Reviewer #3 (Public Review):

1) Appropriate bioinformatics discussions and functional pathway analysis are necessary for the key differentially expressed miRNAs that have been screened out. It is boring to only discuss the differences of miRNA data.

2) In page 5, "We used logistic regression to create such a model in the low-dose setting (N=22 sample pairs). The resulting classifier was based on the expression of miR-150-5p, miR-126-5p and miR-375" , Why the three miRNAs in the low-dose radiation group were selected for modeling instead of the seven overlapping miRNAs in the high and low dose radiation group to classificate the irradiated- and non-irradiated samples ? Please explain in detail.

3) In page 5, "Therefore, the expression of miR-126-5p, miR-150-5p and miR-375 enabled efficient classification of the irradiated- and non-irradiated samples in both settings (Fig. S6C)"; In page 6, "Interestingly, a set of 3 miRNAs quantified by qPCR in all of our previous experiments clearly visually distinguished irradiated from non-irradiated samples in the human analysis (Fig. 5A)",<br> Which three of miRNAs, miR-150-5p miR-375 miR-126-5p mentioned before or miR-150-5p miR-375 miR-215-5p？Please clarify clearly.

4) In page 4, "Since miRNA-containing exosomes.......high dose irradiation", Do you think that the differently expression of serum miRNAs partly results from exosomes? Low dose irradiation is also able to change exosomal miRNA profile，why only high dose irradiation is taken into account in paper while low dose irradiation is not？

5) Are there any miRNAs that can clearly distinguish between high and low dose groups? If so, please clarify them in text.

6) In page 7,"Importantly, similarities were observed in the level of both individual miRNAs and miRNA families", What part of result Comes to this conclusion？Please explain clearly.

7) In page 7, "We found that the most common putative tissue sources for differentially expressed miRNAs were hematopoietic and endothelial cells", Which part of result shows this sentence? Please point it out.

8) Were the patients suffering from cancer or other diseases? How to ensure that the differential expression of miRNA was caused by radiation exposure rather than their own disease? Please explain.

Reviewer #1 (Public Review):

Strength: Excellent statistical methods are employed. Specimens collected from two centers are used.<br> Weakness: It is not clear what new knowledge this follow-up study bring to the audience. The critical biomarker, miR150 they propose for development of biodosimetry assay was already discovered. There are close to dozen publications showing the dose response of miR50, in mouse, rats, non-human primates and humans (including two research papers and and several reviews from authors). The dose response shown in 4b is not appreciable. Introduction and discussion talk about clinical utility for triage after nuclear disaster. Is analysis of miRNAs purified from exosome a viable approach for triage and clinical decision making? If so, please provide convincing argument showing practicality.

Major comments:<br> 1. Longitudinal evaluation of specimens from human patients who received TBI is a plus. However, baseline readings in specimens collected from leukemia patients need to be compared with that in healthy humans. Why several specimens were excluded from analysis?<br> 2. Dose response noted is moderate. Biodosimetry refers retrospective evaluation of absorbed dose and the analysis should include validation using specimens of unknown exposure.<br> 3. Authors says that 1 Gy exposure in humans can cause ARS (paragraph 1, introduction). However their approach do not resolve dose under 4 Gy (around the LD50 value in humans).

Reviewer #2 (Public Review):

The study first compared the profiles of serum miRNA in patients before and after irradiation treatment. Then they selected 8 miRNA markers that showed significant changes in levels for further analysis. Then, they showed that the analysis of these markers by real-time PCR can differentiate the pre- and post-irradiation samples in 12 additional patients.

The objective of the study is unclear.<br> The study only demonstrates that the 8 miRNA markers are useful to differentiate serum samples collected before and after irradiation. This information is not useful as the blood picture would be more accurate and cheap to accomplish this task.<br> The authors also propose that these markers are useful for the identification of subjects exposed to irradiation. As this study has not addressed the specificity of these miRNA markers to irradiation, the claim of having a signature for radiation exposure is not justified.

The key new experiments in this study are the profiling of the serum miRNA in the patients undergoing total body irradiation. The results on mouse model and macaques have been published previously. The consistency of the changes of the miRNA markers is not surprising.

Reviewer #2 (Public Review):

This study investigates how uncertainty about spatial position is represented in hippocampal theta sequences. Understanding the neural coding of uncertainty is important issue in general, because computational and theoretical work clearly demonstrates the advantages of tracking uncertainty to support decision-making, behavioral work in many domains shows that animals and humans are sensitive to it in myriad ways, and signatures of the neural representations of uncertainty have been demonstrated in many different systems/circuits.

However, studies of whether and how uncertainty is signaled in the hippocampus has remained understudied. The question of how spatial uncertainty is represented is already interesting but recent interest in interpreting hippocampal sequences as important for planning and decision-making provide additional motivation.

A variety of experimental paradigms such as recordings in light vs. darkness, dual rotation experiments in which different cues are placed in conflict with another, "morph" and "teleportation" experiments and so on, all speak to this issue in some sense (and as I note below, could nicely complement the present study); and a number of computational models of the hippocampus have included some representation of uncertainty (e.g. Penny et al. PLoS Comp Biol 2013, Barron et al. Prog Neurobiol 2020). However, the present study fills an important gap in that it connects a theory-driven approach of when and how uncertainty could be represented in principle, with experimental data to determine which is the most likely scheme.

The analyses rely on the fundamental insight that states/positions further into the future are associated with higher uncertainty than those closer to the present. In support of this idea, the authors first show that in the data (navigation in a square environment, using the wonderful data from Pfeiffer & Foster 2013), decoding error increases within a theta sequence, even after correcting for the optimal time shift.

The authors then lay out the leading theoretical proposals of how uncertainty can be represented in principle in populations of neurons, and apply them to hippocampal place cells. They show that for all of these schemes, the same overall pattern results. The key advance of the paper seems to be enabled by a sophisticated generative model that produces realistic probability distributions to be encoded (that take into account the animal's uncertainty about its own position). Using this model, the authors show that each uncertainty coding scheme is associated with distinct neural signatures that they then test against the data. They find that the intuitive and commonly employed "product" and "DDC" schemes are not consistent with the data, but the "sampling" scheme is.

The final conclusion that the sampling scheme is most consistent with the data is perhaps not surprising, because similar conclusions have been reached from showing alternating representation of left and right at choice points cited by the authors (Johnson and Redish 2007; Kay et al. 2020; Tang et al. 2021) and "flickering" from one theta cycle to the next (Jezek et al. 2011). So, the most novel parts of the work to me are the rigorous ruling out of the alternative "product" and "DDC" schemes.

Overall I am very enthusiastic about this work. It addresses an important open question, and the structure of the paper is very satisfying, moving from principles of uncertainty encoding to simulated data to identifying signatures in actual data. In this structure, the generative model that produces the synthetic data is clearly playing an important role, and intuitively, it seems the conclusions of the paper depend on how well this testbed maps onto the actual data. I think this model is a real strength of the paper and moves the field forward in both its conceptual sophistication (taking into account the agent's uncertainty) and in how carefully it is compared to the actual data (Figures S2, S3).

I have two overall concerns that can be addressed with further analyses.

First, I think the authors should test which of the components of this model are necessary for their results. For instance, if the authors simply took the successor representation (distribution of expected future state occupancy given current location) and compressed it into theta timescale, and took that as the probability distribution to be encoded under the various schemes, would the same predictions result? Figuring out which elements of the model are necessary for the schemes to become distinguishable seems important for future empirical work inspired by this paper.

Second, the analyses are generally very carefully and rigorously performed, and I particularly appreciated how the authors addressed bias resulting from noisy estimation of tuning curves (Figure S7). However, the conclusion that the "sampling" scheme is correct relies on there being additional variance in the spiking data. This is reminiscent of the discussions about overdispersion and how "multiple maps" account for it (Jackson & Redish Hippocampus 2007, Kelemen & Fenton PLoS Biol 2010), and the authors should test if this kind of explanation is also consistent with their data. In particular, the task has two distinct behavioral contexts, when animals are searching for the (not yet known) "away" location compared to returning to the known home location, which extrapolating from Jackson & Redish, could be associated with distinct (rate) maps leading to excess variance.

Such an analysis could also potentially speak to an overall limitation of the work (not a criticism, more of a question of scope) which is that there are no experimental manipulations/conditions of different amounts of uncertainty that are analyzed. Comparing random search (high uncertainty, I assume) to planning a path to a known goal (low uncertainty) could be one way to address this and further bolster the authors' conclusions.

Reviewer #1 (Public Review):

When theta phase precession was discovered (O'Keefe & Recce, 1993; place cell firing shifting from late to early theta phases as the rat moves through the firing field, averaged over many runs), it was realized that, correspondingly, firing moves from cells with firing fields that have been run through (early phase) to those whose fields are being entered (late phase), with the consequence that a broader range of cells will be firing at this late phase (Skaggs et al., 1996; Burgess et al., 1993; see also Chadwick et al., 2015). Thus, these sweeps could represent the distribution of possible future trajectories, with the broadening distribution representing greater uncertainty in the future trajectory.

Using data from Pfeiffer and Foster (2013), they examine how neurons could encode the distribution of future locations, including its breadth (i.e. uncertainty), testing a couple of proposed methods and suggesting one of their own. The results show that decoded location has increasing variability at later phases (corresponding to locations further ahead), and greater deviation from the actual trajectory. Further results (when testing the models below) include that population firing rate increased from early to late phases; decoding uncertainty does not change within-cycle, and the cycle-by-cycle variability (CCV) increases from early to late phases more rapidly than the trajectory encoding error (TEE).

They then use synthetic data to test ideas about neural coding of the location probability distribution, i.e. that: a) place cell firing corresponds to the tuning functions on the mean future trajectory (w/o uncertainty); b) the distribution is represented in the immediate population firing as the product of the tuning functions of active cells or c) (DDC) the distribution is represented by its overlap with the tuning curves of individual neurons; d) (their suggestion) that different possible trajectories are sampled from the target distribution in different theta cycles.

The product scheme has decreasing uncertainty with population firing rate, so would have to have maximal firing at early phases (corresponding to locations behind the rat), contradicting what was observed in the data, so this scheme is discarded.

The DDC scheme has an increased diversity of cells firing as the target distribution gets wider within each cycle, whereas the mean and sampling schemes do not have increasing variance within-cycle (representing a single trajectory throughout). The decoding uncertainty in the data did not vary within-cycle, so the DDC scheme was discarded.

The mean and sampling schemes are distinguished by the increase in CCV vs TEE with phase, which is consistent with the sampling scheme.

The analyses are well done and the results with synthetic data (assuming future trajectories are randomly sampled from the average distribution) and real data match nicely, although there is excess variability in the real data. Overall, this paper provides the most thorough analyses so far of place cell theta sweeps in open fields.

I found the framing of the paper confusing in a way that made it harder to understand the actual contribution made here. As noted in the discussion, the field has moved on from the 1990s and cycle-by-cycle decoding of theta sweeps has consistently shown that they correspond to specific trajectories moving from the current trajectory to potential future trajectories, consistent with continuous attractor-based models (in which the width of the activity bump cannot change, e.g. Hopfield, 2010). Thus it seems odd to use theta sweeps to test models of encoding uncertainty - since Johnson & Reddish (2007) we know that they seem to encode specific trajectories (e.g. either going one way or the other at a choice point) rather than an average direction with variance covering the possible alternatives.

Thus, the main outcomes of the simulations could reasonably be predicted in advance, and the possibility of alternative neural models of uncertainty explaining firing data remains: in situations where it is more reasonable to believe that the brain is in fact encoding uncertainty as the breadth of a distribution. Having said that, most previous examples of trajectory decoding of theta sweeps have not been for navigation in open fields, and the analysis of Pfeiffer and Foster (2013; in open fields) was restricted to sequential 'replay' during sharp-wave ripples rather than theta sweeps. This paper provides the nicest decoding analyses so far of place cell theta sweeps in open field data. However, there are already examples of theta sweeps in entorhinal cortex in open fields (Gardner et al., 2019) showing the same alternating left/right sweeps as seen on mazes (Kay et al., 2020). Such alternation could explain the additional cycle-by-cycle variability observed (cf random sampling).

Refs not in paper:<br> Burgess N., O'Keefe J. and Recce M. (1993) Using Hippocampal Place Cells for Navigation, Exploiting Phase Coding, Neural Information Processing Systems 5: 929-936.<br> Chadwick A., van Rossum M. C. W. and Nolan M. F. (2015) Independent theta phase coding accounts for ca1 population sequences and enables flexible remapping. eLife 4: e03542<br> Gardner R. J., Vollan A. Z., Moser M.-B., Moser E. I. (2019) A novel directional signal expressed during grid-cell theta sequences. Soc. Neurosci. Abstr. 604.13/AA9<br> Hopfield J. J. (2010) Neurodynamics of mental exploration. PNAS 107: 1648-1653.

Reviewer #3 (Public Review):

Summary of the goals:

The authors set out to test the hypothesis that neural activity in hippocampus reflects probabilistic computations during navigation and planning. They did so by assuming that neural activity during theta waves represents the animal's location, and that uncertainty about this location should grow along the path from the recent past to the future. They next generated empirical signatures for each of the main four proposals for how probabilities may be encoded in neural responses (PPC, DDC, Sampling) and contrasted them with each other and a non-probabilistic representation (scalar estimate of location). Finally, the authors compared their predictions to previously published neural activity and concluded that a sampling-based representation best explained neural activity.

Impact & Significance:

This manuscript can make a significant impact on many fields in neuroscience from hippocampal research studying the functions and neural coding in hippocampus, through theoretical works linking the representation of uncertainty to neural codes, to modeling experimental paradigms using navigation tasks. The manuscript provides the following novel contribution to cognitive neuroscience:

- It exploits the inherent change in uncertainty about a parsimonious internal variable over time during planning to test hypotheses about probabilistic computations.

- A full model comparison of competing hypotheses for the neural implementation of probabilistic beliefs. This is a topic of wide interest and direct comparisons using data have been elusive.

- The study presents substantial empirical evidence for a sampling-based neural representation of the probability distribution over trajectories in the hippocampus, a finding with potential implications for other parts of neural processing.

Strengths:<br> - Creative exploitation of a naturally occurring change in uncertainty over a parsimonious latent variable (location).

- Derivation of three empirical signatures using a combination of analytical and numerical work.

- Novel computational modelling & linking it to neural coding using 4 existing implementational models

- Comprehensive and rigorous data analysis of a large and high-quality neural dataset, with supplemental analyses of a second dataset

- Mostly very clear and high quality presentation

Weaknesses:

- It is unclear to what degree the "signatures" depend on the details of the numerical simulation used by the authors to generate them. At least two of them (gain for the product scheme and excess variability for the sampling scheme) appear very general, but the degree of robustness should be discussed for all three signatures.

- The claims about "efficiency" lack a definition of what exactly is meant by that, and empirical support.

Reviewer #1 (Public Review):

Giove and colleagues find that a perceptual effect, namely whether a flicker is perceived or unperceived, is reflected in metabolic signals measured with functional MRS, but not in BOLD-fMRI. Specifically, perceived but not unperceived flicker led to an increase in lactate and glutamate in early visual cortex (a combination of V1, V2 and V3). BOLD-fMRI did not increase in this same region, suggesting that we are missing important neural signals by focusing on BOLD-fMRI only. The authors also provide a thorough discussion of the potential physiological mechanisms underlying these metabolic effects. I should note that I have no expertise in fMRS, and my assessment is based on knowledge of BOLD-fMRI and perception.

Whether or not the flicker was visible was manipulated by changing the frequency of the flicker. Specific, a low frequency flicker (7.5 Hz) was perceived, but a high frequency flicker (30 Hz) was not. Of course, this means that it is difficult to assess whether the fMRS effects are related to perception itself (visible vs. invisible) or due to the low-level features of the stimulus, e.g. the temporal filtering properties of the visual system. This limitation does not however hinder the main conclusion of the paper, which is that certain neural signals are missed by BOLD-fMRI but can be picked up by fMRS.

In Figure 2B, it looks like BOLD dynamics may differ between the slow and fast flicker blocks, even if the mean amplitude did not. So perhaps there are some more subtle BOLD differences between conditions that the authors do not explore.

The authors themselves also raise a potential partial voluming issue in the fMRS measurement that seems important to consider, given the differential BOLD signal in nearby regions (V2 and V3). Specifically, the volume in which fMRS is measured consists of parts of V1, V2, and V3. There are no significant differences between perceived and unperceived BOLD-fMRI in this volume as a whole, but there are in V2 and V3 in isolation. This raises the possibility that the null effect of BOLD-fMRI in the fMRS volume as a whole is due to it washing out in this larger volume. Could it be that the fMRS effects are also driven by V2 and V3, but are for some reason stronger/more robust, and therefore survive in the larger volume? In other words, I wonder if the BOLD and fMRS effects may actually co-localise, but differ in effect size.

In conclusion, the authors demonstrate an intriguing dissociation between BOLD-fMRI and fMRS, which should prompt further research into this topic, and may ultimately change the way we interpret neuroimaging signals.

Reviewer #2 (Public Review):

In this paper the authors investigate differences in metabolic response in primary visual cortex (V1) to perceptible and imperceptible stimuli using proton magnetic resonance spectroscopy (1h-MRS) and fMRI.

