report detailed and systematic functional, electrocardiographic, electrophysiologic (both in vivo and in vitro experiments) and histological analysis in a large animal (canine) model of exercise to assess risk of ventricular arrhythmia susceptibility. They find that exercise-trained dogs have a slower heart rate (not accounted by heightened vagal tone alone and consistent with recent work from Denmark), an increased ventricular mass and fibrosis, APD lengthening due to repolarisation abnormality, enhanced HCN4 expression and decreased outward potassium channel density together with increased ventricular ectopic beats and ventricular fibrillation susceptibility (open-chest burst pacing). The authors suggest these changes as underlying the risk of VA in athletes, and appropriately caution against consigning the beneficial effects of exercise. In general, this study is well done, reasonably well-written, with reasonable conclusions, supported by the data presented and is much needed. There are some methodological, however, given the paucity of experimental data in this area, I think it would still be additive to the literature.

      Strengths<br /> 1. This is an area with very limited experimental data- this is an area of need.<br /> 2. The study, in general seems to be well-conducted with two clear groups<br /> 3. The use of a large animal model is appropriate<br /> 4. The study findings, in general, support the authors conclusions<br /> 5. The authors have shown some restraint in their conclusions and the limitations section is detailed and well written.

      Weaknesses<br /> 1. There are some methodological issues:<br /> a. Authors should explain what the conditioning protocol was and why it was necessary.<br /> b. The rationale for the exercise parameters chosen needs to be presented.<br /> c. Open chest VF induction was a limitation, and it was unnecessary.<br /> d. A more refined VT/VF induction protocol was required. This is a major limitation to this work.<br /> e. The concept of RV dysfunction has not been considered in the study and its analysis.<br /> f. The lack of a quantitative measure for fibrosis is a limitation.<br /> 2. Statistical analysis requires further detail (checking of normality of the data/appropriate statistical test).<br /> 3. The use of Volders et al. study as a corollary in the discussion does not seem justified given that this study used AV block induced changes as an acquired TdP model.

    3. Reviewer #3 (Public Review):

      This is a well-designed and well conducted study on the effect of 4 months sustained exercise on atrioventricular function and cardiac remodeling in a clinically relevant large animal (canine) model. All methods are well described with proper controls. The findings support the conclusion. Potential limitations are the study are clearly stated. The findings advance the field and provide clear evidence for the susceptibility of ventricular arrhythmia in the canine model of endurance training.

    4. Reviewer #4 (Public Review):

      In the manuscript the author tried to find the cellular level mechanism that causes sudden cardiac death in elite athletes. They found that there are more ventricular fibrosis, ventricular extrasystole burden, longer action potential duration, higher ventricular fibrillation (VF) inducibility, higher HCN4 expression and decreased Ito in sustained trained dog model.

      The author successfully conducted large animal training model, showed bradycardia and ventricular fibrosis as a finding similar in athletes and demonstrated the increased ventricular arrhythmia susceptibility to electrical stimulation. The finding of increased action potential duration can be postulated to be a factor of sudden cardiac death in these athletes. However, the interpretation of these findings should be cautious just like all the animal studies. Human has a more complex interaction with the environment and individual variabilities. Will the higher susceptibility of VF to electrical stimulation be the same in athletes is still hard to answer.

      Still, it is the first study to provide a large animal model of sustained training mimicking trained athletes and to give insights into the cellular level of change in an athlete's heart. The young death of this special group is a tragedy and the importance of these studies cannot be overemphasized.

    1. Reviewer #1 (Public Review):

      This article is aimed at constructing a recurrent network model of the population dynamics observed in the monkey primary motor cortex before and during reaching. The authors approach the problem from a representational viewpoint, by (i) focusing on a simple center-out reaching task where each reach is predominantly characterised by its direction, and (ii) using the machinery of continuous attractor models to construct network dynamics capable of holding stable representations of that angle. Importantly, M1 activity in this task exhibits a number of peculiarities that have pushed the authors to develop important methodological innovations which, to me, give the paper most of its appeal. In particular, M1 neurons have dramatically different tuning to reach direction in the movement preparation and execution epochs, and that fact motivated the introduction of a continuous attractor model incorporating (i) two distinct maps of direction selectivity and (ii) distinct degrees of participation of each neuron in each map. I anticipate that such models will become highly relevant as neuroscientists increasingly appreciate the highly heterogeneous, and stable-yet-non-stationary nature of neural representations in the sensory and cognitive domains.

      As far as modelling M1 is concerned, however, the paper could be considerably strengthened by a more thorough comparison between the proposed attractor model and the (few) other existing models of M1 (even if these comparisons are not favourable they will be informative nonetheless). For example, the model of Kao et al (2021) seems to capture all that the present model captures (orthogonality between preparatory and movement-related subspaces, rotational dynamics, tuned thalamic inputs mostly during preparation) but also does well at matching the temporal structure of single-neuron and population responses (shown e.g. through canonical correlation analysis). In particular, it is not clear to me how the symmetric structure of connectivity within each map would enable the production of temporally rich responses as observed in M1. If it doesn't, the model remains interesting, as feedforward connectivity between more than two maps (reflecting the encoding of many more kinematic variables) or other mechanisms (such as proprioceptive feedback) could well explain away the observed temporal complexity of neural responses. Investigating such alternative explanations would of course be beyond the scope of this paper, but it is arguably important for the readers to know where the model stands in the current literature.

      Below is a summary of my view on the main strengths and weaknesses of the paper:

      1. From a theoretical perspective, this is a great paper that makes an interesting use of the multi-map attractor model of Romani & Tsodyks (2010), motivated by the change in angular tuning configuration from the preparatory epoch to the movement execution epoch. Continuous attractor models of angular tuning are often criticised for being implausibly homogeneous/symmetrical; here, the authors address this limitation by incorporating an extra dimension to each map, namely the degree of participation of each neuron (the distribution of which is directly extracted from data). This extension of the classical ring model seems long overdue! Another nice thing is the direct use of data for constraining the model's coupling parameters; specifically, the authors adjust the model's parameters in such a way as to match the temporal evolution of a number of "order parameters" that are explicitly manifested (i.e. observable) in the population recordings.

      I believe the main weakness of this continuous attractor approach is that it - perhaps unduly - binarises the configuration of angular tuning. Specifically, it assumes that while angular tuning switches at movement onset, it is otherwise constant within each epoch (preparation and execution). I commend the authors for carefully motivating this in Figure 2 (2e in particular), by showing that the circular variance of the distribution of preferred directions is higher across prep & move than within either prep or move. While this justifies a binary "two-map model" to first order, the analysis nevertheless shows that preferred directions do change, especially within the preparatory epoch. Perhaps the authors could do some bootstrapping to assess whether the observed dispersion of PDs within sub-periods of the delay epoch is within the noise floor imposed by the finite number of trials used to estimate tuning curves. If it is, then this considerably strengthens the model; otherwise, the authors should say that the binarisation reflects an approximation made for analytical tractability, and discuss any important implications.

      2. While it is great to constrain the model parameters using the data, there is a glaring "issue" here which I believe is both a weakness and a strength of the approach. The model has a lot of freedom in the external inputs, which leads to relatively severe parameter degeneracies. The authors are entirely forthright about this: they even dedicate a whole section to explaining that depending on the way the cost function is set up, the fit can land the model in very different regimes, yielding very different conclusions. The problem is that I eventually could not decide what to make of the paper's main results about the inferred external inputs, and indeed what to make of the main claim of the abstract. It would be great if the authors could discuss these issues more thoroughly than they currently do, and in particular, argue more strongly about the reasons that might lead one to favour the solutions of Fig 6d/g over that of Fig 6a. On the other hand, I see the proposed model as an interesting playground that will probably enable a more thorough investigation of input degeneracies in RNN models. Several research groups are currently grappling with this; in particular, the authors of LFADS (Pandarinath et al, 2018) and other follow-up approaches (e.g. Schimel et al, 2022) make a big deal of being able to use data to simultaneously learn the dynamics of a neural circuit and infer any external inputs that drive those dynamics, but everyone knows that this is a generally ill-posed problem (see also discussion in Malonis et al 2021, which the authors cite). As far as I know, it is not yet clear what form of regularisation/prior might best improve identifiability. While Bachschmid-Romano et al. do not go very far in dissecting this problem, the model they propose is low-dimensional and more amenable to analytical calculations, such that it provided a valuable playground for future work on this topic.

      3. As an addition to the motor control literature, this paper's main strengths lie in the model capturing orthogonality between preparatory and movement-related activity subspaces (Elsayed et al 2016), which few models do. However, one might argue that the model is in fact half hand-crafted for this purpose, and half-tuned to neural data, in such a way that it is almost bound to exhibit the phenomenon. Thus, some form of broader model cross-validation would be nice: what else does the model capture about the data that did not explicitly inspire/determine its construction? As a starting point, I would suggest that the authors apply the type of CCA-based analysis originally performed by Sussillo et al (2015), and compare qualitatively to both Sussillo et al. (2015) and Kao et al (2021). Also, as every recorded monkey M1 neuron can be characterized by its coordinates in the 4-dimensional space of angular tuning, it should be straightforward to identify the closest model neuron; it would be very compelling to show side-by-side comparisons of single-neuron response timecourses in model and monkey (i.e., extend the comparison of Fig S6 to the temporal domain).

      4. The paper's clarity could be improved.

    2. Reviewer #2 (Public Review):

      The authors study M1 cortical recordings in two non-human primates performing straight delayed center-out reaches to one of 8 peripheral targets. They build a model for the data with the goal of investigating the interplay of inferred external inputs and recurrent synaptic connectivity and their contributions to the encoding of preferred movement direction during movement preparation and execution epochs. The model assumes neurons encode movement direction via a cosine tuning that can be different during preparation and execution epochs. As a result, each type of neuron in the model is described with four main properties: their preferred direction in the cosine tuning during preparation (denoted by θ_A) and execution (denoted by θ_B) epochs, and the strength of their encoding of the movement direction during the preparation (denoted by η_A) and execution (denoted by η_B) epochs. The authors assume that a recurrent network that can have different inputs during the preparation and execution epochs has generated the activity in the neurons. In the model, these inputs can both be internal to the network or external. The authors fit the model to real data by optimizing a loss that combines, via a hyperparameter α, the reconstruction of the cosine tunings with a cost to discourage/encourage the use of external inputs to explain the data. They study the solutions that would be obtained for various values of α. The authors conclude that during the preparatory epoch, external inputs seem to be more important for reproducing the neuron's cosine tunings to movement directions, whereas during movement execution external inputs seem to be untuned to movement direction, with the movement direction rather being encoded in the direction-specific recurrent connections in the network.


      1) Fundamentally, without actually simultaneously recording the activity of upstream regions, it should not be possible to rule out that the seemingly recurrent connections in the M1 activity are actually due to external inputs to M1. I think it should be acknowledged in the discussion that inferred external inputs here are dependent on assumptions of the model and provide hypotheses to be validated in future experiments that actually record from upstream regions. To convey with an example why I think it is critical to simultaneously record from upstream regions to confirm these conclusions, consider two alternative scenarios: I) The recorded neurons in M1 have some recurrent connections that generate a pattern of activity that is based on the modeling seems to be recurrent. II) The exact same activity has been recorded from the same M1 neurons, but these neurons have absolutely no recurrent connections themselves, and are rather activated via purely feed-forward connections from some upstream region; that upstream region has recurrent connections and is generating the recurrent-like activity that is later echoed in M1. These two scenarios can produce the exact same M1 data, so they should not be distinguishable purely based on the M1 data. To distinguish them, one would need to simultaneously record from upstream regions to see if the same recurrent-like patterns that are seen in M1 were already generated in an upstream region or not. I think acknowledging this major limitation and discussing the need to eventually confirm the conclusions of this modeling study with actual simultaneous recordings from upstream regions is critical.