The main strength of this paper is it shows that perceptible stimuli trigger a different metabolic response in V1 than imperceptible stimuli, namely that lactate and glutamate levels both increase for perceptible stimuli but are unchanged for imperceptible stimuli. Weaknesses of the study are that no retinotopic mapping was performed on the subjects so the spectroscopic voxel may contain contributions from early visual cortex outside V1; the assumption that increased BOLD response in V2 is caused by perception is not convincing.

The differences in concentration of lactate and glutamate are striking, and the only plausible explanation is differences in metabolic response in V1. This is the clear and main result of the paper. The argument that an increased activation in V2 is caused by perception is less interesting. More sophisticated experimentation and analysis including connectivity analysis would be required to investigate the interaction between V1 and the rest of the brain.

This could considerably increase the importance of MRS in cognitive neuroscience. It would be fascinating to use dynamic causal modelling or a similar technique to explore connectivity between regions for perceptible/imperceptible stimuli and to combine this with proton spectroscopic imaging.

Reviewer #3 (Public Review):

Di Nuzzo et al demonstrate here that perception of visual stimulation is reflected in dissociable neurometabolic -but not neurovascular- responses in human visual cortex. This work uses human neuroimaging to show the effects of perception on neuronal energy demands and is of great importance for the neuroscience community. The authors carefully designed a task that would elicit similar BOLD response in primary visual cortex (V1) for perceived or unperceived visual flickering. They combined fMRI BOLD measurements with functional MRS, to quantify the functional (BOLD) and metabolic (concentration of lactate and glutamate) responses during visual stimulation. While they found no differences in the BOLD response within V1 for perceived vs. unperceived visual flicker, the authors show increased levels of glutamate and lactate in V1 when the flicker is perceived, suggesting increased energy metabolism during perceived visual stimulation.

While BOLD response within V1 does not differ between perceived and unperceived flicker (Figures 3B, 2C, 3C), the authors find enhanced BOLD in the lateral occipital cortex when the flicker is perceived (Figures 3B, 3D). The authors consider BOLD in secondary visual areas to be a surrogate measure of V1 output, indicating that stimulus processing during perceived stimulation results in enhanced V1 output. The spatial and temporal resolution more commonly used in human neuroimaging do not facilitate building relationships of input-output neuronal activity in a way analogous to animal neurophysiology. The assumption that BOLD activity in secondary visual areas reflects V1 output is very tightly linked to the unique architecture of the visual system; however, the paper would benefit from including the uncertainty of this assumption in the discussion.

The paper would further benefit from following a more standardised way of reporting pre-processing steps of the fMRI data, as well as a more detailed description of the statistical analyses on the fMRI data.

Finally, the authors have provided a series of well-chosen controls to ensure that their findings are not driven by differences in levels of attention between perceived and unperceived stimulation (Figure 1). The authors are commended on the quality of their figures, their choice of detailed graphs and the constructive use of additional media.

Reviewer #1 (Public Review):

This paper investigates what functional properties emerge from training an anatomically-constrained neural network on a specific computational task - detection of looming visual stimuli. Several functional models are identified by optimizing a network model for this task, and one of these models matches several properties observed in the fly neurons that perform the task. The approach and results are interesting. I did feel that several aspects of the work could be described more clearly, and that the potential of the model to reveal important aspects of the computation could be probed more thoroughly.

Inhibitory component of model. The interplay between excitatory and inhibitory components of the model could be explored in more detail. A specific aspect that is interesting is the inclusion of rectification in the inhibitory circuit. Rectification is motivated by the extra neuron in the circuit proving inhibition (lines 155-157), but it is not clear why an additional neuron would require rectification. Are their physiological measurements that indicate that the extra neuron introduces rectification, or is that a speculation? Exploring whether rectification is important would also be interesting - e.g. by removing it from the trained models, and/or training models on circuits in which rectification is absent. Lines 360-362 mention interesting response properties created by inhibition, but do not define what those are. Including some of these extensions of the basic model could highlight the potential of the model to make predictions about specific circuit features that are important for detection of looming stimuli.

Intuition for second model class. One of the key results in the paper is the existence of two classes of solution to the optimization problem - one of which follows the expectation for a detector based on outward optical flow, and the other of which does not. It is important to explain intuitively how the "inward" model is able to detect looming stimuli, given that it seems sensitive to the wrong optical flow features. This should be early - e.g. around lines 214-216.

In general, the results would benefit from developing some arguments in more detail. One example is the paragraph on lines 232-237. The differences in performance in Figures 8C and D stick out to me as a reader, but I am not guided through those differences in the text. Intuition for why you see the change in relative performance of the two solution (lines 266-268) would similarly be helpful. Another example is lines 290-292. These are several examples in which more explanation would be helpful, but you could look at the results in general with this in mind.

The performance of the two classes of solution becomes more similar as the number of neurons increases. A concern is that this reflects saturation of performance rather than actual equivalence of the models. Can you make the task harder, e.g. by adding distracting optical flow? That might help separate performance of the different models and avoid saturation.

Figure 10: how did you chose the specific outward solution used in this figure? More generally, some measure of the similarity of model components with experiment across all outward models is important. Currently the text reads as if you chose one of many models that happened to have components that looked like those measured. This comes up again on lines 310-311 and 313-315.

Are there animals that detect looming stimuli with fewer loom detectors? If so it would be interesting to see if they have adopted a similar or different computation.

Reviewer #2 (Public Review):

The manuscript from Zhou et al. investigates how certain features of looming-detecting neurons can arise from optimizing a shallow neural network to detect imminent collisions. The authors consider architectures that resemble the known anatomy of LPLC2 neurons in Drosophila, with excitatory inputs from the four layers of motion detectors in the lobula plate and inhibitory inputs from the interneurons in those layers. The authors find that some fraction of the trained networks exhibit tuning properties of LPLC2 neurons, including (a) similar response profiles to stimuli that are not present in the training data; (b) similar dependence on the angular size of the looming object as opposed to angular velocity; and (c) similar dependence between peak response time and the ratio of size to speed of the looming object. The authors also find another solution among the trained networks that is very different from the biological circuits. However, they show that this other solution becomes less common as the number of neurons grows, which is the relevant regime for the biological circuit. This paper adds to a body of work that suggests that the structural or functional properties of brain circuits are the solution to an optimization problem implied by the task that they have to perform -- in this case, the ability to detect looming motion.

The conclusions of the paper seem well supported within the class of models that was considered. The choice of class is, however, rather narrow and could be better explained and analyzed.

1. One potentially confusing aspect of the work is that there are in fact three major types of solutions that are found, not only two as described in the abstract: apart from "outward" (similar to LPLC2) and "inward" (dissimilar to LPLC2) there are also "unstructured" solutions that, as far as I understand, basically fail to perform the task -- although their performance isn't adequately discussed. The authors comment on this in the Discussion, suggesting that the unstructured networks are local optima where the stochastic gradient descent algorithm they use for optimization gets stuck. They argue that evolutionary processes would be unlikely to linger there, implying that it might be fine to ignore these solutions. While reasonable, this claim is difficult to assess without more discussion of these results. These solutions are not a rare occurrence: according to the Methods, over half of the trained networks end up in the "unstructured" pile.<br> 2. The stimuli used in the paper are very simple: basically rigid, featureless objects moving in a straight line and at constant velocity, or rotating at constant angular velocity. Naturalistic stimuli are likely to be much more complex, which could hurt the training process. This is only briefly touched upon in the Discussion, leaving open the question of how the results of this work would change in more natural settings.

3. The authors impose a 90-degree rotation symmetry as well as a reflection symmetry on the connection weights to the four layers of motion detectors that are sensitive to the four cardinal directions. Given that the training data that is used also has these symmetries, the question arises whether imposing these symmetries by hand was necessary. This is unfortunately not discussed in the paper.

4. One highly confusing aspect is that there is, in fact, an additional symmetry: the same filters are used for all the subunits. The difference between the different subunits seems to be only in the inputs that they receive -- i.e., that they are responding to different parts of the visual field. This is only really apparent from the Methods. Given again the rotational symmetry of the inputs, it would be reasonable to assume that this symmetry could be learned, but this isn't discussed or explained properly.

5. The authors say that the "outward" model reproduces biology but I'm not sure that the details of the lobula plate circuitry match this claim. For instance, LPi neurons typically have broad arbors, making location specific inhibitory inputs unlikely. And is there evidence that the inhibitory inputs are limited to a small region, like in the model?

6. Why not test the predictions of the model by analyzing the inputs onto the LPLC2 neurons using connectomics datasets?

Reviewer #3 (Public Review):

Although collision detecting neurons have been identified across animals, the computations they perform remain unresolved. Here, Zhou et. al train artificial neural networks to predict collisions across a diverse set of visual stimuli and constrain network geometry using the known anatomy of a Drosophila looming detector cell type, LPLC2. Zhou et al demonstrate that trained networks converge upon three solution types: an unstructured solution, a solution where inward motion is excitatory, and a solution where outward motion is excitatory. Interestingly, the solution excited by outward motion is also inhibited by inward motion as predicted for LPLC2 computations, and the output of these trained networks is highly similar to measured LPLC2 responses across stimuli.

1. Strengths:<br> a. The novelty of this study is that the networks are trained to solve a problem (collision detection) instead of being trained on neural data, but as a result are able to reproduce neural data.<br> b. The authors investigate how collision detection solutions change when looming is computed by a single neuron versus a population of neurons. This is particularly interesting because looming detectors have been identified at both population and single neuron levels. These results shed light on why many different collision detection computations have been proposed across neurons and across species, as they may face different anatomical constraints. The results also provide novel computations that can be further investigated in vivo.<br> c. The manuscript is well written, the figures are clear, and the movies are very helpful in understanding the approach and the results.

2. Limitations:<br> a. The findings could be strengthened by a more thorough characterization across the different solutions. For example, only two of many outward solutions are compared to actual neural data, and there is no explanation for why these two solutions were selected and whether they are representative of the entire category of outward solutions. There is also no metric for evaluating how well these solutions match the neural data.<br> b. The inward solutions are dropped from the last section of the paper; however, it would be very interesting to see the output of example inward solutions in comparison to actual neural data.<br> c. Within outward solutions, there are cases where inward inhibition is completely absent which does not follow what is known about LPLC2. The authors should mention this and also provide a comparison between outward solutions with or without inhibition.

Reviewer #1 (Public Review):

This group has examined basement membrane composition using sophisticated technical methods previously. Here they have methodically examined kidney organoids for their resemblance to mammalian foetal kidneys in the temporal expression of membrane proteins. They continue this through to adulthood and use peripheral blood leucocytes to demonstrate the effect of a COL4A5 mutation on the expression of basement membrane components.

The manuscript's strengths are its methodical nature and the number of proteins examined, as well as building on previous work.

Its weaknesses are that we do not know how good a model the organoid is for Alport syndrome and whether it results in an intact glomerular basement membrane. So far this manuscript has demonstrated that the organoids are consistent with what we know - but can it also tell us new things? In addition it has only examined one pathogenic Alport COL4A5 variant and this person also had a COL4A4 variant and thus complicated disease.

Reviewer #2 (Public Review):

Morais et al provide a convincing model for understanding basement membrane (BM) biology and interactions of BM components. The key findings of this paper are to establish a model that recapitulates the same biology and chemistry that is occuring during equivalent kidney development in humans, primarily. Utilizing kidney organoids, the authors characterize the spatiotemporal relationship of the proteins within kidney organioids as they form distinct basement membrane structures. They kidney is vital system in itself for understanding basement membranes among many different organs/tissues as they kidney has served as a genesis for discoveries over the last 60 years. Here the authors describe not only the timing of proteins in the development of kidney organoid BMs, but also the spatial relationships. Importantly, as a kidney BM model, the authors recapitulated the disease state of Alport syndrome (AS), a syndrome involving the disruption of the collagen IV α345 network in kidneys, an essential component of kidney BMs. Furthermore, they find that this model of kidney organoids derived from AS patients had the same hallmarks during development as AS in a human patient, including laminin overcompensation as a result of α345 network disruption.

This manuscript provides an invaluable model for understanding overall BM biology and disease progression, and especially so for kidney BM biology and kidney diseases. The potential for this model to study any number of missense variations within any number of proteins within a tractable and functionally identical BM is worth noting and exploring by other researchers.

In general, the weaknesses of this article are insignificant as this manuscript aims to provide functional proof of concept of kidney organoids as a model for understanding human BM disease. Importantly, however, is the assumption that kidney BMs might represent all BMs. The diversity of BMs across tissues within humans alone is significant. Amongst different organisms from a broader evolutionary standpoint than just fly, C. elegans, mouse, and human, BMs are very likely exceptionally diverse from the earliest animal BMs to different human tissues BMs. While this model provides an important model for understanding BM biology, a caveat that a kidney BM will functionally differ from a lens BM should be apparent and noted. However, the open-ended question of how to create tractable models like kidney organoids in other tissues systems will be of use in stimulating the matrix, proteomic, and structural biology fields.

Reviewer #3 (Public Review):

The emergence of methods to convert human induced pluripotent stem cells (iPSCs) into cultured kidney organoids that phenocopy the normal progression of embryonic and fetal differentiation represent a major advance in the study of normal and defective renal morphogenesis. This progress has been enriched by the addition of temporal/cell-type specific proteomics.

The current study largely focusses on the site-specific compositional changes that occur in basement membranes (BM) that form on different abluminal cell surfaces as differentiation advances. A general model of BM assembly from earlier studies provides a foundation upon which to interpret organoid kidney development. Laminins initiate BM assembly by binding to cognate cell-surface receptors, polymerizing, and binding to secreted nidogens, proteoglycans and collagen type IV, the last forming a second stabilizing polymer network. The iPSC differentiation system reveals the assembly and turnover of BM components consistent with the above, but now provides detailed information on the accumulation and turnover of different components in the key cell types through the different steps of differentiation with proteomic correlation. The approach also enables the analysis of the assembly defects and consequences arising from human congenital diseases as was shown with a type IV collagen alpha 5 subunit in organoids derived from Alport cells.

In combining organoid kidney culturing with laser microdissection and proteomic analysis, the authors have advanced use of the new tool compared to a 2018 study (Hale LJ et al., Nat. Communications), pushing the model from 18 to 25 days of differentiation and focusing more on BM formation during development. Evidence is presented to show that the major cell types, importantly including vascular endothelial cells, appear in the organoids in a temporal sequence. Relevant changes in BM-associated components are also shown. BM staining patterns are shown to change with emergence of laminin alpha5, laminin beta2 and collagen-IV alpha3 (replacing laminin beta1 and collagen-IV alpha1/2) at later stages. Organoids generated from iPSC cells derived from an X-linked missense variant of COL4A5 generated glomeruli containing alpha3/4/5, but with increases in laminin beta2, a known compensatory outcome.

The evaluation of later renal differentiation stages is particularly critical for the study of the glomerulus in which BM components undergo isoform switches that normally correlate with glomerular vascularization. A limitation of previous differentiation glomerular models has been the inability to show formation of the vascular tuft and the associated morphological changes as well as to show podocytes form inter-digitations. In that light, the current study could be strengthened by showing the ultrastructure of the day 25 glomeruli with identification of the BMs and different glomerular cell types (noting in particular if vascular endothelial cells are beginning to organize into the morphology of vascular tufts), and revealing the appearance of podocyte processes. It would also benefit the reader to enumerate the strengths as well as limitations with the culture model and how this work compares to previous studies.

Reviewer #2 (Public Review): 

The overall approach of this study is to compare gametocyte related parameters of infected blood samples from asymptomatic children, in some cases followed over time, with matched samples from uncomplicated malaria infections during the transmission season. A variety of parameters are analysed to investigate which mechanisms are used by the parasite to ensure that gametocytes are present in adequate numbers at the end of the dry season to be efficiently taken by mosquitoes reappearing at the start of the rainy season. 

Authors analyse expression levels of 333 P. falciparum gametocyte specific transcripts from a previously published transcriptome study on infected blood samples from asymptomatic children and from children with uncomplicated malaria. They conduct a Principal Component Analysis on 12 samples from each condition, revealing that PCA1 explains 65% of the seasonality variance, and identify 146 transcripts upregulated and 59 downregulated in the asymptomatic vs symptomatic carriers. This result may be expected as gametocyte physiology is likely different in the two very different conditions. 

The following step is investigating male and female gametocyte densities and relative proportions on asexual stages by Realtime analysis of specific transcripts. The main conclusions of this part of the manuscript are 1- that symptomatic malaria cases in the wet season have higher parasite levels and consequently gametocyte densities than asymptomatic cases in the dry season and 2- that the lower parasitaemia levels in the asymptomatic cases show a comparatively higher proportion of sexual stages, precisely of female and, albeit not significantly, of male gametocytes. 

Some observations and unclear points on this part of the work are the following: 

• More details are needed to clarify how density and proportion of the different parasite stages presented in Figure 1 panels C, D and E have been derived from the Realtime experiments.

• Male and female gametocyte numbers are calculated using published calibration curves for genes pfs25 (female) and pfMGET (males). Incidentally, these are published in https://malariajournal.biomedcentral.com/articles/10.1186/s12936-018-2584-y and not in reference 32.

• The calibration curves for the pfGLY and the pf17 transcripts, used to quantify total parasites and total gametocytes, respectively, and those used to quantify sexual commitment are instead missing and should be shown or referred to. This is relevant both for the determination of parasite numbers and for the use of the DeltaDeltaCt method.

• In panel C, it is intriguing that total parasite densities coincide with gametocyte densities. Comparing total gametocyte density in panel C with those of male and female gametocytes in panel D, it is puzzling that the former, quantified by the gametocyte specific marker pfg17, dramatically differs from those determined using pfs25 and pfMGET.

• It would be interesting to see if the proportion of male and female gametocytes on total parasites, calculated by the DeltaDeltaCt method in panel E, is comparable with the value that can be calculated from the density data in panels C and D.

• Finally, it would be valuable to derive from these data the gametocyte sex ratio to more synthetically describe the different infections.

A general point on this part of the work is that the transcriptome work was conducted on 12 vs 12 samples from asymptomatic and clinical cases, whereas the subsequent experiments were conducted on 35 samples from the dry season and 27 clinical cases. This is not very clear from the text and it rather looks like that the latter experiments were designed to better characterise the samples used in the transcriptome analysis. Also, the subsequent analysis of phospholipid levels and of the sexual commitment transcript levels appear to be performed on a subset of the samples. <br> In the following section, authors describe a longitudinal study on a smaller cohort of children from the above asymptomatic cases. Analysis of parasite densities, of phospholipid levels and of transcript levels of genes involved in sexual commitment essentially confirms the single point analyses described in the first part of the manuscript. The issues raised above on the determination of male and female gametocyte numbers and relative proportions conducted on these samples apply here as well. 

As 2-4fold reduced levels of phospholipids, described to affect rate of gametocytogenesis, were measured in the samples from uncomplicated malaria, sera from symptomatic and asymptomatic cases were compared to measure whether those with reduced phospholipid levels induced sexual commitment of in vitro cultivated parasites. However, as authors themselves state, the observed differences in Lyso-PC levels between the sera were predictably too small to produce an effect in these experiments, as compared with the fold difference described to see effect on sexual commitment. 

Having concluded so far that level of early gametocyte and sexual commitment markers do not change in the wet vs the dry season and that the low parasitaemias associated with the latter condition show a higher proportion of sexual stage transcripts/cells, another mRNA expression comparison has been conducted on a set of 163 transcripts selected as some "define early stages" and others "are characteristic of late stages". Assumptions, rationale and conclusions of this analysis are not entirely clear. The predictive value of the gene sets is not clear: available gametocyte stage-specific transcriptomic data fail to identify large numbers of transcripts unambiguously and specifically associated to early vs late gametocytes, so use of the 163 genes requires details on their stage specific diagnostic power. The analysis shows that 43 out of 54 "late gametocyte" transcripts vs 42 out of 109 "early gametocyte" transcripts are upregulated in the samples from asymptomatic cases, which leads authors to propose that this condition has "an effect during the 8-12 days of gametocyte development, in both/either the sexual and/or the asexual parasite compartments". This part requires attention and a clearer formulation of rationale and conclusions. 

In conclusion, this work produced evidences consistent with the hypothesis that the efficient clearance of asexual stages in low parasitaemia asymptomatic infections explains the increase in the proportion of mature circulating gametocytes.

Reviewer #1 (Public Review): 

Demonstrating that the proportion of asexual parasites that convert to gametocytes is similar in asymptomatic and symptomatic infections using the established biomarker ap2-g, suggests that the apparent increase in gametocyte density observed during symptomatic infection is due to an increase in total parasitemia, not a higher conversion rate. Moreover, they show that that mean conversion rate is not changed even though lysophosphatidylcholine (lysoPC) concentrations are significantly lower during a symptomatic infection. LysoPC deficient media has previously been shown to stimulate gametocyte conversion in vitro, but they confirm in vitro that lysoPC levels did not drop low enough in the symptomatic plasma to enhance conversion. Together the results suggest that under these conditions environmental factors are not influencing gametocyte production at the population level, which is a major question in the design of transmission blocking strategies. 

Additional information about the quantification of parasite and gametocyte numbers is needed to critically analyze the data comparing asexual and total, male, and female gametocyte proportions in symptomatic infections and overtime in asymptomatic infections. As they state in line 287, these comparisons are complicated by the long development time for P. falciparum gametocytes. The mature gametocytes in a blood samples have been sequestered for ~10 days as they develop prior to release into the blood stream, while the asexual parasitemia continues to expand through 4-5 life cycles. Consequently, it is difficult to compare the asexual/gametocyte ratio in a sample from a chronic ongoing infection and an acute symptomatic infection. This difference also complicates the transcriptome analysis. The symptomatic children would have to be followed overtime to track gametocyte dynamics. Another advance is the development of a model that recapitulates their data and should allow future analysis of factors that would have the most impact on reducing gametocyte production.

Reviewer #1 (Public Review):

In the manuscript by Li Enjie. et al., the authors determine that METTL3 is important for HR and deletion sensitizes MCF10A cells to Doxorubicin. Importantly, the authors show that METTL3 functions by regulating RAD51 via m6A modifications of RNA which changes how EGF expression factors regulating cellular invasion. This is important for cancer biology as m6A RNA methylation levels, METTL3 and RAD51 proteins are elevated in certain cancers and are targets for therapeutic intervention. The authors show that METTL3 knockdown in MCF-7 cells sensitize cells to the chemotherapeutic and topoisomerase II inhibitor Adriamycin (Doxorubicin). MCF-7 and MB-231 cells treated with Doxorubicin have elevated pro-apoptotic factors. The authors then go on to show that MCF-7 and MB-231 cells have increased gamma-H2AX protein levels and foci when METTL3 is knocked down, treated with Doxorubicin or Etoposide. Using a dr-GFP reporter assays the authors show that METTL3 knockdown decreases repair via HR. Using RNA-seq coupled with meRIP-qPCR the authors identified targets of METT3-mediated m6A modification to identify EGF. The authors then go on to show that EGF is a novel regulator of RAD51 expression. This is a significant finding as RAD51 is the key protein involved in HR that catalyzes homology search and strand exchange. The authors then went on to use EGF inhibitors to show in cells treated with ADR, METTL3 modulates HR via the EGF-RAD51 axis. They further define important m6A readers to include (YTHDC1) which regulates EGF mRNA stability and promotes EGF synthesis. The results of the study are potentially important for those interested in signaling and cancer invasion, as well as the DNA repair community, as the work characterizes transcript regulation by m6A binding and links it to HR and EGF signaling.

Strengths:<br> • The authors identified a novel protein involved in HR whose depletion sensitizes cells to TOPII inhibitors.<br> • The authors identified a new mechanism of regulation of key DNA repair proteins, mainly RAD51.<br> • They also find that this protein is regulated by EGF a known protein involved in cancer progression as well as DNA Repair.

Weaknesses:<br> • While doxorubicin was a standard therapy in breast cancer treatment, Olaparib is now commonly used in the clinic. It is unknown how this cancer treatment therapeutic would have on the mechanism proposed by the authors.<br> • Gamma H2AX phosphorylation is a global marker of DSBs and stalled forks. The authors failed to note that H2AX phorylation is present and a marker of stalled replications forks.<br> o PMID: 11673449, PMID: 20053681, doi:10.1101/gad.2053211, https://doi.org/10.1016/j.cell.2013.10.043 etc.

Reviewer #2 (Public Review):

In this study, the authors sought to define the influence of the METTL3 m6A methyltransferase on cellular resistance to Adriamycin (topo2 poison) and homologous recombination, along with studies aimed at linking such effects to METTL3 being important for expression of EGF, which in turn promotes expression of the RAD51 recombinase. A final set of experiments also examine the link of the m6A reader TYHDC1 to EGF expression.

The authors use a series of DNA damage response assays, and find some interesting results with METTL3 overexpression and depletion, and EGF expression. Since homologous recombination is a major end point of the study, which is most active in S/G2, then cell cycle analysis needs to be employed to determine if the effects are due to cell proliferation changes. This is particularly relevant for experiments with growth factor manipulations, such as EGF expression. Also, while interesting results are shown with METTL3 overexpression and depletion, not many of the experiments test both manners of manipulating METTL3 levels, which is important for interpreting the results (namely are effects only observed with overexpression for some and only depletion for others?). Finally, a major mechanistic idea in the manuscript is centered on increased levels of the RAD51 recombinase, but then testing whether the phenotypes can be mimicked by overexpression of RAD51 should be tested, particularly since the literature is not consistent about whether elevated RAD51 indeed causes more homologous recombination.

Reviewer #3 (Public Review):

METTL3 is part of a N6-methyltransferase complex that methylates adenosine residues at the N6 position of some RNAs and regulates various vital processes including DNA damage response. To date, the potential mechanism of METTL3 in the chemotherapeutic response is poorly defined. A previous study reported that METTL3 participates in the regulation of homologous recombination (HR) based DNA damage repair. Interestingly, in this paper, Li and colleagues added a piece into a puzzle by describing a parallel pathway where METTL3 promoted EGF expression through m6A modification, which further upregulated RAD51 expression, resulting in enhanced HR activity. Additionally, the m6A reader YTHDC1, bound to the m6A modified EGF transcript promoted EGF synthesis, which enhanced HR and cell survival during Adriamycin treatment. Hence, METTL3 knockdown results in DNA damage accumulation, which renders breast cancer cells sensitive to ADR treatment. This study establish the rationale for the development of METTL3 inhibitors for cancer treatment. The paper is well-written and the authors' claims and conclusions are justified by their data. Some figures will benefit from editing and proper quantification of the data.

Joint Public Review:

The authors have previously reported on the effects on skeletal muscle of Flt1 loss of function in endothelial cells, including amelioration of the dystrophic phenotype. To better understand this mechanism, they now identify a low level of endothelial cell gene signature in satellite cells and report that VEGFA-FLT1 axis functions as a pro-survival signal in SCs in response to acute muscle injury and in a mouse model of Duchenne muscular dystrophy. Bulk and scRNA sequencing analyses with robust controls were employed and results confirmed using RNA scope on sections. Interestingly, FLK1 has been considered as the receptor for VEGFA signaling, and FLT1 is considered to be a decoy receptor, with stronger affinity for VEGFA, but weaker cytoplasmic signaling. The authors determine that VEGFA binding to Flt1 in SCs promotes AKT-signaling cascade to support cell survival.

However, the results obtained in vivo need to be reinforced. Specifically, is not clear whether the authors measured satellite cells specifically, or survival, or total muscle cell (myocyte, endothelial, inflammatory, smooth muscle....) survival. Also, since the authors used Pax7tdT mice they should be able to track the fate of VegfA+/Hyper, VegfaKO or Flt1KO satelite cells upon injury and to quantify the number and/or percentage of tomato+ myocytes after injury in each model reflecting the specific consequences of impaired VEGFA signaling in satellite cells. While measuring fibrosis may indeed reflect a myocyte loss, measuring myocyte diameter more likely reflects myocyte differentiation which is proposed (in vitro) not to be modulated by VEGFA-FLT1 signaling. Some methods also need quantification and clarification. Overall the results support their conclusions and the work will be useful for research groups interested in understanding the developmental links between skeletal muscle and blood vascular lineages.

Reviewer #1 (Public Review):

In the present work Valperga and de Bono performed a forward genetic screen to identify candidate genes that would fulfill two criteria when mutant: 1) enhance an escape response to high ambient oxygen but 2) without modifications in the respective oxygen sensing neurons. They found that qui-1 mutants meet these criteria. qui-1 is known to act in the nociceptive neurons ASH and ADL (among others). The authors show that in qui-1 mutants ADL neurons are defective in normal chemo-sensation and upregulate neuropeptide secretion. This is associated with increased gene expression of neurosecretion components in ADL, among them two GPCR receptors (npr-22 and tkr-1); mutants in these receptors partially phenocopy the neurosecretion phenotype. The authors suggest an intriguing model in which ADL, upon loss of its normal sensory properties, relays peptidergic input from oxygen sensory circuits to peptidergic output towards yet unidentified downstream circuitry. This novel mechanism of sensory cross modality expands on on previous work on cross modality in C. elegans, where until now only one example been demonstrated, and where a different mechanisms than in the present study was described (Rabinowitch 2016). These findings could serve as generalizable models for other systems where cross-modal plasticity has been observed. Although many conclusions in this work are substantiated by cell specific rescue of qui-1 in ADL others are made based on correlated observations only. The study therefore would benefit from additional experiments that demonstrate a causal link between elevated neurosecretion in ADL and the associated changes in behavior. This could be achieved by ADL cell ablation experiments and specific interference with ADL neurosecretion.

Reviewer #2 (Public Review):

Loss of one sensory modality is often compensated with an increase in another sensory modality. Valperga and de Bono identify a possibly conserved mechanism that appears to heighten the worm's sensitivity to O2 while dampening other sensory responses. The mechanism that they discover suggests that increased neuropeptide secretion could be responsible for the overcompensation for a loss of a sense. The combined data based on forward genetic screening and behavioral analysis, imaging and genomics are convincing and interesting.

1. I very much enjoyed reading a manuscript that uses 'good old' forward genetics to make an interesting discovery!

2. The paper is well written and very easy to follow. The data quality and their display in the figures are very convincing, too.

3. The proposed mechanism of using enhanced neuropeptide secretions for compensating the loss of one sensory modality with an increase of function of another is novel and could indeed be conserved.

Reviewer #3 (Public Review):

The work by Valperga and de Bono aims to uncover molecular components of cross-modal plasticity, a system-wide form of neuronal remodeling that responds to sensory loss by altering the performance of remaining sensory modalities. The study focuses on the interplay between oxygen-sensing and pheromone detection in C. elegans. The data presented are mostly convincing and revealing. However, the message and the overall context within which the findings are framed are problematic.

The authors rightly assert that the molecular processes underlying cross-modal plasticity are not fully understood. However, they emphasize that the important challenge is to reveal genetic lesions that result in sensory loss and drive cross-modal plasticity. I find this to be over-specific and imprecise. There are many possible causes for sensory loss, some are genetic, some are non-genetic (e.g., certain diseases and injuries). In any case, the causes for sensory loss are usually independent of the processes that give rise to cross-modal plasticity. The genetics behind cross-modal plasticity enables the response to sensory loss, it does not cause the sensory loss. Genetic lesions to genes involved in cross-modal plasticity disrupt cross-modal plasticity, they don't induce it. Curiously, the authors sought to find single genes whose removal is simultaneously associated with both the loss of a sensory modality and the enhancement of another. This was done using a forward genetic screen for C. elegans mutants displaying enhanced oxygen sensation.

The analysis was further complicated by the fact that the screen was performed on strains whose oxygen sensitivity is already modified due to dysregulated activity in the RMG hub-and-spoke neural circuit, which integrates diverse sensory signals to control locomotion. Mutagenesis was performed on either the N2 strain, exhibiting RMG suppression, and thus decreased oxygen sensitivity, or flp-21 mutants, displaying excessive RMG activation, and increased oxygen sensitivity.

The screen yielded a gene, qui-1, whose dysfunction led to enhanced oxygen sensing (it is unclear if this is in the N2 or flp-21 background). The authors found that increased neuropeptide release from the pheromone-sensing neuron ADL underlies the increase in oxygen sensitivity. Furthermore, the qui-1 mutation was shown to diminish ADL pheromone responses. Therefore, a very particular genetic coupling between loss of pheromone sensation and enhanced oxygen sensitivity was revealed.

To generalize this finding, several additional mutant genes (not from the screen) were examined, including genes from the BBS family as well as wrt-6 and fig-1. They too displayed enhanced oxygen sensing linked to increased ADL neuropeptide secretion. However, their effects on ADL pheromone sensation were not reported. The main conclusion I draw from these findings is that the ADL neurons are able to modulate oxygen sensitivity by relaying information about oxygen levels from the RMG circuit to locomotor circuits via neuropeptide secretion. It is not at all clear that loss of pheromone sensation in the qui-1 case is the cause for increased neuropeptide release, or whether it is just one out of the many outcomes of mutating this gene. A much cleaner and more revealing experiment could have been, for example, to examine worms lacking the functional pheromone receptor OCR-2 in ADL. In fact, unlike qui-1 mutants who showed diminished oxygen responses in ADL, previous work from the de Bono group (Fenk and de Bono 2017) demonstrated that ADL O2 response are normal in ocr-2 mutants, indicating a profound difference between loss of pheromone sensitivity due to receptor dysfunction (ocr-2) and the unknown and broad effects of qui-1.

In fact, it would be interesting if the authors could explain or speculate how qui-1 eliminates ADL O2 responses, and how neuropeptide signaling from the RMG circuit via the NPR-22 neuropeptide receptor bypasses this lack of response and drives enhanced neuropeptide secretion in ADL, as they report.

The work includes a transcriptomic analysis comparing ADL-specific gene expression between wild type and the qui-1 mutant. Unlike other experiments in the study, in which the specific effects of mutations were confirmed through rescue experiments and the use of additional alleles, thus eliminating potential confounds with background mutations, the transcriptomic experiment did not apply such controls. Therefore, it is hard to conclude whether the reported changes in transcription are due solely to the qui-1 mutation or to other unrelated genetic modifications in the mutant strain.

Overall, except for where mentioned, the data presented are solid and consistent. However, the conclusion that the study reveals a molecular pathway for cross-modal plasticity is less convincing. The chain of events does not include some form of sensory loss, leading to subsequent, independent neural plasticity, as expected for cross-modal plasticity. Rather, a very broad genetic switch is described that can simultaneously change receptor abundance and neuropeptide release. Thus, an equally interesting and more coherent framing of the data could be that the study uncovered a genetic regulator, yet to be fully characterized, of oxygen-dependent behavior in a non-oxygen sensing neuron, adding to previous literature on neural circuit cross-talk.