      2) The ring network model used in this work implicitly relies on the assumption that cosine tuning models are good representations of the recorded M1 neuronal activity. However, this assumption is not quantitatively validated in the data. Given that all conclusions depend on this, it would be important to provide some goodness of fit measure for the cosine tuning models to quantify how well the neurons' directional preferences are explained by cosine tunings. For example, reporting a histogram of the cosine tuning fit error over all neurons in Fig 2 would be helpful (currently example fits are shown only for a few neurons in Fig. 2 (a), (b), and Figure S6(b)). This would help quantitatively justify the modeling choice.

      3) The authors explain that the two-cylinder model that they use has "distinct but correlated" maps A and B during the preparation and movement. This is hard to see in the formulation. It would be helpful if the authors could expand in the Results on what they mean by "correlation" between the maps and which part of the model enforces the correlation.

      4) The authors note that a key innovation in the model formulation here is the addition of participation strengths parameters (η_A, η_B) to prior two-cylinder models to represent the degree of neuron's participation in the encoding of the circular variable in either map. The authors state that this is critical for explaining the cosine tunings well: "We have discussed how the presence of this dimension is key to having tuning curves whose shape resembles the one computed from data, and decreases the level of orthogonality between the subspaces dedicated to the preparatory and movement-related activity". However, I am not sure where this is discussed. To me, it seems like to show that an additional parameter is necessary to explain the data well, one would need to compare fit to data between the model with that parameter and a model without that parameter. I don't think such a comparison was provided in the paper. It is important to show such a comparison to quantitatively show the benefit of the novel element of the model.

      5) The model parameters are fitted by minimizing a total cost that is a weighted average of two costs as E_tot = α E_rec + E_ext, with the hyperparameter α determining how the two costs are combined. The selection of α is key in determining how much the model relies on external inputs to explain the cosine tunings in the data. As such, the conclusions of the paper rely on a clear justification of the selection of α and a clear discussion of its effect. Otherwise, all conclusions can be arbitrary confounds of this selection and thus unreliable. Most importantly, I think there should be a quantitative fit to data measure that is reported for different scenarios to allow comparison between them (also see comment 2). For example, when arguing that α should be "chosen so that the two terms have equal magnitude after minimization", this would be convincing if somehow that selection results in a better fit to the neural data compared with other values of α. If all such selections of α have a similar fit to neural data, then how can the authors argue that some are more appropriate than others? This is critical since small changes in alpha can lead to completely different conclusions (Fig. 6, see my next two comments).

      6) The authors seem to select alpha based on the following: "The hyperparameter α was chosen so that the two terms have equal magnitude after minimization (see Fig. S4 for details)". Why is this the appropriate choice? The authors explain that this will lead to the behavior of the model being close to the "bifurcation surface". But why is that the appropriate choice? Does it result in a better fit to neural data compared with other choices of α? It is critical to clarify and justify as again all conclusions hinge on this choice.

      7) Fig 6 shows example solutions for 2 close values of α, and how even slight changes in the selection of α can change the conclusions. In Fig. 6 (d-e-f), α is chosen as the default approach such that the two terms E_rec and E_ext have equal magnitude. Here, as the authors note, during movement execution tuned external inputs are zero. In contrast, in Fig. 6 (g-h-i), α is chosen so that the E_rec term has a "slightly larger weight" than the E_ext term so that there is less penalty for using large external inputs. This leads to a different conclusion whereby "a small input tuned to θ_B is present during movement execution". Is one value of α a better fit to neural data? Otherwise, how do the authors justify key conclusions such as the following, which seems to be based on the first choice of α shown in Fig. 6 (d-e-f): "...observed patterns of covariance are shaped by external inputs that are tuned to neurons' preferred directions during movement preparation, and they are dominated by strong direction-specific recurrent connectivity during movement execution".

      8) It would be informative to see the extreme case of very large and very small α. For example, if α is very large such that external inputs are practically not penalized, would the model rely purely on external inputs (rather than recurrent inputs) to explain the tuning curves? This would be an example of the hypothetical scenario mentioned in my first comment. Would this result in a worse fit to neural data?

      9) The authors argue in the discussion that "the addition of an external input strength minimization constraint breaks the degeneracy of the space of solutions, leading to a solution where synaptic couplings depend on the tuning properties of the pre- and post-synaptic neurons, in such a way that in the absence of a tuned input, neural activity is localized in map B". In other words, the use of the E_ext term, apparently reduces "degeneracy" of the solution. This was not clear to me and I'm not sure where it is explained. This is also related to α because if alpha goes toward very large values, it would be like the E_ext term is removed, so it seems like the authors are saying that the solution becomes degenerate if alpha grows very large. This should be clarified.

      10) How do the authors justify setting Φ_A = Φ_B in equation (5)? In other words, how is the last assumption in the following sentence justified: "To model the data, we assumed that the neurons are responding both to recurrent inputs and to fluctuating external inputs that can be either homogeneous or tuned to θ_A; θ_B, with a peak at constant location Φ_A = Φ_B ≡ Φ". Does this mean that the preferred direction for a given neuron is the same during preparation and movement epochs? If so, how is this consistent with the not-so-high correlation between the preferred directions of the two epochs shown in Fig. 2 c, which is reported to have a circular correlation coefficient of 0.4?

    3. Reviewer #3 (Public Review):

      In this work, Bachschmid-Romano et al. propose a novel model of the motor cortex, in which the evolution of neural activity throughout movement preparation and execution is determined by the kinematic tuning of individual neurons. Using analytic methods and numerical simulations, the authors find that their networks share some of the features found in empirical neural data (e.g., orthogonal preparatory and execution-related activity). While the possibility of a simple connectivity rule that explains large features of empirical data is intriguing and would be highly relevant to the motor control field, I found it difficult to assess this work because of the modeling choices made by the authors and how the results were presented in the context of prior studies.

      Overall, it was not clear to me why Bachschmid-Romano et al. couched their models within a cosine-tuning framework and whether their results could apply more generally to more realistic models of the motor cortex. Under cosine-tuning models (or kinematic encoding models, more generally), the role of the motor cortex is to represent movement parameters so that they can presumably be read out by downstream structures. Within such a framework, the question of how the motor cortex maintains a stable representation of movement direction throughout movement preparation and execution when the tuning properties of individual neurons change dramatically between epochs is highly relevant. However, prior work has demonstrated that kinematic encoding models provide a poor fit for empirical data. Specifically, simple encoding models (and the more elaborate extensions [e.g., Inoue, et al., 2018]) cannot explain the complexity of single-neuron responses (Churchland and Shenoy, 2007), and do not readily produce the population-level signals observed in the motor cortex (Michaels, Dann, and Scherberger, 2016) and cannot be extended to more complex movements (Russo, et al., 2018).

      In both the Introduction and Discussion, the authors heavily cite an alternative to kinematic encoding models, the dynamical systems framework. Here, the correlations between kinematics and neural activity in the motor cortex are largely epiphenomenal. The motor cortex does not 'represent' anything; its role is to generate patterns of muscle activity. While the authors explicitly acknowledge the shortcomings of encoding models ('Extension to modeling richer movements', Discussion) and claim that their proposed model can be extended to 'more realistic scenarios', they neither demonstrate that their models can produce patterns of muscle activity nor that their model generates realistic patterns of neural activity. The authors should either fully characterize the activity in their networks and make the argument that their models better provide a better fit to empirical data than alternative models or demonstrate that more realistic computations can be explained by the proposed framework.

      Major Comments<br /> 1. In the present manuscript, it is unclear whether the authors are arguing that representing movement direction is a critical computation that the motor cortex performs, and the proposed models are accurate models of the motor cortex, or if directional coding is being used as a 'proof of concept' that demonstrates how specific, population-level computations can be explained by the tuning of individual neurons.<br /> If the authors are arguing the former, then they need to demonstrate that their models generate activity similar to what is observed in the motor cortex (e.g., realistic PSTHs and population-level signals). Presently, the manuscript only shows tuning curves for six example neurons (Fig. S6) and a single jPC plane (Fig. S8). Regarding the latter, the authors should note that Michaels et al. (2016) demonstrated that representational models can produce rotations that are superficially similar to empirical data, yet are not dependent on maintaining an underlying condition structure (unlike the rotations observed in the motor cortex).<br /> If the authors are arguing the latter - and they seem to be, based on the final section of the Discussion - then they need to demonstrate that their proposed framework can be extended to what they call 'more realistic scenarios'. For example, could this framework be extended to a network that produces patterns of muscle activity?

      2. Related to the above point, the authors claim in the Abstract that their models 'recapitulate the temporal evolution of single-unit activity', yet the only evidence they present is the tuning curves of six example units. Similarly, the authors should more fully characterize the population-level signals in their networks. The inferred inputs (Fig. 6) indeed seem reasonable, yet I'm not sure how surprising this result is. Weren't the authors guaranteed to infer a large, condition-invariant input during movement and condition-specific input during preparation simply because of the shape of the order parameters estimated from the data (Fig. 6c, thin traces)?

      3. In the Abstract and Discussion (first paragraph), the authors highlight that the preparatory and execution-related spaces in the empirical data and their models are not completely orthogonal, suggesting that this near-orthogonality serves an important mechanistic purpose. However, networks have no problem transferring activity between completely orthogonal subspaces. For example, the generator model in Fig. 8 of Elsayed, et al. (2016) is constrained to use completely orthogonal preparatory and execution-related subspaces. As the authors point out in the Discussion, such a strategy only works because the motor cortex received a large input just before movement (Kaufman et al., 2016).

    1. Reviewer #1 (Public Review):

      DeRisi and colleagues used a new phage-display peptide platform, with 238,068 tiled 62-amino acid peptides covering all known P falciparum coding regions (and numerous other entities), to survey seroreactivity in 198 Ugandan children and adults from two cohorts. They find that breadth of responses to repeat-containing peptides was twofold higher in children living in the high versus moderate exposure setting, while no such differences were observed for peptides without repeats. Additionally, short motifs associated with seroreactivity were extensively shared among hundreds of antigens, with much of this driven by motifs shared with PfEMP1 antigens.

      Malaria immunity is complex, and this new platform is a potentially valuable addition to the toolkit for understanding humoral responses. The two cohorts differed in fundamental ways: 1) high versus moderate exposure to infective bites; 2) samples drawn at the time of malaria for most donors in the high zone versus ~100 days after the last malaria episode in the moderate zone. The effect of acute malaria to boost short-term cross-reactive antibodies can confound the ability to draw inferences when comparing the two cohorts, and this should be further explored to understand its role in the patterns of seroreactivity observed.

    2. Reviewer #2 (Public Review):

      This work profiles naturally acquired antibodies against Plasmodium falciparum proteins in two Ugandan cohorts, at incredibly high resolution, using a comprehensive library of overlapping peptides. These findings highlight the ubiquity and importance of intra- and inter-protein repeat elements in the humoral immune response to malaria. The authors discuss evidence that repeat elements reside in more seroreactive proteins, and that the breadth of immunity to repeat-containing antigens is associated with transmission intensity in children.