Reviewer #1 (Public Review):

This paper provides a new method, smfATAC-Seq, which can be used to identify epigenetic alterations that occur in myofibers during muscle regeneration or muscle diseases. Through extracting single myofibers from the Extensor Digitorum Longus (EDL) muscle fiber followed by ATAC-Seq, the chromatin accessibility profile of a single myofiber can be obtained for further analysis. Using the smfATAC-Seq, sufficient reads can be obtained from one myofiber containing around 200-300 myonuclei. This method allows for a small input amount and is easy to follow. The authors show that the chromatin accessibility profile of myonuclei is different from that of muscle stem cells (MuSCs) and changes upon injury and disease. Further analysis of the smfATAC-Seq data may allow for the identification of active regulatory elements in muscle fibers.

Although the approach does have strengths in principle, the design of the experiments and data analyses performed are superficial. Notably, the data size (i.e., the number of myofibers evaluated) is insufficient to conclude to support the claims in the manuscript.

Major points:

1) One muscle contains hundreds of myofibers, while the authors only show 2-4 myofiber replicates for each condition. The authors claim that this method can be used to distinguish different fiber types. However, there is no evidence to support such a claim. Instead of using EDL, the approach should be applied to a muscle that contains a ratio of both fast and slow fiber types, indicating the heterogeneity among myofibers in one muscle in different conditions. In addition, the myofibers from the injured or disease muscles are highly heterogeneous in terms of their regeneration status. What is the rationale for choosing the myofibers? Were all the myofibers injured with central nuclei from end to end? Or is it partial? What is the diameter of this muscle fiber? Can the smfATAC-seq be a method to tell us about the maturity of the myofibers? Unfortunately, the design of the experiments did not provide any interesting biological insights.

2) The authors claim that their data "revealed a repertoire of active cis-regulatory elements", but no supporting evidence is provided. In the manuscript, only the smfATAC-Seq signal coverage across the genes known to be functional in muscle development was shown. Identifying the active cis-regulatory elements is essentially impossible without combinational analysis with other epigenetic profiles (e.g., H3K27ac ChIP-Seq, Hi-C). The results presented serve as validation but not an exploration of the regulatory elements for MuSCs and myofibers.

3) Muscle regeneration is a long-term process that could take a long time to complete depending on the age of the animal and the severity of the injury. The authors examined the chromatin accessibility profile of the myofibers in uninjured and 7 days post-injured muscle. This short time frame does not provide sufficient information to interpret the chromatin accessibility changes of myofibers during the whole regeneration process. It is difficult to understand the result of these experiments (comparing the uninjured fibers to injured fibers or WT fibers to MDX fibers). What does the ATAC-seq data add to our understanding that these myofibers are different? From a molecular point of view, Can this analysis provide a set of biomarkers of the myofiber cell states during regeneration and disease, for example? Again, the design of the experiments is superficial.

Reviewer #2 (Public Review):

In this paper, Sahinyan and colleagues developed a method for analyzing chromatin accessibility in single murine myofibers. This goal was achieved by adapting the previously published OMNI-ATAC protocol to the specific properties of the myofiber environment. To demonstrate the validity of this method, they isolated myofibers from uninjured and regenerating murine EDL muscles dissected from wild type animals. In a second experiment, this method was applied to isolate myofibers from mdx mice, a model of Duchenne Muscular Dystrophy. The resulting datasets were further compared to the one generated from purified muscle stem cells.

Strengths<br> In general, the authors provided robust quality controls for these datasets, which ensures the validity of their observations. Analysis of chromatin accessibility using this protocol enabled the identification of subsets of peaks specific for each experimental group, which were further analyzed to determine enriched biological processes.

Weaknesses<br> While the experiments are well executed, the resulting data are descriptive and do not provide further insights into the biological processes under investigation. A more comprehensive analysis could significantly increase our knowledge of the molecular pathways controlling skeletal muscle response to acute or pathogenic injuries.

Reviewer #3 (Public Review):

In this study, the authors adapt the OMNI-ATACSeq for single muscle fiber ATACSeq. This technique, dubbed "smfATAC-Seq" is used to demonstrate epigenetic changes in injury and the mdx dystrophic mouse model. Single-cell ATACSeq has been used to characterize cellular epigenetic heterogeneity of other tissues. However, the fused, multinucleated nature of muscle fibers makes skeletal muscle intractable to this method. Thus epigenetic studies have been restricted to either whole muscle, or single cells within the tissue (eg muscle stem cells, endothelial and immune cells). While the study is primarily descriptive, the smfATAC-Seq method presented is technically thorough, and will be valuable for the muscle field. Furthermore, the data produced is a good resource that can be mined to generate testable hypotheses in the future.

Strengths<br> The methods are presented clearly, and the most often studied conditions in the muscle field (injury and dystrophy) are appropriately used as examples to demonstrate the utility of smfATACSeq. In addition, the authors show that the data generated is reproducible, and can capture known aspects of myofiber heterogeneity.

Weaknesses<br> Since the paper is largely focused on a new method, more emphasis should be placed on what kind of questions can be uniquely answered by smfATCSeq. The authors show selected tracks between smfATAC-Seq and DESeq, however, this is a qualitative comparison. Authors should also provide a more systematic comparison of smfATAC seq with other published epigenetic datasets in whole skeletal muscle (DNaseq, ATACSeq, etc) -- for example, compare quality metrics such as sequencing depth or quantify overlap between identified peaks across methods.

Reviewer #1 (Public Review):

The study shows nicely that the calcium binding protein KChiP3 is associated with poor survival of the colorectal cancer (CRC) cohort analyzed and could be a potential prognostic marker. Decreased KChiP3 expression also reduced cell survival upon FOLFIRI treatment making its impact interesting. KChiP3 has previously been indicated by this group to regulate mucus release from the used colon carcinoma cell line and has in other studies been implicated in regulation of calnexin and potassium channels. The inhibitors Benzamil, an Amiloride derivative, and SN-6, a NCX inhibitor, both further reduced survival of colon cancer cells treated with FOLFIRI. These findings reveal an interesting potential of positive effects by manipulating calcium control in the cells to enhance the effect of FOLFIRI treatment of CRC patients.

The mechanism is speculated to involve modulation of the protective mucus secreted by the cells. The experimental setup is largely based on two subclones of the cancer derived cell line HT29 (M6 and 18N2). These cells produce and secrete different gel-forming mucins, mainly MUC2 and MUC5AC but this could vary depending on confluency, differentiation, and polarization. The inhibitory effect of mucus was addressed by the studying the effect of FOLFIRI treatment on MUC5AC production and secretion. The initial experiments showed increased production of MUC5AC using western blot analysis with a peak intensity of the protein after three hours of FOLFIRI treatment. The band, as determined by the size, corresponds to the non-glycosylated newly produced mucin product from the endoplasmic reticulum. These results are supplemented with detection of increased intracellular intensity in cells stained for MUC5AC, but these findings do not seem to be reproduced in the following control experiments studying effects of Benzamil or SN-6. The amounts of transcriptionally expressed or mature secreted mucus is however not determined upon FOLFIRI treatment nor in the experiments manipulating mucus secretion by altered KChiP3 expression or inhibited NCX function. In this context the extracellular mucus could be a combination of truly secreted mucus and intracellular components from detached or dead cells, but how this contributes to cell survival is not known. The inhibitor effect of mucus on treatment of CRC cells is speculated to be though binding of the drugs to the mucus. This is studied indirectly by binding of albumin to cells with or without secreted mucus. Mucus binding is however influenced by several factors as mucus composition, organization, and environmental impacts and as the binding affinity is dependent on the molecule the difference between albumin and FOLFIRI is a concern. The provided images also indicate mucus free areas that could allow for compound access if not covered by other unstained mucins. Organoids were used to assess the effect of SN-6 which also in this system reduce cell survival combined with FOLFIRI treatment but the link to mucus secretion was not explored. Taken together the basis for the arguments that mucus is influencing the accessibility of the supplemented drugs is not fully supported by the experiments provided.

The CRC survival analysis shows that MUC5AC expression levels were not found to be associated with disease outcome but the influence of other mucins was not analyzed. Some previous studies suggests that MUC2 expression is associated with better survival while MC5AC show opposite effects. As the analysis do not link mucus secretion to treatment efficacy combined with analyses of a role for other regulatory proteins it is not clear how this adds to the aim of the study.

Reviewer #2 (Public Review):

The main goal of this work is to test the hypothesis that (a) secreted mucins act as a barrier to protect CRC cells from toxic drugs, and (b) the level of KChIP3 is a determinant of mucin secretion and therefore of chemoresistance in CRC. Experiments with cultured CRC cells and organoids provide solid support for the hypothesis. Apart from some suggestions about presentation as indicated below, the science is persuasive, and it is likely to prove useful for developing improved treatment options for CRC and other mucinous tumors.

Perhaps the most interesting biological question, touched on but not answered in the manuscript, is why chemotherapeutic drugs increase mucin secretion in CRC cells. The implication is that CRC cells have a built-in resistance pathway for toxic drugs, but it is hard to see how such a pathway could have evolved specifically to promote chemoresistance. A possible explanation is that increased mucin secretion is a general stress response in certain cell types, both normal and cancerous. Experiments to explore this concept would enhance the paper by linking cell biology to potential clinical approaches.

Reviewer #1 (Public Review):

This paper by Lu et al. is an exciting new contribution to our understanding of transport, especially in the context of oogenesis. Two stages of transport from nurse cells to oocytes, in the Drosophila egg chamber, have long been accepted: early selective transport and late dumping. The authors clearly establish that transport by dynein-driven advection is occurring before dumping begins. Using a combination of genetic knock downs, labeling and live imaging, the authors show that dynein and BicD are necessary for proper oocyte growth and focus on stage 8/9 of development. They proposal a model in which dynein-driven microtubule transport pulls cytoplasm along with it. Supporting this model, they have beautiful movies directly demonstrating movement of microtubules and several cellular components through the ring canals. The go on to show that passive flow of artificial cargo that does not interact with dynein or microtubules is transported as well. Importantly, they show that a minus end directed motor construct, with no dynein cargo interaction domains, when directed to the nurse cell cortex, is sufficient to drive this transport as well. Careful controls, such as eliminating a role for myosin are included.

The claim is made that this movement of microtubules and associated cytoplasm is robust in stage 9 and maybe 8 but data are not presented regarding this timing. How early the streaming starts should be established, in order to determine whether or not there really is a slow selective stage first.

The authors note that the cytoplasmic rings necessary for such transport between cell types, are observed in developing oocytes of many species. Cytoplasmic streaming is seen in other contexts, including later in Drosophila oogenesis, within the oocyte, and within the plant syncytium. Thus I wonder how many other cases we are still missing.

Reviewer #2 (Public Review):

In this manuscript, Lu and colleagues demonstrate a novel role for Dynein in mediating bulk cytoplasmic transport into the oocyte. This function of Dynein is generated as a result of microtubule gliding rather than specific cargo transport. The conclusions are well-supported by the data. Appropriate controls have been included and the data have been quantified. In addition, the numerous live movies that have been included really help to visualize this process of bulk transport. The notion of bulk transport is best supported by the experiment utilizing GEMs (Figure 3). In addition, the ability of the chimeric cortically anchored plant Kinesin motor to restore oocyte growth in Dlic depleted egg chambers underscores the notion that cargo-independent bulk transport is critical for this process.

Reviewer #3 (Public Review):

The authors address the role of dynein, a key minus-end directed microtubule motor in cytoplasmic polarized transport to the oocyte in the developing Drosophila egg chamber. Nurse cells to oocyte mediated transport is essential for oocyte growth and previous studies have shown that dynein mediated transport is important for this process. The authors with this present work shed a new light on this process.

The dynein motor is required sequentially during oocyte and nurse cell development. To circumvent the requirement of the dynein motor complex in oocyte specification, the authors study oocyte growth by taking advantage of germline-specific Gal4 that is expressed after oocyte specification with different RNAi against dynein components. They thus illustrate that dynein core components and regulators are indeed essential for oocyte growth. By analyzing the putative requirement of BicD, Spindly and Hook, they show that BicD is the most important dynein activating adaptor for oocyte growth.

By combining photoconversion of a-tubulin and fluorescently labeled microtubule-associated proteins, they interestingly show that microtubules are very dynamic both in the nurse cells and in the ring canals that separate nurse cells and oocyte. They provide evidence that the movement of microtubules through the ring channels is not part of myosin-II-induced nurse cells contraction, but requires the function of dynein, supporting the idea that dynein causes microtubules to slide into the nurse cells and through the ring canals to the oocyte.

As previously shown, the authors report a dynein-dependent cytoplasmic flux and movement of organelles from nurse cells to the oocyte. However, by combining an ingenious assay with genetically encoded nanoparticles that display Brownian motion in tissue culture cells, they were able to show that direct dynein-cargo interaction is not necessary for this process and that neutral particles can be efficiently transported through the ring channels by dynein microtubule-associated movement of cytoplasmic flow.

The authors next investigate how the dynein motor complex glides microtubules from the nurse cell to the oocyte. They show that in the nurse cells dynein presents a cortical localization.

In order to study this possibility, the authors carried out two particularly ingenious types of construction (F-tractin-BicD with BicD-RNAi and F-Tractin-Kin14Vib motor domain with Dlic-RNAi), which enabled them to highlight that cortically-anchored microtubule minus-end motors that cannot directly transport cargoes drive oocyte growth. They further provide evidences that the C terminus part of Dlic subunit could localize the dynein complex to the cell cortex independently of the dynein activating adaptors BicD, dynactin and Lis1.

Overall this work provides several interesting and new findings illustrating a novel mechanism for dynein-associated cytoplasmic transport. In the nurse cells, and in the ring canals, dynein motors anchored to the actin associated cortex glide microtubules that generate a cytoplasmic flow that move cargoes to the growing oocyte. This corresponds to a new mechanism of cargo transport required for Drosophila oocyte growth. The conclusions of this paper are well supported by the data. Several experiments have been carried out with original and innovative tools that reinforce the scope of this work. The quality of the results presented is very good in general. The figures and the movies are generally of good quality and well explained.

However, the authors should also discuss that previous studies using fluorescent labelled RNA injection into nurse cells have shown that specific mRNAs are delivered in the oocyte through a dynein-mediated transport distinct from cytoplasmic flow (doi:10.1242/dev.02832).

They should also discuss more about the velocity range of the various cargoes they analyzed as they passed through the ring canals.

Reviewer #1 (Public Review):

Lopes et al present an exploration of the functional interactions between developing CD4+ T cells and mTEC in the thymus. The study is interesting both because the precise differentiation stages for mTEC development is currently in flux with substantial recent discoveries, and because the manner in which developing T cells influence this development is also currently in question. The finding that CD4 cells induce a transcriptional reprogramming in mTEClo cells to induce cells suitable for orchestrating T cell development is therefore novel and interesting. The identification of a propensity to multi-organ immunity adds further to the impact of the work.

The authors are to be commended in applying the analysis to two different mouse models of disrupting CD4 cells ((deltaCD4 and deltaMHCII). However, both are indirect means of removing CD4 cells, and presumably leave open the possibility of additional non-CD4-related disruptions.

A weakness of the work is that, given the complexity of the two-way interaction between CD4 cells and mTEC, some of the experimental interventions are somewhat blunt, leading to conclusions that are not well supported by the results. For instance, the differences in NFkB signalling described in Fig 1 are measured on total mTEClo cells from 'deltaCD4 mice. Given that the premise of the study is heterogeneity in mTEClo cells, it seems important to address whether these differences relate to the differences in representation of the different mTEClo populations (which might exhibit different NfkB signalling) before inferring a direct effect on signalling. Similarly, since the knockout was directed to the mTEC, it is not clear that the phenotype relates to CD4 deficiency. Thus, the phenotype might well be influenced by subtle changes in mTEC composition rather than direct effects on signalling.

More generally, the single cell analysis that forms the major part of the manuscript is difficult to interpret given the context - that dynamic changes in the differentiation state of this heterogeneous population of cells is likely to lead to differences in gene expression states, but the 'snapshot' analyses inherent to this single cell analysis does not allow for dissection of cause and effect. For instance, Figs 1- 3 convincingly demonstrate that the mTEC composition is different in the different mice, and that signalling and transcription is different in the mTEClo precursors. Demonstration of a functional connection between these two observations would add substantially to these findings.

Reviewer #2 (Public Review):

The study by Lopes et al focuses on the role of thymocyte-epithelial cell cross-talk in the thymus and aims to determine the role of thymocyte derived signals in the differentiation of thymic epithelial cells. The study uses three different knockout models in which the thymocyte derived signals are defective and studies the resulting effect on mTEC maturation. The study suggests that these signals indeed play a crucial role in mTEc maturation and proposes a novel mechanism by which the developing T-cells direct the functionality of the thymic stromal compartment.<br> The study is mostly well designed and performed and the manuscript well written. Although the conclusions are largely based on substantial scientific evidence few points should be addressed in order to make the message of the study more precise and clearer:

1) According to the recent scRNA sequencing studies (reviewed in Kadouri et al 2020), the mTEClow mTECs contain at least two distinct subpopulations: the functionally mature CCL21-producing mTEC I and the immature mTECs giving rise to mTEC II and III. In its current form, however, the manuscript largely ignores the presence of mTEC I cells. The authors should make effort to analyze the changes in this population in the knockout models (by sequencing and/or qPCR) and cover this population also in the introduction and discussion.

2) As together with the classical mTEC classification (mTEChigh etc), the new scRNAseq based classification of mTECs (mTEC I etc) is used more and more often, it would be helpful to give in parallel the names of the subpopulations according to both of these classifications, at least when different mTECs are described/introduced in the beginning of the manuscript.

3) In Figures 2 and 4 the authors show data only on selected chemokines, cytokines and adhesion molecules and make a conclusion suggesting that these groups of proteins are down-regulated. To make this conclusion, however, the authors should analyze/show the whole groups i.e. all chemokines, cytokines and adhesion molecules (as they do for TRAs). Alternatively, the authors should be more careful/specific with their conclusions. The same is true for HDAC3 regulated transcriptional regulators and transcription factors.

Reviewer #1 (Public Review):

Summary:

In this work, the authors develop a tool for personalising prostate cancer treatment using a Boolean model. The model is extremely complex and describes the regulation of invasion, igration, cell cycle, apoptosis, androgen and growth factors signalling in prostate cancer using 133 nodes (genes and our metrics) and 449 edges (regulation pathways. Using their model, they were able to grade the effect of combined treatments for each of the 488 patients for already-developed drugs and find several genes suitable for intervention in most of the 488 patients. The predications from their model could help develop a patient-tailored treatment that could boost success of pancreatic cancer treatments in clinical practice.

Strengths:

The authors clearly achieved their aims of predicative prostate cancer modelling and have added value to the field of prostate cancer personalisation.<br> Calibrating and then validating predications of a model, as this work does, is a fundamental part of systems biology and mathematical modelling. By using a cell line to investigate predictions that AKT is the top hit for prostate cancer, validates the utility of their model and also shines a light on how useful models like this can be in oncology. The methodology in this paper provides a guide for future modelling work in this area.<br> Providing a detailed Supplementary Information and additional links to the code and fundamental modelling platform publications, helps to provide readers with a tool that may be applied in other settings. However, while this is a strength of the publication, the model is extremely complex and relies heavily on readers spending time comprehending pre-published work and doesn't provide a single contained body of work.<br> The methodology they are presenting could have significant impact on the field of cancer treatment, but would need to be testing clinically to validate that personalising treatment in this manner does improve outcomes.

Weaknesses:

While it is a strength of this work that such a detailed, and complex model is developed for prostate cancer, and that the code is provided, the weakness of this work is that the model is not easily accessible, and a lot of the techniques used in model development feel brushed over. The work relies heavily on other works and does not provide detailed descriptions of the underlying algorithm, requiring readers to absorb knowledge from our places This could be a challenge if an experimentalist wishing to implement this methodology in a different cancer treatment.<br> The protein/genes in the model are not presented in a way that it can be easily validated as such, the complexity of such a Boolean model comes into question. How sure are we in the model predications and are there are any potential weaknesses to modelling the network in such an extensive manner? For such a model like this, it is crucial to demonstrate its sensitivity to initial conditions and node additions/removals so some work could be done to demonstrate this so that the readers have an idea of how many over/under predications there might be in the model.<br> As they test so many drugs and combination regimes it is also hard to extract information about which key drugs should be repurposed. It could be useful to the readers to have this spotlighted more in the model so that it is easily discernable.

Suggestions:

Another way to validate the cohort level predications could have been to examine the efficacy of the predicted personalised protocols, or sensitive parts of the Boolean network, in a new prostate cancer patient cohort. Do we see the same sensitive pathways if we examine a different cohort of prostate cancer patients?

Reviewer #2 (Public Review):

Montagud et al. present a very successful experiment - modeling feedback loop: the authors develop a Boolean model of the major signaling pathways deregulated in prostate cancer, use molecular data from patient samples to personalize this model, use drug response of cell lines to validate the model, predict 15 actionable interventions based on the model, and test nine of these interventions, confirming four.

The premise of the work is well-supported by prior work by the team and the wider community. The methods are sound, well integrated and thoroughly documented, with one notable omission. The process through which the logic functions of the nodes were determined/decided is not described. The Appendix file indicates "The model is completed by logical rules (or functions), which assign a target value to each node for each regulator level combination.". The interested reader would want to know what information is used and what considerations are the basis of these assignments, and what would change if an assignment were different.

The manuscript makes a number of testable predictions of actionable single and combinatorial therapeutic interventions for prostate cancer. Equally important, the combination of information and methodologies used in this paper offers a roadmap for future development of predictive and personalized models. Such models are much needed in precision oncology.

Reviewer #3 (Public Review):

This paper tries to establish a model for drug (and combination) selection for individual prostate cancer patient based on a prior signal network knowledge base and genomic/transcriptomic profiling data. This is of great clinical potential. However, whether this approach could be robustly applied in clinic is not validated. Limited validation using cell line is provided. Most tumors have complex structure including tumor cells and surrounding microenvironment. The model is mainly built from onco-signaling pathways. The contribution of microenvironment including immunity is unclear.

Reviewer #1 (Public Review):

Strengths: The manuscript is very thorough and convincingly homes in on key circuit elements and mechanisms that likely underpin this unexpected linearity in the On-parasol circuit.

Weaknesses: perhaps this is just me, but I found the MS is quite "hard work" to parse. To some extent I suspect this just is what it is, given the complexity of the circuit, but perhaps the occasional explanatory statement and further streamlining of the figures might help get the key points across a little more easily. I was also a little unsure about what exactly each figure and statistical measure depicted, something that presumably is easily fixed by expanding the legends accordingly. I offer some specific suggestions in the author section.

Finally, the manuscript rather glances over "the middle" of the circuit, i.e. the bipolar cells, and their computations. The circuit insights gained from the study do not require spatially nonlinear computations in bipolar cells (e.g. via amacrine cells), rather, they seem to require the bipolar cells to be linear in this regard. However bipolar cells, like cones or ganglion cells, also have many sources of nonlinearities, some well understood and probably several others less well understood. Is it then not surprising that they should not play a role in the circuit as well? It would be good to see some discussion/acknowledgement of this topic.

Reviewer #2 (Public Review):

The manuscript by Yu, Turner, Baudin, and Rieke investigates spatial integration of visual information by On parasol cells in the primate (macaque) retina. Having previously shown that these cells integrate natural stimuli linearly over space, but show characteristics of nonlinear spatial integration when probed with standard reversing gratings, the authors here study the mechanisms behind this surprising stimulus-dependent difference of spatial integration characteristics. Using patch-clamp recordings of synaptic inputs into On parasol cells as well as computational modeling, they reveal that crossover inhibitory inputs, which are elicited by Off-type stimulation, linearizes spatial integration and that this crossover inhibition is strongly triggered by natural stimuli as well as at the onset of a spatial grating stimulus, but is reduced during ongoing stimulation with reversing gratings, leading to nonlinear integration. The authors further show that this shift in crossover inhibition may follow from adaptation dynamics in the cone photoreceptors, which shift the balance between excitation triggered by brightening and inhibition triggered by darkening. Experiments with modified grating stimuli, designed to mimic photoreceptor signals without adaptation, confirm that subtle asymmetries in the signaling of bright and dark stimulus parts can fundamentally alter the integration characteristics of these retinal ganglion cells.

This is an excellent manuscript, which combines state-of-the-art experiments with sophisticated modeling and data analysis to tackle an important question about signal processing in the retina. It provides a convincing line of reasoning, from clear characterizations of linear and nonlinear stimulus integration under different stimulus contexts via a clever evaluation of the relative contributions of feedforward and crossover inhibition to the model-based evaluation of adaptation dynamics in cone photoreceptors. The idea that photoreceptor adaptation evokes a systematic shift in balance between excitation and crossover inhibition under reversing gratings, which represent a widely used standard stimulus, is thought-provoking, and the clever use of a computational model to transform the visual stimulus for probing retinal processing without photoreceptor adaptation is compelling. My only major comment is that the manuscript provides relatively little information about the photoreceptor adaptation model. Explaining how the adaptation dynamics relate to the asymmetry in the responses and how the non-adapting, linear version of the model compares to the full model, would help understand why the linear cones create a scenario that's more similar to the steady-state response under reversing gratings than to the onset response.

Reviewer #3 (Public Review):

This manuscript reports an interesting result regarding retinal ganglion cells in primates, although the presentation of the main findings was slightly confusing. There is a classic distinction between ganglion cells that integrate linearly vs. nonlinearly over space. The primary test has been the presentation of a high spatial frequency contrast-reversing grating, which generates no response in a linear cell - because the responses to bright and dark bars cancel each other. A nonlinear cell would instead respond at twice the frequency of reversal, because (e.g., for an ON cell) the increased response to the bright bars on each phase of the grating cannot be canceled by the decreased response to the dark bars. The explanation has been a nonlinearity at the output of bipolar cells, such that increases in glutamate release to the preferred contrast cannot be canceled by decreases in release to the non-preferred contrast, either because the release rate at baseline is very low (i.e., rectified) or else there is some difference in the dynamics to increases vs. decreases in release even when the output is not strongly rectified. This latter idea is illustrated in Borghuis et al. (2013; Fig. 7) - which the authors might consider citing.

The authors report that On parasol cells in primate show the nonlinear behavior to a high contrast grating. Although - and this an important detail - there is no such response at the first half-cycle of the grating (Fig. 5C). Likewise, there is no response to a grating flashed briefly from a gray background (Fig. 1C). There could be an advantage to presenting these results together at the front of the paper - indeed they seem to show results from the same cell (at least, Fig. 1C and 5A spike traces look identical).

The main section of the paper shows that natural images presented either in sequence or flashed briefly from a gray background lack nonlinear response behavior - that is, a linear receptive field seems to be encoding the image, because structure in the image can be replaced by the mean luminance across the receptive field center. This makes natural images and gratings seem like they bring out truly unique behaviors in the ganglion cell. But what could be so special about a grating?

It seems, simply, that the natural image dataset did not match the high contrast grating in the detailed properties of the stimulus, and the proposed nonlinearity in the cone is also obviously important. For example, when the grating is first presented in the contrast-reversing stimulus sequence (i.e, the first half-cycle) or the grating is simply flashed briefly, the On parasol cell is excited by the presynaptic On bipolar cells that see a gray-to-white transition. But the release is relatively weak, because the cones are partly adapted to the gray background. The inhibition driven by 'crossover' circuitry (i.e., driven by the Off pathway) is relatively strong at the gray-to-black transition, and the responses cancel.

What is less obvious is that for subsequent cycles of the grating, the On bipolar cells that see black-to-white transitions are now releasing very strongly (because the cone response is relatively strong coming from an un-adapted state), which can no longer be canceled by the Off-pathway inhibition. The grating is the optimal stimulus to reveal this nonlinearity because there are simultaneous full-contrast changes in luminance, in opposite directions, at different points across the image. There is no reason why a similar sequence in a natural movie would not drive a similarly strong nonlinear response. It just happens that the natural images and movies (i.e., image sequences) used here apparently never contain these patterns of light changes across the receptive field center.

So there is a fundamental feature of the circuitry that determines the degree of nonlinear spatial summation in On parasol cells (in an interesting way that differs from Off parasol cells). My main recommendation is that the authors present the findings with gratings in a more compact fashion rather than alternating between gratings and natural images. For example, the results in FIg. 1B and C are puzzling until one learns of the result in Fig. 5A and realizes that Fig. 1B is showing the average across many cycles, which does not capture the unusual response in the first half cycle that is completely consistent with the result in Fig. 1C.

There is also a clear difference in the contrast of the grating vs. the contrasts within the natural scenes (which are more difficult to define, perhaps). It could be interesting to explore the responses to lower contrast gratings to see if a single model can explain both grating and natural image responses over a range of contrasts in a satisfying way.

Reviewer #1 (Public Review):

This manuscript describes the finding that the voltage-dependence and kinetics of one component of K+ current in parvalbumin positive cortical interneurons in a mouse model of Alzheimer's disease (5xFAD) is different from that in the same neurons of wild type mice. Using computational modeling and dynamic clamp recordings, the work supports the concept that this alteration in K+ current reduces the firing rate of these inhibitory interneurons in response to weak stimulating currents, and that this may lead to the overall hyperexcitability of cortical circuits in the 5xFAD model. The work then goes to make two major claims about the nature of the change in K+ current: i) that the change in voltage-dependence and kinetics does not involve any change in the expression of K+ channels and ii) that it represents a posttranslational modification of Kv3 channels that alters their biophysical properties. Overall, the work is clear and interesting but more would be required to substantiate these two claims.

1) The possibility that different levels of channel subunits or their auxiliary subunits exist in the wild type mice and the 5xFAD mice has not been ruled out.

2) It has not been established completely that the component of current that differs between the two strains of mice solely represents the activity of Kv3 channels.

3) While many posttranslational mechanisms linked to cellular signaling pathway are known to modify Kv3 channels, and some of them could account for the observed differences, none of these are tested or discussed in the manuscript.

Reviewer #2 (Public Review):

In this study, the authors used a viral-tagging method to compare in the somatosensory cortex, the excitability of parvalbumin (PV) interneuron of young adult 5xFAD mice (an early-stage Alzheimer's disease [AD] mouse model) and WT mice. In current clamp recording from cortical slices, they found that PV interneurons from 5xFAD mice displayed strongly dampened spike discharge near-threshold and modified action potential (AP) waveforms. Extensive examination of several AP firing parameters, computational modeling, and PV-specific qPCR experiments indicated that changing Na+ channel availability was not responsible for the changes in AP firing. Outside-out patch-clamp recording and quantitative qPCR revealed alterations in the voltage dependence of Kv3 K+ channel activation and kinetics in AD mice, without significant changes in K+ channel gene expression. Using dynamic clamp and PV interneuron modeling, they found that the left-shift in Kv3 channel activation could recapitulate the observed firing phenotypes in 5xFAD mice with a near-threshold hypoexcitabilty of PV interneurons. Kv3 modulation reduces synaptically-evoked AP firing in PV interneurons. Using a reduced cortical network model, the authors showed that the changes in PV interneuron firing induced by Kv3 channel biophysical alterations were sufficient to produce gamma-frequency specific network hyperexcitability.

Strengths:<br> The work deciphers in detail the intrinsic excitability properties of PV interneurons, which allowed the identification of an interesting difference in the biophysical properties of the predominant Kv3 current of AD mice (5xFAD) as compared to WT mice. Exploiting the state-of-the-art dynamic clamp technique and cortical network modeling, the authors could elegantly reproduce the firing features of AD mice and examine their consequences to cortical network hyperexcitability. This work represent an interesting cellular mechanism with a causal link to overall circuit hyperexcitability, with a potential therapeutic approach to combat AD progression in its early stages.

Weaknesses:<br> Although the work exhibits the strength principles, it suffers from insufficient and sometimes inappropriate analyses that are necessary to fully support the key claims of the manuscript. In addition, plausible alternative explanations related to the firing features of AD mice and their cellular mechanism need to be considered. Therefore, some aspects of the experiments and data analyses need to be extended and clarified.

Reviewer #1 (Public Review): 

The authors address the influence of genetic variation, captured as genome wide SNPs on multiple parameters of the T cell receptor repertoire, including differential variable gene usage, base pair deletions (trmming) and base pair additions at the VDJ junctions. 

The authors uncover signficant genetic associations between multiple SNPs in the T cell beta locus and the MHC on variable gene usage. They develop a model which conditions on gene usage to identify SNPs within the gene coding Artemis protein which associate with deletions; and other SNPs in the TdT gene which associate with additions. The authors attempt a validation cohort, although because of the lack of SNP overlap the validation observed is very limited. 

The study contributes to our understanding of the genetic control of the T cell receptor repertoire and its diversity. However, the advance in understanding of the process of repertoire creation is quite modest. Genetic effects on variable gene expression have been seen before. Genetic effects on basepair deletion and addition is new.

Reviewer #2 (Public Review): 

In this study, Russell et al. combined T cell receptor (TCR) repertoire sequencing data with SNP array genotype data to infer genetic polymorphisms which impact upon the process of TCR generation. Using these data, the authors looked for loci with polymorphisms which associate with different V(D)J recombination probabilities, i.e. sites in the genome which impact upon the chances of TCRs with different properties being produced when they change. 

Beyond the expected sites in the TCR and MHC loci, the authors observed strong associations with distant sites. One was with DCLRE1C, which encodes Artemis, the endonuclease responsible for cutting the TCR loci during recombination, while the second was DNTT, the site encoding the enzyme Tdt, which is responsible for addition of nucleotides to cut V(D)J during recombination. This is the first time that such SNP associations have been described to my knowledge, and yet make perfect sense: DCLRE1C variations were associated with the amount of trimming V and J genes underwent during recombination, while DNTT polymorphisms associated with the number of inserted nucleotides. The authors also report, after assigning donors an associated ancestry based on clustering of their genotype data, that certain inferred ancestries associate with different TCR repertoire properties. In this analysis 'Asian-associated' TCR repertoires had fewer non-templated nucleotide insertions, along with a corresponding greater incidence of the DNTT polymorphisms associated with differences in insertions, relative to other groups. 

Strengths: 

This manuscript is exceedingly well written. Both the TCR biology and the statistical considerations of the genetic analyses are extremely complex topics, mired in arcane terminology, which often end up somewhat impenetrable to non-expert readers. However both have been introduced and explained with admirable clarity throughout, including the caveats and implications of analyses that would not be intuitive to many readers not already expert in both fields. 