      A key strength and value added to publicly available data are the breadth of proteome coverage and unprecedented resolution from using tiling peptides. The authors point out that a known limitation of PhIP-seq is that conformational and discontinuous-linear epitopes cannot be detected with short linear peptides. In addition, disulfide linkages and post-translational modifications would be absent in the T7 representations.

      Several significant conclusions drawn from the results in this study are based on the humoral response to repeat elements that are present in multiple locations, including different genes. If antibodies to these regions are cross-reactive as described, it is not clear how the assay can differentiate antibodies that were developed against one or many of these loci. This potential confounding could change the conclusions about inter-protein motifs.

    3. Reviewer #3 (Public Review):

      This work provides a new tool, a comprehensive PhIP-seq library, containing 238,068 individual 62-amino acids peptides tiled every 25-amino acid peptide covering all known 8,980 proteins of the deadliest malaria parasite, Plasmodium falciparum, to systematically profile antibody targets in high resolution. This phage display library has been screened by plasma samples obtained from 198 Ugandan children and adults in high and moderate malaria transmission settings and 86 US controls. This work identified that repeat elements were commonly targeted by antibodies. Furthermore, extensive sharing of motifs associated with seroreactivity indicated the potential for extensive cross-reactivity among antigens in P. falciparum. This paper provides a new proteome-wide high-throughput methodology to identify antibody targets that have been investigated by protein arrays and alpha screens to date. Importantly, only this methodology (PhIP-seq library) is able to investigate repeat-containing antigens and cross-reactive epitopes in high resolution (25-amino acid resolution).

      Strengths:<br /> 1) Novel technology<br /> Firstly, the uniqueness of this study is the use of novel technology, the PhIP-seq library. This PhIP-seq library in this study contains >99.5% of the parasite proteome and is the highest coverage among existing proteome-wide tools for P. falciparum. Moreover, this library can identify antibody responses in high resolution (25 amino acids).<br /> Secondly, the PhIP-seq converts a proteomic assay (ie. protein array and alpha screen) into a genomic assay, leveraging the massive scale and low-cost nature of next-generation short-read sequencing.<br /> Thirdly, the phage display system is the ability to sequentially enrich and amplify the signal to noise.<br /> Finally, a high-quality strategic bioinformatic analysis of PhIP-seq data was applied.

      2) Novel findings<br /> The major findings of this study were obtained only by using this novel technology because of its full-proteome coverage and high resolution. Repeat elements were the common target of naturally acquired antibodies. Furthermore, extensive sharing of motifs associated with seroreactivity was observed among hundreds of parasite proteins, indicating the potential for extensive cross-reactivity among antigens in P. falciparum.

      3) Usefulness for the future research<br /> Importantly, plasma samples from longitudinal cohort studies will give the scientific community important insights into protective humoral immunity which will be important for the identification of vaccine and exposure-marker candidates in the near future.

      Weaknesses:<br /> Although the paper does have strengths in principle, the weaknesses of the paper are the insufficient description of the selected parasite proteins and seroreactivity ranking of the selected proteins such as TOP100 proteins.

    1. Reviewer #1 (Public Review):

      Osteoclasts, giant multinucleated bone-resorbing cells, are crucial regulators of bone homeostasis and pathology. An underestimated aspect of their biology is that they are very heterogeneous, with at least 2 sub-populations (inflammatory osteoclasts and tolerogenic osteoclasts) existing, and exerting different actions, especially in the context of inflammatory bone loss. In this report, Madel, Halper (co-first authors), and colleagues present an interesting report investigating this heterogeneity, and showing that the probiotic yeast S. boulardii (probably through β-glucans) may be useful in managing inflammation-mediated bone loss, including oestrogen deprivation-mediated osteoporosis, as the authors show in vivo using an OVX mouse model.

      The authors first evaluate the differences in the transcriptional landscape of tolerogenic vs inflammatory osteoclasts with RNAseq, and then they evaluate the differences in miRNA expression between the two. Finding that some of the pathways/genes that vary are related to pattern recognition receptors (PRRs), specialized in recognizing non-self antigens including those arising from bacteria and yeasts, they wonder if the probiotic yeast S. boulardii could influence the balance between tolerogenic and inflammatory osteoclasts. Indeed, when the authors treated OVX mice, characterised by an increase in inflammatory osteoclasts and estrogen deprivation/inflammation-induced bone loss, with the probiotic, the bone loss is avoided and inflammatory osteoclasts are reduced. This challenges the classical way in which osteoclast-mediated bone loss is treated, since targeting specifically the inflammatory osteoclasts could allow the good osteoclasts to keep working and improving bone health and immunity, while only the bad osteoclasts are targeted. Current treatments are not able to distinguish between the two, which can cause a paradoxical degradation in bone health and atypical fractures. The report is therefore potentially very important for the field, and although quite focused on a specific strain, it can pave the way to treating bone diseases with probiotics, or specific molecules derived from them including beta-glucans.

    2. Reviewer #2 (Public Review):

      The authors apply their previously developed concept that osteoclasts exist in at least two flavors, tolerogenic and inflammatory osteoclasts towards the treatment of osteoporosis. They suggest that selectively targeting inflammatory osteoclasts attenuates ovariectomy-induced bone loss by agonists of pattern recognition receptors (PRR) that are higher expressed on inflammatory osteoclasts. The vision would be that the tolerogenic osteoclasts are still functioning, allowing bone remodeling with high bone quality, while the strong resorbing inflammatory osteoclasts are resorbed. By expression profiling, they detected PPR differentially expressed and confirmed these by flow cytometry and RT-QPCR. The activation of the Tlr2, Dectin-1, and Mincle reduced inflammatory osteoclast generation in vitro and affected their resorptive activity. Dendritic syk cell-specific deletion abrogated the differentiation of this osteoclast subset as well. The application of yeast Saccharomyces boulardii (Sbb) into mice attenuated trabecular bone loss (but not cortical) and seemed to inhibit in vitro the generation of inflammatory osteoclasts.