As best I can determine, the analyses themselves are also extremely rigorous, with each step carefully taken and justified in the text, involving numerous corrections at multiple scales (e.g. for TCR productivity, TCR gene usage, specific TRDB2 genotype, population substructure, and more). The major findings have also been validated in a completely separate cohort, using a different analysis pipeline. While the authors point out that such genome-wide association efforts looking at TCR gene expression have been undertaken before, the major innovation presented here lies in applying those data to investigating specific V(D)J recombination probabilities. Thus the findings are novel, and the conclusions well supported by the data. 

The data visualisation have all been plotted in a sensible and easily interpretable manner. The majority of data themselves are all already publicly available, having been published in prior studies. The TCRseq data for the validation code has been assigned a BioProject accession, which I presume will go live at the time of publication. The code is also appropriately hosted on Github, and are mostly adequately commented and documented enough so as to be repeatable. 

Limitations: 

There are very few if any obvious technical limitations or weaknesses that I can see that are not intrinsic to the data themselves. While the authors do mention these limitations, I wonder if they should be devoted some more attention somewhere in the text of the manuscript; relatively few researchers are expert in both TCR biology and the technicalities of genome-wide association studies, so I think more explicit consideration of these issues would be helpful. 

In particular, I think the difficulty of studying these loci with standard techniques could be underlined, along with what implications that might have for this study. The highly repetitive nature of the TCR loci can certainly make any analysis looking at short sequences problematic, which has implications for both the TCRseq and genotyping aspects of this study. Combined with the fact that most studies focus on certain populations, polymorphisms in the TCR loci are very likely being relatively undersampled by the field (a hypothesis supported by the ongoing discovery of novel exonic polymorphisms in TCRseq data itself, e.g. as demonstrated in this pre-print by Omer et al. https://doi.org/10.1101/2021.05.17.444409). The consequences of SNP polymorphism coverage in SNP arrays has already been considered for IgH (https://doi.org/10.1038/gene.2012.12): while this is an admittedly more polymorphic locus, the underlying causes of these issues are mostly all true of the TCR loci as well. Similarly, while the authors do appropriately point out that issues with V(D)J gene assignment could infer biases it may be worth noting that the TCRseq technology used to produce their main dataset uses relatively short read sequencing, that is unable to distinguish a substantial fraction of even known TCR gene- and allele-level diversity (see Fig. 1C of the Omer et al. pre-print). Thus there may be a whole dimension of TCR polymorphism that is not well captured by either platform. 

Overall, I think this is an extremely considered and digestible study, which will be of great interest across and beyond the field. As the wider community comes to grips with how best to incorporate TCR and BCR polymorphisms into their analyses of the adaptive immune loci themselves (and how this might impact upon recombination, expression, and downstream immune functions) this serves as a timely reminder that we should not forget the polymorphisms elsewhere in the genome that might also be relevant.

Reviewer #3 (Public Review): 

In this manuscript, Russel et al propose an inference method to link genetic variations with TCR repertoire feature variations, based on observations from previous studies showing similarities at various level of the repertoire in monozygotic twins. To that end, they used a unique publically available dataset, which combines TCRb immunosequencing data as well as whole genome SNPs data. The method is elegant and sheds light on the importance of combining different type of data to better understand the complexity of TCR repertoire generation and selection. However, unfortunately, while their discovery data set provide some associations between SNPs and TCR repertoire features, they were almost unable to recapitulate the results with their validation dataset. The main reasons could be that the donor demographics are highly divergent between the two cohorts (81% Caucasian in the discovery vs. mainly Hispanic in the validation), the immunosequencing data were generated using RNA based method for the validation while the discovery dataset was obtained from gDNA templates and finally the SNPs array were discordant between the two datasets. Nonetheless, the approach and the study deserve attention and might be improved by additional experiments or analyses and by providing additional information.

Reviewer #1 (Public Review): 

This manuscript describes the structural and functional analysis of a discrete clade of magnesium transporters related to the well-studied NRAMP divalent metal ion transporters. The structural approach - cryo-EM for high resolution structure determination coupled with crystallography to precisely identify ion binding sites via anomalous scattering - is excellent, combining the different strengths of the two techniques. The structures clearly represent the culmination of a lot of tenacious biochemical work to achieve a structurally tractable, nanobody-stabilized transporter. 

The structures are coupled with thoughtful and extensive substrate transport data to investigate the effects of various mutants on substrate transport. The authors make the surprising observation that Mg2+ transport is not coupled to proton movement. However, they make a convincing case that, in the physiological context, the electrical and chemical gradients are such that proton-coupling is not energetically required, and structural observations corroborate this interpretation. Although the functional experiments were mostly thorough and well-controlled, it appears that different membrane potentials were applied in proton- and metal-uptake experiments, and the functional insights could be augmented by considering how these potentials might influence transport and controlling for the membrane potential. 

Together, these data complement existing functional and structural observations for other members of the SLC11 family with different substrate specificities, leading to a comprehensive mechanistic picture of divalent metal ion binding and selectivity among this family of transporters.

Reviewer #2 (Public Review): 

There are two important findings in this manuscript. First, the authors discussed the unique ion selectivity property of prokaryotic NRAMPs for Mg2+ and proposed that Mg2+ would be recognized in a nearly fully-hydrated form. Second, based on the proteoliposome experiments, the authors show that EleNRMT is not driven by proton and maybe more like a uniporter. Based on the structural comparison, they proposed the replacement of His with Trp may be crucial for this difference. I basically agree with these views and have several comments as shown below. 

1) I basically agree that EleNRMT would recognize Mg2+ ion in a nearly fully-hydrated form, but to conclude it, the resolution of the EleNRMT structures is a kind of moderate. I think extra efforts, such as the MD simulation, would be too much work and not necessary, but it would be helpful if the authors can mention the limitation of the resolution of the present structures, where water molecules are not clearly visible, somewhere in the manuscript. 

2) As this is a great success of the thermostabilization of membrane proteins, it would be helpful to other people in this research field if the authors provide a more detailed protocol for the thermal shift assays (Figure 3C).

Reviewer #3 (Public Review): 

The paper begins with a robust set of transport assays comparing the ion specificity of an E. coli NRAMP (EcoDMT) with an NRMT from the bacterium Eggerthella lenta (EleNRMT). These assays confirm Mn transport for both proteins, though EleNRMT has 10-fold lower affinity and a notable competition with Mg. Mutation of active site residues previously proposed to account for these differences fail to convert EcoDMT into a Mg transporter, although these mutations do allow Ca transport. This result provides a strong premise for structural studies of NRMT. 

EleNRMT was chosen from a broad screen of NRMT homologs and was then optimized for structural studies by the introduction of 3 thermostabilizing mutations. The authors have been careful to document that thermostabilization has not significantly altered Mn transport by the protein. They document the expected Mg transport activity together with a lack of Ca transport. Thus, despite having the same 4 active site residues as the mutated EcoDMT, EleNRMT has distinct transport properties which call for detailed structural studies of the ion binding site. 

To facilitate cryo-EM on such a small target (~50kD), nanobodies were raised by immunization of alpacas followed by phage-display selection. They settled on imaging a remarkable trimeric complex composed of NRMT and two of the nanobody binders, which led not only to two cryo-EM structures in the presence and absence of Mg, but also to anomalous X-ray scattering data from modestly ordered crystals. The cryo-EM structures define the architecture of the protein, though the modest resolution of the Mg-bound structure seems to limit definition of the site. Nevertheless, anomalous X-ray scattering definitively confirms the location of the ion binding site. The molecule adopts an inward-facing conformation of NRMT that is very similar to previous work on NRAMPs. However upon closer inspection, the authors describe an expanded ion binding site that is able to accommodate a hydrated Mg ion with a coordination shell that clearly differs from the dehydrated Mn ion seen in NRAMPs. Although it is not explicitly discussed, the transport data imply that the binding site of EleNRMT is flexible enough to adopt either configuration depending of the ligand it encounters. 

Armed with this structure, the authors introduce mutations to several key residues within the ion binding pocket. The results from transport assays confirm the authors' ideas about Mg selectivity and ultimately bring the paper full circle by generating the Ca permeability as seen in the mutated EcoDMT protein. 

Although the functional data clearly show the lack of proton coupling in EleNRMT, the structural explanation is not so convincing. It is not clear that the proposed proton pathway for NRAMPS is so specific that the authors conclusion about the replacement of His228 with a Trp is clear cut. This issue seems complex and deserves further discussion and perhaps mutational data to support the critical importance of this His residue. 

However, taken together, the extensive set of results presented in this paper provide a robust and comprehensive explanation of ion selectivity amongst the broader SCL11/NRAMP family.

Reviewer #1 (Public Review):

The authors employed a causal inference method known as Mendelian randomization (MR) to evaluate potential causal relationships between physical activities and obesity.

The main strength is that MR can provide evidence for causal relationships as it is less susceptible to reverse causality and confounding. The authors also employed a recently developed technique (CAUSE) that can leverage a very large number of genetic variants with increase in power.

There are no major flaws in the methodology as far as I observe. One limitation is that MR provides an effect size based on lifelong (from birth) exposure of the risk factor and may not reflect effects of increased physical exercise for a limited period of time (eg 3 months or 1 year). The findings are useful but may not be very novel or surprising given that many previous epidemiological studies also reach the same conclusion (that physical exercise help to reduce weight). Nevertheless, overall speaking, the methodology is sound and the results are supported by the analysis.

This work is clinically relevant as it provides further proof that physical exercise is causally related to lower BMI, while previous studies are not explicitly designed for causal inference. The study also provides estimates of (causal) effect sizes although the estimates may be larger than those from short-term interventions, as MR estimates reflect longer-term exposure.

Reviewer #2 (Public Review):

The manuscript by Carrasquilla and colleagues applied Mendelian Randomization (MR) techniques to study causal relationship of physical activity and obesity. Their results support the causal effects of physical activity on obesity, and bi-directional causal effects of sedentary time and obesity. One strength of this work is the use of CAUSE, a recently developed MR method that is robust to common violations of MR assumptions. The conclusion reached could potentially have a large impact on an important public health problem.

Major comments:

(1) While the effect of physical activity on obesity is in line with earlier studies, the finding that BMI has a causal effect on sedendary time is somewhat unexpected. In particular, the authors found this effect only with CAUSE, but the evidence from other MR methods do not reach statistical significance cutoff. The strength of CAUSE is more about the control of false positive, instead of high power. In general, the power of CAUSE is lower than the simple IVW method. This is also the case in this setting, of high power of exposure (BMI) but lower power of outcome (sedentary time) - see Fig. 2B of the CAUSE paper.

It does not necessarily mean that the results are wrong. It's possible for example, by better modeling pleiotropic effects, CAUSE better captures the causal effects and have higher power. Nevertheless, it would be helpful to better understand why CAUSE gives high statistical significance while others not. Two suggestions here:

(a) It is useful to visualize the MR analysis with scatter plot of the effect sizes of variants on the exposure (BMI) and outcome (sedentary time). In the plot, the variants can be colored by their contribution to the CAUSE statistics, see Fig. 4 of the CAUSE paper. This plot would help show, for example, whether there are outlier variants; or whether the results are largely driven by just a small number of variants.

(b) CAUSE is susceptible to false positives when the value of q, a measure of the proportion of shared variants, is high. The authors stated that q is about 0.2, which is pretty small. However, it is unclear if this is q under the causal model or the sharing model. If q is small under the sharing model, the result would be quite convincing. This needs to be clarified.

(2) Given the concern above, it may be helpful to strengthen the results using additional strategy. Note that the biggest worry with BMI-sedantary time relation is that the two traits are both affected by an unobserved heritable factor. This hidden factor likely affects some behavior component, so most likely act through the brain. On the other hand, BMI may involve multiple tissue types, e.g. adipose. So the idea is: suppose we can partition BMI variants into different tissues, those acted via brain or via adipose, say; then we can test MR using only BMI variants in a certain tissue. If there is a causal effect of BMI on sedentary time, we expect to see similar results from MR with different tissues. If the two are affected by the hidden factor, then the MR analysis using BMI variants acted in adipose would not show significant results.

While I think this strategy is feasible conceptually, I realize that it may be difficult to implement. BMI heritability were found to be primarily enriched in brain regulatory elements [PMID:29632380], so even if there are other tissue components, their contribution may be small. One paper does report that BMI is enriched in CD19 cells [PMID: 28892062], though. A second challenge is to figure out the tissue of origin of GWAS variants. This probably require fine-mapping analysis to pinpoint causal variants, and overlap with tissue-specific enhancer maps, not a small task. So I'd strongly encourage the authors to pursue some analysis along this line, but it would be understandable if the results of this anlysis are negative.

Minor comments

- The term "causally associated" are confusing, e.g. in l32. If it's causal, then use the term "causal".

Reviewer #3 (Public Review):

Given previous reports of an observational relationship between physical inactivity and obesity, Carrasquilla and colleagues aimed to investigate the causal relationship between these traits and establish the direction of effect using Mendelian Randomization. In doing so, the authors report strong evidence of a bidirectional causal relationship between sedentary time and BMI, where genetic liability for longer sedentary time increases BMI, and genetic liability for higher BMI causally increases sedentary time. The authors also give evidence of higher moderate and vigorous physical activity causally reducing BMI. However they do note that in the reverse direction there was evidence of horizontal pleiotropy where higher BMI causally influences lower levels of physical activity through alternative pathways.

The authors have used a number of methods to investigate and address potential limiting factors of the study. A major strength of the study is the use of the CAUSE method. This allowed the authors to investigate all exposures of interest, in spite of a low number of suitable genetic instruments (associated SNPs with P-value < 5E-08) being available, which may not have been possible with the use of the more conventional MR methods alone. The authors were also able to overcome sample overlap with this method, and hence obtain strong causal estimates for the study. The authors have compared causal estimates obtained from other MR methods including IVW, MR Egger, the weighted median and weighted mode methods. In doing so, they were able to demonstrate consistent directions of effects for most causal estimates when comparing with those obtained from the CAUSE method. This helps to increase confidence in the results obtained and supports the conclusions made.<br> This study is limited in the fact that the findings are not generalizable across different age-groups or populations - although the authors do state that similar results have been found in childhood studies. As the authors also make reference to, due to the nature of the BMI genetic instruments used, the findings of this study can only inform on the lifetime impact of higher BMI, and not the effect of a short-term intervention.

The findings of this study will be of interest to those in the field of public health, and support current guidelines for the management of obesity.

Reviewer #1 (Public Review):

The authors explore the effects of progressive depolarization of the resting potential upon excitation contraction coupling seeking to attribute its failure to either failure of action potential generation or of its subsequent coupling to Ca2+ release. The authors report a persistence of action potential generation in the face of a failure of Ca2+ release. The observation is empirically a helpful one in itself, but the further interpretation in the paper must address a number of issues require resolution bearing upon (a) whether there is a possible tubular action potential conduction failure despite persistent surface membrane excitation and (b) the relative inactivation properties of Na channel function and of excitation contraction coupling.

Reviewer #2 (Public Review):

Summary

The objective of this study is to further understand the mechanism by which increases in extracellular [K+] affect action potential, calcium transient and twitch force. They show that the decreases in action potential overshoot and twitch force occur over a very narrow range of resting potential as others have demonstrated. What is new and interesting are the relationship between resting potential, action potential and peak calcium during twitch contractions. They demonstrated the gradual decrease in calcium peak over a narrow range of resting potential and a linear relationship between action potential area and peak calcium.

Strengths

The strongest aspect of this study is about the relationship between resting potential, action potential and calcium transient. This is the first study demonstrating such relationships and this was quite overdue.

Weaknesses

There is a misunderstanding about the all or none concept. The authors argued that some researchers had proposed that the potassium effects on force was an all or none effect, which is not true. It is true that the range of potassium concentrations and resting potentials over which the decreases in twitch forces is very narrow, it does not implicate an all or none concept because the decreases are steep but still gradual. The all or non concept for action potential does not implicate that the action potential shape is constant in all physiological conditions; instead for a given physiological condition the action potential shape is constant but changes between physiological condition.

There is a major lack of information about the calcium indicator that has been used for this study. Consequently, it is difficult to validate the relationship between peak calcium and resting potential or overshoot relationship. It is not clear whether the indicator get saturated with calcium and there is a major issue about a short twitch contraction in msec and a calcium transient lasting minutes.

Finally the experimental temperature was 25oC when during exercise muscle temperature often exceeds 37oC. So, it will not be possible to use any of the relationships provided in this study to understand the role of potassium in fatigue because the potassium effects on all three parameters are very temperature sensitive.

Impact of the study

Although the data about the effects on twitch force is nothing new, the data for the relationship between membrane potential and calcium peak is the first of his kind and extremely important for our understanding of the potassium effect on contraction and its role during fatigue.

Reviewer #3 (Public Review):

This paper by Wang et al., presents an elegant approach to study the role of depolarization of the resting membrane potential for excitation-contraction-coupling in skeletal muscle with relevance for basic physiology and muscle disease. The study aims to provide a more detailed understanding of how depolarization affects skeletal muscle fiber excitability and action potential waveform, and how this in turn reduces Ca2+ release from the sarcoplasmic reticulum eventually leading to the loss of force that is associated with depolarization.