      Strength:<br /> - The expression profiling between very defined in vitro generated osteoclasts, which are somehow extreme phenotypes, provides a good tool to discern gene signatures on the osteoclast level.<br /> - The candidate of PPR were evaluated in their expression at the protein level by flow cytometry and their function was evaluated by loss of function studies.<br /> - The effect of S.b. treatment is striking and exploiting such probiotic fungi could be an elegant way to treat osteoporosis.

      Weakness:<br /> - The osteoclasts are generated in vitro in the presence of M-CSF to induce tolerogenic osteoclasts or GM-CSF / Il-4 to generate inflammatory osteoclasts. The demonstration of these cell populations in the S.b. treated mice in vivo is not present, despite the challenge to do this. The author tried to tackle this, by analyzing the differentiation potential of bone marrow progenitor cells of S.b. treated animals, which provides some information.<br /> - The effect on tolerogenic osteoclasts could have been further evaluated, whether they are not affected at all, or whether there are also effects.

      The authors strikingly show that agonists for PPR are affecting strongly GM-CSF/IL-4 progenitor-derived osteoclasts. They show that t-Ocl and i-Ocl differ in their gene signature and convincingly show the differential expression of the PPR, with exception of mincle which is clearly acknowledged. The molecular mechanism of how Sb treatment acts via the receptors remains obscure since it might act via changes in the gut permeability or by components directly released by the fungus. The kinase syk could play a role, at least some data in vitro suggest this.

      Conceptionally the authors tried to utilize the previously generated knowledge by the group published 2016 and 2020 into an approach. If the use of a probiotic fungus would be beneficial indeed this could be a suitable drug with few side effects much superior to current treatments of osteoporosis.<br /> For me, an intriguing question arises from this study, in case these i-Ocl express these receptors and are thus so "vulnerable" to the agonists to decrease their activity, evt. a negative feedback to prevent overshooting reactions?

    3. Reviewer #3 (Public Review):

      The general objective of this work is the dissection of osteoclast diversity; in particular, the authors intend to identify the specific features and properties that distinguish inflammatory and steady-state (tolerogenic) osteoclasts. To this end, the authors perform a transcriptional analysis of inflammatory and tolerogenic osteoclasts and identify the pattern recognition receptors TLR2, Dectin-1, and Mincle as differentially expressed genes. Agonists of these receptors or yeast probiotics regulating the elicited mechanisms in vitro and in vivo caused a specific inhibition of the differentiation of inflammatory rather than tolerogenic osteoclasts, thus highlighting the preferential use of different differentiation pathways by the two distinct osteoclast populations.

      The project is based on the previous knowledge and know-how of the authors on this peculiar skeletal cell population. The work is well conceived; the experiments are clearly designed and exploit state-of-the-art technologies. The results confirm the heterogeneity of osteoclasts and provide new insights in this respect. The in vitro and in vivo studies suggest that osteoclast heterogeneity can be purposedly modulated; which might be useful and advisable for therapeutic purposes. Overall, the work provides hints for further implementation and future broad applications to diseases featuring pathological bone loss.

    1. Reviewer #1 (Public Review):

      This well-done platform trial identifies that ivermectin has no impact on SARS-CoV-2 viral clearance rate relative to no study drug while casirivimab lead to more rapid clearance at 5 days. The figures are simple and appealing. The study design is appropriate and the analysis is sound. The conclusions are generally well supported by the analysis. Study novelty is somewhat limited by the fact that ivermectin has already been definitively assessed and is known to lack efficacy against SARS-CoV-2. Several issues warrant addressing:

      1) Use of viral load clearance is not unique to this study and was part of multiple key trials studying paxlovid, remdesivir, molnupiravir, and monoclonal antibodies. The authors neglect to describe a substantial literature on viral load surrogate endpoints of therapeutic efficacy which exist for HIV, hepatitis B and C, Ebola, HSV-2, and CMV. For SARS-CoV-2, the story is more complicated as several drugs with proven efficacy were associated with a decrease in nasal viral loads whereas a trial of early remdesivir showed no reduction in viral load despite a 90% reduction in hospitalization. In addition, viral load kinetics have not been formally identified as a true surrogate endpoint. For maximal value, a reduction in viral load would be linked with a reduction in a hard clinical endpoint in the study (reduction in hospitalization and/or death, decreased symptom duration, etc...). This literature should be discussed and data on the secondary outcome, and reduction in hospitalization should be included to see if there is any relationship between viral load reduction and clinical outcomes.

      2) The statement that oropharyngeal swabs are much better tolerated than nasal swabs is subjective. More detail needs to be paid to the relative yield of these approaches.

      3) The stopping rules as they relate to previously modeled serial viral loads are not described in sufficient detail.

      4) The lack of blinding limits any analysis of symptomatic outcomes.

      5) It is unclear whether all 4 swabs from 2 tonsils are aggregated. Are the swabs placed in a single tube and analyzed?

      6) In supplementary Figure 7, both models do well in most circumstances but fail in the relatively common event of non-monotonic viral kinetics (multiple peaks, rebound events). Given the importance of viral rebound during paxlovid use, an exploratory secondary analysis of this outcome would be welcome.

    2. Reviewer #2 (Public Review):

      This manuscript details the analytic methods and results of one arm of the PLATCOV study, an adaptive platform designed to evaluate low-cost COVID-19 therapeutics through enrollment of a comparatively smaller number of persons with acute COVID-19, with the goal of evaluating the rate of decrease in SARS-CoV-2 clearance compared to no treatment through frequent swabbing of the oropharynx and a Bayesian linear regression model, rather than clinical outcomes or the more routinely evaluated blunt virologic outcomes employed in larger trials. Presented here, is the in vivo virologic analysis of ivermectin, with a very small sample of participants who received the casirivimab/imdevimab, a drug shown to be highly effective at preventing COVID-19 progression and improving viral clearance (during circulation of variants to which it had activity) included for comparison for model evaluation.

      The manuscript is well-written and clear. It could benefit however from adding a few clarifications on methods and results to further strengthen the discussion of the model and accurately report the results, as detailed below.

      Strengths of this study design and its report include:<br /> 1. Selection of participants with presumptive high viral loads or viral burden by antigen test, as prior studies have shown difficulty in detecting effect in those with a lower viral burden.<br /> 2. Adaptive sample size based on modeling- something that fell short in other studies based on changing actuals compared to assumptions, depending on circulating variant and "risk" of patients (comorbidities, vaccine state, etc) over time. There have been many other negative studies because the a priori outcomes assumptions were different from the study design to the time of enrollment (or during the enrollment period). This highlight of the trial should be emphasized more fully in the discussion.<br /> 3. Higher dose and longer course of ivermectin than TOGETHER trial and many other global trials: 600ug/kg/day vs 400mcg/kg/day.<br /> 4. Admission of trial participants for frequent oropharyngeal swabbing vs infrequent sampling and blunter analysis methods used in most reported clinical trials<br /> 5. Linear mixed modeling allows for heterogeneity in participants and study sites, especially taking the number of vaccine doses, variant, age, and serostatus into account- all important variables that are not considered in more basic analyses.<br /> 6. The novel outcome being the change in the rate of viral clearance, rather than time to the undetectable or unquantifiable virus, which is sensitive, despite a smaller sample size<br /> 7. Discussion highlights the importance of frequent oral sampling and use of this modeled outcome for the design of both future COVID-19 studies and other respiratory viral studies, acknowledging that there are no accepted standards for measuring virologic or symptom outcomes, and many studies have failed to demonstrate such effects despite succeeding at preventing progression to severe clinical outcomes such as hospitalization or death. This study design and analyses are highly important for the design of future studies of respiratory viral infections or possibly early-phase hepatitis virus infections.

      Weaknesses or room for improvement:

      1. The methods do not clearly describe allocation to either ivermectin or casirivimab/imdevimab or both or neither. Yes, the full protocol is included, but the platform randomization could be briefly described more clearly in the methods section.<br /> 2. The handling of unquantifiable or undetectable viruses in the models is not clear in either the manuscript or supplemental statistical analysis information. Are these values imputed, or is data censored once below the limits of quantification or detection? How does the model handle censored data, if applicable?<br /> 3. Did the study need to be unblinded prior to the first interim analysis? Could the adaptive design with the first analysis have been done with only one or a subset of statisticians unblinded prior to the decision to stop enrolling in the ivermectin arm?<br /> 4. Can the authors comment on why the interim analysis occurred prior to the enrollment of 50 persons in each of the ivermectin and comparison arms? Even though the sample sizes were close (41 and 45 persons), the trigger for interim analysis was pre-specified.<br /> 5. The reporting of percent change for the intervention arms is overstated. All credible intervals cross zero: the clearance for ivermectin is stated to be 9% slower, but the CI includes + and - %, so it should be reported as "not different." Similarly, and more importantly for casirivimab/imdevimab, it was reported to be 52% faster, although the CI is -7.0 to +115%. This is likely a real difference, but with ten participants underpowered- and this is good to discuss. Instead, please report that the estimate was faster, but that it was not statistically significant. Similarly, the clearance half-life for ivermectin is not different, rather than "slower" as reported (CI was -2 to +6.6 hours). This result was however statistically significant for casirivimab/imdevimab.<br /> 6. While the use of oropharyngeal swabs is relatively novel for a clinical trial, and they have been validated for diagnostic purposes, the results of this study should discuss external validity, especially with respect to results from other studies that mainly use nasopharyngeal or nasal swab results. For example, oropharyngeal viral loads have been variably shown to be more sensitive for the detection of infection, or conversely to have 1-log lower viral loads compared to NP swabs. Because these models look for longitudinal change within a single sampling technique, they do not impact internal validity but may impact comparisons to other studies or future study designs.<br /> 7. Caution should be used around the term "clinically significant" for viral clearance. There is not an agreed-upon rate of clinically significant clearance, nor is there a log10 threshold that is agreed to be non-transmissible despite moderately strong correlations with the ability to culture virus or with antigen results at particular thresholds.<br /> 8. Additional discussion could also clarify that certain drugs, such as remdesivir, have shown in vivo activity in the lungs of animal models and improvement in clinical outcomes in people, but without change in viral endpoints in nasopharyngeal samples (PINETREE study, Gottlieb, NEJM 2022). Therefore, this model must be interpreted as no evidence of antiviral activity in the pharyngeal compartment, rather than a complete lack of in vivo activity of agents given the limitations of accessible and feasible sampling. That said, strongly agree with the authors about the conclusion that ivermectin is also likely to lack activity in humans based on the results of this study and many other clinical studies combined.

    3. Reviewer #3 (Public Review):

      This is a well-conducted phase 2 randomized trial testing outpatient therapeutics for Covid-19. In this report of the platform trial, they test ivermectin, demonstrating no virologic effect in humans with Covid-19.

      Overall, the authors' conclusions are supported by the data.

      The major contribution is their implementation of a new model for Phase 2 trial design. Such designs would have been ideal earlier in the pandemic.