Methodologically, two intracellular electrodes are inserted into EDL muscle fibers from mice. The electrodes are used to trigger and record single fiber action potentials in conjunction with recordings of intracellular Ca2+ transients reported by genetically encoded Ca2+ sensing GCAMP6f. Muscle movement is reduced using an inhibitor of the contractile filaments.

The study clearly shows that depolarizing muscle with elevation of extracellular K+ from 3.5 to 10 mM causes increased twitch force. When further elevating extracellular K+, the twitch force declines and at 14 - 16 mM K+ the twitch force is largely abolished. The increase in force at moderately elevated K+ (10 mM) remains unclear especially because it is associated with clear reduction in action potential peak. At high extracellular K+ (12-16 mM) the action potentials become further compromised. In a separate series of experiments, muscle fibers are excited to trigger action potentials every 5 seconds for 7 mins either at 3.5 mM K+ throughout or initially at 3.5 mM K+ but during the 7 minutes a sudden elevation of extracellular K+ to 16 mM is imposed. This series of experiments is productive because it gives both action potential wave form characteristics and Ca2+ transients from the same muscle fibers as the resting membrane potential changes throughout the experiment. The approach therefore provides co-temporal measurements of action potentials and Ca2+ transients that enable correlations to be made between action potential waveform characteristics and the amplitude of the Ca2+ transients.

A fundamental and important finding of the study is that Ca2+ release is highly dependent on the action potential shape showing that changes in action potential waveform such as reduced amplitude and altered width of the action potential are decisive of Ca2+ release. The authors analyze this in more details and arrives at the conclusion that the integral of action potential waveform more depolarized than -30 mV is the best predictor of Ca2+ release.

The data is discussed in relation to loss of force in depolarized muscles, and the authors challenge an existing notion of loss of force in depolarized muscle being caused by all-or-none excitation failure. This notion stems from the close correlation between force and propagating compound muscle action potentials in isolated muscle at elevated extracellular K+ that has been reported in other studies. The present study shows that changes in action potential waveform with reduced waveform integral more depolarized above - 30 mV could reduce Ca2+ release and force without representing a complete loss of excitation. This is then taken to a discussion on how to best define an action potential.

Some considerations regarding the methodology deserve to be mentioned. The electrodes that are inserted in the muscle fibers appear to be at very close proximity and hence the recordings would not be propagating action potentials but rather action potentials or local responses around the recording electrode. Similarly, the Ca2+ transients are also recorded in this same small part of the muscle fiber. On this basis it remains unclear whether the compromised action potentials at extracellular K+ above 10 mM would propagate along the entire muscle fiber to trigger contractions or merely be local responses at the site of the electrodes. Another methodological aspect to mention is that the sampling in the Ca2+ imaging appears slow relative to the rapid upstroke and spike of the Ca2+ transient. The authors acknowledge and discuss these limitations but still draw conclusions based on the data of the recorded amplitude of the Ca2+ transient. Additional experiments that measures propagating action potentials and Ca2+ imaging/fluorescence systems with higher sampling rate and better suited for detecting peaks of Ca2+ transient will be needed to fully validate some of the rather strong conclusions by the authors.

Reviewer #1 (Public Review):

By recording the sleep and movement patterns of a group of baboons (a highly social animal), this study showed that ecological and social pressures affect individuals' sleep patterns. Results highlight the importance of studying animal sleep under natural conditions by using sophisticated electronic tags. The main strength of this paper is that the authors took a novel approach for recording GPS positions and sleep patterns (based on body acceleration) of a group-living species for relatively long periods. The datasets allowed them to examine several important aspects of animal sleep (e.g., the effect of changing sleeping sites and the presence of other group members, testing the sentinel hypothesis). A weakness is that they did not measure sleep directly; however, they are aware of this weakness and carefully discussed it. I had already provided several comments at pre-submission stage, and the authors revised the manuscript accordingly. Therefore, I only have relatively minor, specific comments.

Reviewer #2 (Public Review):

A fundamental aspect of sleep is that it is homeostatically regulated. Sleep homeostasis manifests along two dimensions, sleep duration and sleep intensity. Following sleep loss, animals can recover lost sleep by sleeping longer, sleeping deeper, or doing both. In many species, sleep duration can be measured behaviorally (e.g., the animal is immobile in a sleep posture with closed eyes) or electrophysiologically, via measuring brain activity (electroencephalogram, EEG) along with other physiological parameters that change between wakefulness and sleep. Sleep intensity is measured by determining the amount of stimulation needed to awaken an animal. As this method is disruptive to sleep, the quantity (spectral power) of slow waves in the EEG, which correlates with sleep intensity in mammals, is often used to measure sleep intensity.

Most of our understanding about sleep homeostasis is derived from controlled laboratory experiments where ecological pressures are minimized. However, little is known about how animals reconcile ecological pressures for wakefulness with the homeostatic need for sleep. In this paper, the authors used a neck mounted accelerometer and GPS device to characterize the sleep behavior of a troop of baboons living in the wild. Although EEG devices are now being used to record sleep in a variety of animals in the wild (i.e., sloths, sandpipers, owls, and frigatebirds), this method is not suitable for wild baboons, as they would likely attempt to remove the equipment. However, accelerometry can be used to estimate the time spent sleeping in the wild, once validated against either direct measures of sleep behavior (posture, eye state, immobility) or EEG measures of sleep. In the current paper, accelerometry was validated against behavioral measures of sleep obtained with thermal cameras; immobile baboons sitting with the head hung were considered to be asleep.

Using accelerometry, the authors investigated the impact that various ecological factors had on the time spent sleeping. Baboons slept more when spending the night in their preferred trees. They slept less when sleeping in larger groups, apparently due to awakenings of individuals disrupting the sleep of other group members. Indeed, the timing of bouts of sleep and wakefulness were more synchronous within groups sharing the same tree. The time spent sleeping on a given night was not influenced by how much time the baboons spent sleeping the previous night or how much time they were active during the day. As one might expect the baboons to recover lost sleep following a night with less sleep by sleeping longer, the authors suggest that ecological factors interfere with the homeostatic regulation of sleep. The lack of a relationship between the time spent traveling and sleep duration on the following night was also interpreted as evidence for ecological factors interfering with the homeostatic regulation of sleep.

The results support the authors' interpretation if one only considers sleep duration as a means for recovering lost sleep. However, effective sleep homeostasis could have been mediated by modulating sleep intensity. Baboons that lost some sleep on one night might recover this sleep by sleeping deeper, but not longer, on the subsequent night. Without measuring sleep intensity, it is not possible to have a full accounting of the amount of sleep obtained under the various ecological conditions.

Despite this limitation, the authors present a large and unique dataset on the ecology of sleep in highly social animals living in the wild. The finding that baboons sleep less when sleeping in trees with which they are less familiar, illustrates that safe sleep sites are an important resource for animals. The synchrony between bouts of wakefulness and sleep among group members is also interesting, as it suggests that the group does not maximize vigilance by sleeping and waking asynchronously, as some researchers had proposed. In addition, it suggests that a potential cost of sleeping in groups is disrupted sleep. As such, this study adds to our understanding of the ecology of sleep in socially sleeping animals.

Reviewer #1 (Public Review):

By presenting the detrimental effect of accumulative heterozygous mutations on the sperm head morphology, this report by Martinez and colleagues brings new attention to a wildly accepted paradigm in male germ cells that genetically haploid spermatids are phenotypically diploid, suggesting that multiple heterozygous mutations can lead to unexplained male infertility. The merit of this manuscript is the conceptual advance - oligogenic mutations as the possible cause of male infertility - the manuscript proposes, the strong rationale and reasoning of the motivation of the study, and development of a new tool to visually and quantitatively assess sperm head morphology which will benefit the field in general. The weakness that offsets these strengths is that the sperm phenotypes of the multiple heterozygous mice - while significant - are quite subtle in morphological changes and lack physiological phenotype. The study also does not provide data to support molecular mechanisms such as changes in the protein levels or localizations in their animal models. Due to these limitations, at currently presented the study remains rather descriptive and speculative. It would also be better to avoid excessive novelty claims.

Reviewer #2 (Public Review):

Digenic and oligogenic inheritance are extensions of monogenic disease models, in which effects of variation at two loci (digenic) or a few loci (oligogenic) contribute to the overall phenotype of an individual. The existence of oligogenic inheritance has been appreciated in human genetics for decades, and has been especially well documented for rare disorders with extensive locus heterogeneity, such as retinal degeneration, a condition for which more than 250 loci have been identified (Kousi and Katsanis 2015). Male infertility, itself a collection of diverse and often severe disorders affecting sperm count and sperm morphology, is likely to be driven by as many or even more loci as retinal degeneration, and is thus likely to feature oligogenic inheritance in some familial cases. Indeed, hypogonadic hypogonadism is one of the earliest and best examples of a human disease displaying digenic inheritance. Nonetheless, numerous challenges abound in the identification of digenic or oligogenic causes of male infertility, and validated examples in humans and model organisms are badly needed. In this study, Martinez et al. demonstrate oligogenic inheritance of sperm abnormalities by breeding a series of KO strains known to feature multiple morphological abnormalities of the flagella (MMAF). This is a significant paper for both the sperm abnormality field and for the broader male infertility community. the experiments and analyses are straightforward and the manuscript is well written.

My primary concerns are simply about the description of the experiments and analyses themselves.<br> 1. There are numerous references to the "% of abnormal cells", "% of head abnormalities", "% flagellum abnormalities" (Figures 1B, 2B, 3B, 4B, 5, 6 and elsewhere). There are no clear definitions of how a cell is classified as "abnormal" or a head is classified as "abnormal" or a flagellum is classified as "abnormal". Are these all defined from manual classification of images? This seems essential to know if someone would like to reproduce this experiment.

2. In order to be of most value to the community, it would be helpful to provide the individual-level data behind Figures 5,6,7, indexed by genotype. Currently the supplementary tables just contain the summary statistics for each group. Further, for the individual level data, it would be good to decode the labels from "two genes" and "three genes" to the actual genotypes, since there are multiple genotypes in those groups. These data could be used for fitting genetic models to each of the traits (e.g. to estimate additive effects, epistatic effects, etc).

Reviewer #1 (Public Review):

The study by Dr. Marcheva and colleagues aims at identification of the pathways and specific proteins involved in regulation of circadian metabolic disruption. With the idea of discovering novel therapeutic targets for beta cell failure in diabetic context, the authors launch high-throughput screening of 2640 pharmacologically active compounds employing beta cell line that is lacking function Bmal1 protein and is thus arrhythmic (clonal line of Bmal1 deficient Beta-TC-6). Insulin NanoLuciferase has been introduced to the cells via lentiviral transduction and served a readout for insulin secretion capacity. The authors validate functionality of this elegant system by correlating bioluminescence signals emanating from insulin driven NanoLuc and actual levels of insulin secretion, at basal conditions and in response to high glucose concentration. Insulin-stimulating capacity of each of 2640 screened small molecules has been evaluated in these clock-deficient Beta-TC cells. The authors identified 19 initial hits that strongly increased secretion of insulin in this system. Majority of these initial hits targeted ion channels or GPCRs. Most of thus identified small molecules were not further considered due to either high dosage of the compound that was required to obtain the effect, hepatotoxicity, or effects on insulin secretion in the presence of basal levels of glucose when tested on WT mouse islet cells. Among these primary hits, the authors thus focused on ivermectin (IVM), by validating its insulin secretion stimulatory effects on mouse islets isolated from arrhythmic Bmal1KO mice. Furthermore, the authors reported that IVM conferred improved glucose tolerance in vivo in Akita mice, and enhanced insulin secretion in response to glucose by the islets isolated from these mice. Importantly, the authors provide mechanistic insights into the effect of IVM on insulin exocytosis by demonstrating that it modulates glucose stimulated flux of Ca and increases capacitance in Bmal1KO primary beta cells, as well as in human beta cells. Moreover, the authors conducted extensive search for the factors that promote peptide exocytosis, hypothesizing that those might be efficient in rescuing insulin secretion perturbed upon circadian disruption. Employing RNAseq, ChIP analyses and additional approaches, the authors identified purinergic receptor P2Y1 as a target of IVM, that was also rhythmically expressed. Indeed, by blocking P2Y1 signaling pharmacologically, the effects of IVM on insulin secretion were abrogated. In line with this finding, beta cells lacking P2Y1 failed to respond to IVM application. To summarize, based on the large-scale drug screening, the authors identify novel function of the P2Y1 receptor that is driven by Bmal1, and that regulates glucose stimulated Ca influx and insulin exocytosis in mouse and human beta cells in response to IVM. This discovery represents an important advance in our understanding of regulatory machinery of insulin secretion by cell-autonomous clocks operative in beta cell in mice and in humans, and it is of fundamental clinical relevance in context of novel therapeutic targets for diabetes management.

Overall, the study is very well designed and carefully controlled, the authors used state-of-the art approaches, and the work has an important translational potential. The manuscript is very structured, and it is written in a clear and concise manner.

Reviewer #2 (Public Review):

In this study Marcheva, Weidemann, Taguchi et al. aimed to identify pharmacologically-active compounds that augment insulin secretion in a model of ß cell failure related to circadian disruption (i.e. in pancreatic ß cells and islets that lack the circadian core clock gene Bmal1). To this end they performed an unbiased high-throughput screen, including 2,640 small molecule compounds, in Bmal1 knockout (KO) ß cells expressing an insulin-Nanoluciferase reporter gene. Bioluminescence intensity following simultaneous treatment with glucose and pharmacological compounds was used to quantify changes in insulin secretion and identify those compounds that rescue insulin secretion from ß cells upon Bmal1 KO. They were able to narrow down a single compound hit by additional rounds of validation in ß cells, mouse and human islets and finally uncovered a potential mode of action by which ivermectin positively regulates insulin secretion. They suggest that the purinergic P2Y1 receptor is regulated by the circadian clock (via BMAL1 dependent transcription) and mediates ivermectin's insulinotropic activity. These data add an interesting way to the ongoing connection between circadian disruption and the development of type 2 diabetes based on ß cell failure resulting from perturbed circadian oscillation in these cells.

Overall, the authors present a scientifically sound study. In some cases the conclusions are supported by the data, in other cases hypothesis and experimental approaches need to be clarified and extended and some additional controls need to be provided.

Strengths:<br> The authors developed an insulin-nanoluciferase reporter system that allows to monitor changes in insulin secretion in a sensitive, cost-effective, and high-throughput manner. This system can be useful to study ß cell function in different context than circadian disruption or pharmacological compound screens. They use a variety of different methods (e.g. bioluminescence and calcium imaging, membrane capacitance measurement, RNA-Seq, in vivo ivermectin treatment and glucose tolerance test) and tools (e.g. Bmal1 KO ß cells, P2ry1 KO ß cells, mouse and human wild type islets, and pancreatic Bmal1 KO mouse islets) to characterize the mode of action of the pharmacological compound they identified. In many parts of the manuscript, the conceptualization and methodology is logical and comprehensible.

Weaknesses:<br> While the authors initially set out to identify pharmacological compounds that are able to enhance insulin secretion in a context of circadian disruption (Bmal1 KO ß cells recapitulate secretory defects observed in primary clock-deficient islets), they later also focus on the activity of ivermectin in wild type ß cells and islets to show that this compound enhances insulin secretion. Thus, the question whether the compound hit and mode of action they identified are specific to "circadian ß cell failure" remains. Moreover, in 2015, Burns et al. (Cell Metabolism) published a similar study using an insulin-gaussia luciferase reporter system to measure changes in insulin secretion in a model of ß cell failure by a 1,600-pilot compound screen. Why their nanoluciferase system is be advantageous over the previously published gaussia luciferase reporter system needs to be clarified by the authors. Lastly, even though Bmal1 KO is a commonly used model for circadian clock disruption, BMAL1 transcription factor may also exhibit biological functions independent of its role as circadian clock component. It would interesting to see if other models of circadian ß cell failure (e.g. Clock∆19 or Cry1,2 dKO) recapitulate the effects of ivermectin treatment.

Reviewer #1 (Public Review):

The authors used two different genetic approaches of reducing ndrg1a in Zebrafish to investigate its role in metabolic suppression in adaptation to hypoxia. Major strengths include high-quality phenotypic data on organismic survival, striking changes of NKA pattern and abundance in vivo by both hypoxia and ndrg1a genetic manipulations. A major weakness is lack of direct data to characterize "metabolic suppression" and evidence causally linking NKA degradation and metabolic suppression to organismic adaptation and survival against hypoxia. Ndrg family proteins have been previously implicated in many other cellular pathways and functions that are independent of NKA regulation. Thus, while roles of ndrg1a in regulating NKA abundance are novel and clearly shown, the underlying mechanisms and physiological significance remain firmly established.