    1. Reviewer #1 (Public Review):

      In the study, Zhao et al. investigated loop conformational changes in the active site of L1 Metallo-beta-lactamase. Antibiotic resistance is on the rise and beta-lactamases are enzymes that cleave a lactam ring. Authors investigate class B3 MBLs since these could be used for designing drugs for treating antibacterial resistance. Authors find specific loops that act as gates to the shape and access to the active site of the enzymes. They study these loops via MD simulations, Markov state models, and CVAE-based deep learning to experimentally reveal how each residue affects activity as well as remodeling of the active site.

      Strengths<br /> - The authors make a good case for why MD is important for this scaffold and protein class. The study performs MD simulations coupled with Markov State Models - this coupled with CVAE to understand the different states the protein exists in shapes the state-of-the-art study. Authors are able to isolate three different states that the protein exists in and pinpoint which interactions cause a reshaping of the active site.<br /> - Furthermore, they isolate the likely states that also correspond with lower free energy indicating why these states might be more populated. This study adds to the depth of their work.

      Weaknesses<br /> - Overall, the impact of work on the currently used antibiotic classes is unclear since the total market presence of all antibiotics is discussed not the carbapenem-based antibiotics class. Statistics related to broad antibiotic class reduce the impact statement instead of improving it.<br /> - Finally in the experimental testing only a few variants at each position were tested, leading to limited learning of the impact of active site interactions.<br /> - Authors state from previous studies on TEM-1 that disruption of the salt bridge between the two loops would alter the binding site, thus reducing antibiotic resistance. The authors continue on to hypothesize that this would hold true for the structure in consideration for this paper as well. While a good hypothesis, this cannot be inferred until we see experimental evidence for the same or a sequence comparison discussing how similar TEM1 is to the L1 MBL in question.<br /> - The authors do not explain how different splits of this data in terms of splitting (80:20 vs 70:30 or others) and reducing interaction matrix lower than 22 x 22 residues can impact their results. Also, the effect of changing the distance shell (8A) for matrix generation is not described. This variation is unaccounted for and can enable authors to pressure test their method and learnings.

    2. Reviewer #2 (Public Review):

      Zhao, Shen et al. ran molecular dynamics simulations, followed by the application of Markov State Model analysis and deep machine learning dimensional reduction, to study the dynamical behavior of two loops close to the catalytic site of L1 Metallo-β-lactamase (MBL).

      The simulations are carefully executed and of sufficient length to build a representative kinetic model. Using a dimensional reduction of the loop conformational sampling based on backbone dihedral features followed by tICA embedding, the authors obtain a Markov state model that identifies the main conformational states of the loops in the absence of bound ligands and provides estimates of the timescales for the transitions between them. Next, the authors employ an alternative way to cluster the conformations of the loops, using unsupervised dimensional reduction, implemented as a convolutional VAE applied to residue distances followed by tSNE embedding. This second step gives results that are not consistent with the clustering used for the kinetic modeling (for instance, supplement 2 of figure 4 shows that the 7 macrostates obtained by the MSM analysis don't always correspond to different areas in the CVAE+tSNE embedding).

      Moreover, an inspection of the results from both analysis techniques just confirms the role of interactions that are readily observed in the available crystal structure stabilising the most populated, closed conformation of the loops. The sophisticated computational analysis does not elucidate much of the role of the loop dynamics beyond the intuitive conclusion that disruption of the key interactions keeping the loops in the closed state would affect the function. For instance, it does not clarify what is the role of the other observed metastable states.

      Finally, the authors propose and test mutations that would likely disrupt the stability of the closed state and find that they have variable effects on the ability of the enzyme to contrast the antibiotic effect of a panel of substrates. These experimental results look useful and can potentially be used to elucidate the role of the loops in the recognition and activity of the enzyme, and for the design of inhibitors. However, no additional attempt is made to clarify the experimental results based on the mechanistic model of loop dynamics: why do different mutations have different effects? why do some mutations affect all substrates, other mutations only some substrates, and for others, no substrate is affected? What is the role of the tetrameric arrangement?

    3. Reviewer #3 (Public Review):

      The authors provide a molecular dynamics (MD)-based detailed evaluation of the contribution of the two elongated loops (alpha3-beta7 and beta12-alpha5), present near each active site of the tetrameric Stenotrophomonas maltophilia class B Metallo-beta-lactamase (MBL) L1, towards the L1's lactamase activity with the premise that a better understanding of the categorical conformational states sampled by the loops would ultimately help in the design of a better lactamase inhibitor. This is to then ultimately alleviate the public health crisis arising from β-lactam antibiotic resistance. Using enhanced sampling MD, Markov state modeling (MSM), and convolutional variation autoencoder (CAVE)-based deep learning, the authors identify five key interacting residues in these two loops which contribute to the conformational states of loops.

      The major strength of the study is that the authors carry out a detailed study (e.g., enhanced sampling MD, Markov state modeling, and convolutional variation autoencoder-based deep learning) of the conformational landscape of an important enzyme as these findings would help further experimental studies (e.g., NMR dynamics) for ligand binding, better design of inhibitory ligands of an important class of enzyme. One weakness would be that MBL L1 is a good representative of the class of MBL enzymes or not needs clarification.

      The authors achieve the goal of capturing the various conformational states of the L1 enzyme loops and their computational results support the conclusion about the various loop conformations sampled during the dynamics. However, how the mutagenesis experiment supports the existence of different conformational states will likely benefit from more clarification. Further clarification on how detecting the existence of multiple conformers benefits better inhibitor design will be very beneficial.

      Since details on macromolecular motion are often neglected in macromolecular experimental studies, the detailed MD methods described here will be a very useful companion in experimental studies of proteins and their interactions.

      A discussion on how the study of one particular enzyme could benefit in understanding the molecular properties of a class of enzymes would enhance the generality of the study.