Reviewer #2 (Public Review):

The manuscript entitled "N-myc downstream regulated gene 1 (NDRG1) functions as a molecular switch for cellular adaptation to hypoxia" seeks to understand the function of NDGRG1 during induced hypoxia as it relates to the decreased expression of an energy demanding sodium-potassium pump expressed in the pronephric duct and ionocytes. The authors create a novel mutation in the NDRG1 gene that is not associated with lethality and use the mutant to characterize expression of a protein component of the sodium-potassium channel in question. Utilization of zebrafish is a major strength of this manuscript as the question would be more difficult to address in other systems as zebrafish are known for their anoxic survival. The authors also use some very nice techniques that include measuring of kidney clearance, measurements of ATP, lactic acid quantification, and PLA to support their hypothesis. Overall, the data suggest that expression of NDGR1 in the PD and ionocytes is required to initiate the degradation of ATP1A1A, a protein used as a readout for degradation of the sodium potassium pump in question. the authors show colocalization of NDGR1 and ATP1A1 that increases after lactic acid build up and presumably during a switch from aerobic respiration to glycolysis. Additional studies also suggest lysosomal and proteosomal pathways that break down the ATP1A1 protein during hypoxia. Some weaknesses associated with the work include a somewhat small sample size across in situ hybridization/immunohistochemistry (3-9 total animals in replicates) for zebrafish related studies. Some of the graphs and quantification lack significance or at a minimum lack an indication of significance and most of the in situ hybridizations from supplemental and some figures are unclear due to what appears as either a growth delay or different stages. This impacts correct interpretation. Many images are also mounted at different angles preventing correct analysis. Overall, the data support a function for NDGR1 in the PD and during hypoxia, the connection with lactate and the proteosome is less strong albeit very interesting. The conclusions suggesting no role for NDRG1 in kidney function and development from early figures is not substantiated and therefore the conclusions should be toned down. The manuscript requires careful re-examination of stages and better image quality, however, in the event that all embryos can be stage matched through somite counts, the paper will have an impact to closely related fields.

Reviewer #1 (Public Review):

This manuscript investigates CypD gene regulation in OB differentiation. CypD regulates MPTP opening in IMM and hence its significance in regulating Mt function, OxPhos and cell differentiation.

The data demonstrate presence of potential Smad-binding elements in the upstream region of Ppif gene and functionally confirmed by down regulation of CypD mRNA expression in the presence of Smad1 over expression, while Runx2 expression was induced. Using ChIP assay the authors then attempted to demonstrate interaction of Smad1 with Ppif promoter in response to BMP and it's functionality using appropriate Luc reporter construct. Additional inhibitory approaches such as Smad7 and noggin were considered to validate this response. The data further show that Ppif gene expression and CypD protein expression are reduced during osteogenic differentiation. Conversely, expression of caCypD inhibited osteogenesis.<br> Overall, the findings of this study shed some additional light on the transcriptional activity of Ppif in osteoblasts, yet only moderately advance current knowledge related to the role of CypD as a regulator of osteogenesis. Specifically, the role of CypD in osteogenesis has been documented demonstrating benefits of CypD loss of function to bone. Hence, demonstrating the negative impact of CypD gain of function on osteogenesis and bone formation is only confirmatory and appears to not advance current knowledge. However, although certain aspects of the transcriptional regulation have been investigated, this study adds additional important insights.

Reviewer #2 (Public Review):

Eliseev and colleagues investigated the hypothesis that BMP signaling indirectly regulates the function of the mitochondrial permeability transition pore by controlling the transcription of cyclophilin D, a component of the pore. As mitochondria dysfunction correlates with increased pathophysiology in many diseases, understanding how mitochondrial function is regulated is of significant interest. Here the authors use the skeleton as a model system, and employ a variety of methodologies, including mouse loss and gain of function studies, cell based assays, bulk RNAseq, and phenotype rescue experiments, to examine the relationship between BMP signaling, cyclophilin D expression and bone formation. This experimental approach produced data that convincing demonstrates that BMP signaling can inhibit opening of the mitochondrial permeability transition pore to regulate the function of bone forming cells and ultimately enhance bone formation. However, in several instances the data provided do not completely support the authors' conclusions, and the specific impact on the study of age-related bone loss still remains to be determined. An additional significance of this work is the ability to test if the regulation of the mitochondrial permeability transition pore by BMP signaling is a general phenomenon of cell physiology. As BMP signaling occurs in all cell types, and this proposed function has not been previously studies, the authors data support a conceptual advance in our understanding of how the many distinct BMPs exert control over biological functions.

Reviewer #3 (Public Review):

The TGF-beta superfamily BMP/Smad transcription factors play a pivotal role in multiple pathophysiological processes including bone morphogenesis, remodeling. This study attempts to show that CypD (Ppif) gene is an important target of BMP/Smad. In particular, the authors stress that down regulation of CypD gene expression by overactivation of BMP/Smad negatively regulates the transcription of Ppif gene thereby suppressing MPTP, a process which leads to cell death. The study also suggests that the pathway becomes defective in aging which may be the cause of bone defects.

Reviewer #1 (Public Review):

Strength:<br> Based on the previously published data (Binda et al., 2018), the authors focused their analysis on two subcortical ROIs, ventral pulvinar and LGN, and disclosed short-term ocular dominance plasticty in the ventral pulvinar but not in the LGN following monocular deprivation. The analysis method is generally sound, and the writing is clear. They primarily performed an FFT analysis, combining two more traditional analyses. The main finding in the ventral pulvinar was supported by all the three methods.

Weakness:<br> Although the paper does have strengths in principle, the weaknesses of the paper are the insufficient analyses and some writings that might potentially bias the main conclusion.

Line 72: Bourne & Morrone's (2017) review paper introduces the connectivity between the early visual cortex and ventrolateral subdivision of lateral Pulvinar. That is perhaps why the authors hypothesize that ventral Pulvinar may support ocular dominance plasticity. However, readers may wonder why it is the ventral Pulvinar but not the dorsal if they are not familiar with that review paper. For example, I was a little confused when I first read this paper.

A related question thus arises. Did the authors try the similar analysis on the dorsal division of Pulvinar? Showing the results there (even if they are negative) may help further understand the function of Pulvinar.

A further related question. Did the authors check different divisions of LGN? Different parts of LGN may have different connectivity patterns, too. LGN receives direct and robust feedback from the cortex. There are different feedback connectivities between the layers of LGN and the layers of V1. For example, in macaque monkey, cells in cortical layer 6A were reported to project to the LGN P layers (or neighboring K layers, K3-K6) while cells in 6B were reported to project primarily to the LGN M (or neighboring K layers, K1-K3). In the present work, the BOLD response was calculated and reported only for the entire LGN, without separating its different layers. Not clear if different layers of LGN would show distinct BOLD response patterns.

What about the phase results? The current manuscript only reports the amplitude results for the FFT analysis.

Line 272: With Bold results, one cannot clarify the effect is feedforward or feedback, as the authors also proposed. Therefore, for now it seems not safe to say the plasticity originates in the pulvinar. Also at line 58, it is not clear why the authors propose that possibility in the Introduction. Input signals of each monocular pathway do not converge until they reach the visual cortex. Cortical changes of neural activity may be fed back to LGN though. Without feedback modulations from the cortex, it is hard to imagine why ocular dominance plasticity can originate at LGN.

Reviewer #2 (Public Review):

While this is an interesting paper using a clever behavioral paradigm to induce and measure short-term ocular dominance plasticity in humans, there are some limitations of the current manuscript, which limit the strength of claims made about the relative roles of the visual pulvinar and LGN in this plasticity:

1. Established major differences between LGN and vPulv properties and connectivity: LGN relay neurons receive their strongest driving input from a single eye, and are considered monocular. While there may be cross-talk between eye-specific information at the level of the LGN (because of intrinsic circuitry, cortical input, or both), this stands in stark contrast with vPulv neurons, which are largely binocular and receive their driving input from a range of visual cortical areas. A concise review of the literature on these subjects would help better define the hypotheses, and modulate the interpretation of results obtained.<br> 2. A key animal study previously showed how binocular rivalry correlated with changes in LGN versus vPulv firing rates (Wilke et al. Proceedings of the National Academy of Sciences Jun 2009, 106 (23) 9465-9470) presenting results that dramatically parallel those reported here, and also in the context of binocular rivalry - a discussion of those findings and their implications for the present paradigm seems necessary and useful for interpretation of findings.<br> 3. Other fMRI work in humans reporting strong BOLD signal modulation in the LGN associated with periods of perceptual dominance and suppression during binocular rivalry should also be reported and discussed (Haynes JD, Deichmann R, Rees G. 2005. Eye-specific effects of binocular rivalry in the human lateral geniculate nucleus. Nature 438(7067):496-499; Wunderlich K, Schneider KA, Kastner S. 2005. Neural correlates of binocular rivalry in the human lateral geniculate nucleus. Nat Neurosci 8(11):1595-1602).

Overall, while the results advance the field by presenting evidence of changes in vPulv [but not LGN] activity in concert with changes in perceptual performance reflective of ocular dominance plasticity, they do not reach the level of evidence needed to claim that these changes (or lack thereof) are causal, even differentially so. Nonetheless, the insights provided are useful, especially if the authors could expand on their [albeit speculative] discussion of how differences in circuitry, connectivity and physiological properties of the vPulv versus LGN could underlie the observed phenomena.

Reviewer #1 (Public Review):

This manuscript uses computational approaches to study in live cells the epithelial to mesenchymal transition (EMT) , which is relevant for cancer, development, and wound healing. The manuscript's strengths are its powerful combination of interdisciplinary approaches to define cell paths during phenotypic transitions in live cells, with an ability to discern two possible cell transition path categories without multi-color labeling or other advanced experimental approaches, which could be impactful. The manuscript's weaknesses are reading difficulty arising from the complexity of the computational analysis and the uncertainty of the method's applicability to other cell lines, proteins, and cellular transitions. The claims are justified by the data, given this particular set of analysis steps, but the question remains if fewer or simpler steps could lead to the same conclusion, or how the methods and findings relate to existing knowledge of bifurcations and cellular phenotypic transitions.

Reviewer #2 (Public Review):

I think this is a very interesting and timely contribution to the literature. It combines a dynamical systems perspective and single cell data in a very neat and exciting combination in order to identify aspects of the EMT process and dynamics.

This is an ambitious and multi-faceted study and draws on a wide range of experimental, data science, and modelling tools and techniques. Overall I really liked the scope and focus of the study. I do believe that there are a few points where the arguments can be tightened and I will focus on those aspects.

General Comments:

In order to capture the dynamics the authors should perhaps engage with the arguments in Cruel and Flandoli (J Dynamics Diff Equations) which prove that additive noise destroys a pitchfork bifurcation. Related to this I think the arguments in PMC3372930 should be considered. They make a case against the pitchfork bifurcation on purely dynamical grounds. In PMID: 27616569 the arguments are not made quite as forceful but this is an excellent background reference. Against this background it is probably not surprising that the dynamics are best explained by saddle node bifurcations.

One potential concern relates to the construction of the Langevin equation. Additive noise is a very specific choice and needs to be clearly justified. It is convenient, but not based on any physical reasoning in this case. We know that multiplicative noise (e.g. in the chemical Langevin equation, or geometric noise) will qualitatively alter the dynamics compared to the deterministic model. Much of the discussion in lines 250-260 is therefore limited or restricted to the case of additive noise and this needs to be made explicit. If additive noise is chosen because reaction coordinates can only be easily defined in this framework then this limitation should be specified.

I can see that the simple additive noise makes the integrations in the calculation of the potential 486-499 easier, but again the limitations of this approach should be addressed either by pointing them out, or by considering a model with multiplicative noise.

The most intriguing result to my mind is the existence of multiple reaction paths. I would like to see to what extent this is robust to e.g. multiplicative noise and other factors in the analysis.

Reviewer #3 (Public Review):

Building upon their established framework that reconstructs the transition pathways between cell phenotypes from time-lapse live-cell imaging, Dr. Xing and colleagues quantified the epithelial-to-mesenchymal transition (EMT) - central to development and pathology - in which epithelial cells lose their "fellowship" and begin to migrate and reproduce on their own. They directly demonstrated for the first time that EMT proceeds through two parallel pathways, i.e., the original epithelial state evolves into two metastable states that then gravitate toward the same mesenchymal state. This is an epic work that provides novel insight of EMT and developmental processes, and defines a new, promising direction of systems biology.

The strength/significance of the work lies at the following aspects. First, instead of relying on data mining as commonly practiced in the field, the authors built their quantitative framework with a solid theoretical underpinning and a clear physical perspective. Using deep learning/AI to extract information from imaging, they extended the techniques of transition path sampling - well-established in physical chemistry - to reconstruct the reaction coordinates and the associated pseudopotentials characterizing the pathways of cell phenotypic transitions. Second, rather than a few static snapshots, the authors tested their theory with the live-cell time-lapsed imaging, in which the spatial-temporal correlation intrinsic to the system can be better preserved. This greatly constrains the possible interpretations of data and hence enables more faithful determination of the mechanisms. Last but not the least, the authors present a coherent pipeline with a proven success (in EMT) that is capable of quantitatively characterizing how cells evolve from one phenotype to another in any system, with the potential of mapping out the entire landscape for development. The weakness of this manuscript pertains to the clarity in writing, including but not limited to the theoretical assumption, presentation of methodology, and potential connections with existing experiments. Overall the key conclusion is justified with the data presented.

Reviewer #1 (Public Review):

Peter Dietrich and his collaborators performed a complex experimental study aiming at exploring an interactive effect of selection history (offspring of plants grown in low- and high-diversity plots), soil origin (soil from low- and high-diversity plots) and experimental treatments (drought or nitrogen addition) on performance of four grass species. The authors did so to examine eco-evolutionary feedbacks between plant community diversity and global change drivers. Specifically, the authors hypothesize that decline in species richness due to the drivers can induce a selection regime that will select for traits that will make species more vulnerable to the further effects of global change drivers, amplifying thus the initial diversity decline. The authors indeed found that all three factors, and their interaction can affect plant performance, though the effects detected here were often species-specific.

Reviewer #2 (Public Review):

The authors present work from a greenhouse experiment testing the influence of plant and soil histories on seedling responses to global change. They grew seedlings of 4 grass species via seeds collected from different historical levels of plant community diversity (2 vs 6 species) as well as in home and away soil inoculum and a combination of these. The authors find that certain plant species respond differently to global change depending on the historical plant diversity (6 vs 2 species) and to a lesser extent the soil history. These effects were primarily species specific and affected plant traits rather than biomass.

Strengths of this study include the thorough experimental approach and novel question regarding how plant diversity may modulate plant-soil interactions under global change. Weaknesses of this study include weak or unclear support for several of the proposed hypotheses as well as lack of clear results to support the main conclusions and title of this paper.

Reviewer #1 (Public Review): 

Overall the findings are novel in that these are the first sensors for a panel of 4 nicotinic receptor partial agonists and their characterization is thorough; the experimental results support the conclusions. The main strengths are in that organellar partitioning is likely an essential aspect of the mechanism of nicotine addiction, and is therefore also essential to understand in its treatment. We lacked tools to follow the anti-addiction candidates in cells. This study provides these tools and shows enough cellular characterization to suggest they will be useful. A perceived weakness is that the current compounds for which sensors were developed are not promising anti-addiction drugs or leads; however, the study provides a template for what appears to be facile development of bespoke sensors for promising lead compounds.

Reviewer #2 (Public Review): 

This beautifully crafted paper describes what may be a true paradigm shift. Only time will tell. The paper bridges very different fields; the design and use of genetically encoded fluorescent biosensors, pharmacology, and medicinal chemistry. Here we can see for the first time signaling dynamics caused by different compounds at the nicotinic acetylcholine receptor. The compounds act on receptors on many membranes, including intracellular ones in fascinating ways. The current world of drug discovery typically involves cell homogenates and end point assays, crude measurements to be sure. Here we glimpse the subcellular and temporal patterns of activation of the different biosensors which begins to provide a richer, more mechanistic view of the compounds action. While this paper focuses on just one small set of ligands and one target, it is easy to imagine that in the future this sort of approach could become common place in the search for the better drugs our society needs.

Reviewer #3 (Public Review): 

The manuscript by Nichols and collaborators presents a description of the development and characterization of new biosensors for detection of a set of nicotinic ligands that may play a role in smoking cessation. The sensors can be targeted to the plasma membrane (PM) or endoplasmic reticulum (ER) providing a means to enable cell-based measurements of concentration and kinetics of acetylcholine and nicotinic ligands at these sites. This work is an extension of previous work describing development of sensors (iNicSnFRs) for acetylcholine, nicotine and varenicline, an approved smoking cessation treatment. These sensors combine a periplasmic binding protein (PBP) coupled to a circularly permuted GFP variant such that nicotinic ligand binding to the PBP leads to an increased fluorescence emission from the GFP variant. The authors used both a structural approach to guide sensor development and a site-saturation mutagenesis tactic to evaluate and optimize sensor design leading to a set of iDrugSnFR sensors for four nicotinic partial agonists dianicline, cytisine, - 10-fluorocytisine, and 9-bromo-10-ethylcytisine. These sensors complement previously developed sensors for nicotine and acetylcholine. The approach is rigorous, and a variety of complementary techniques were used to evaluate the novel sensors. 

These novel sensors can provide important information to guide development of improved smoking cessation agents. Previous work has suggested a framework for interpreting the acute effects of nicotine and nicotine dependence, in which acute effects and reward are mediated by plasma membrane nicotinic acetylcholine receptors (nAchR) and dependence may be, in part, based on nicotine effects on trafficking of intracellular AchRs acting via an "inside out" mechanism. Prior work suggests that partial agonists with lower efficacy than nicotine acting at the cell surface could possibly serve as effective smoking-cessation drugs. However, compounds with physiochemical properties that enable CNS exposure may also enter neurons and engage the "inside out" mechanism, limiting efficacy. The novel sensors described in this paper now provide a means to measure concentrations of these nicotinic ligands, and of structurally related compounds, in these compartments with sufficient sensitivity and temporal response to evaluate compound performance. This toolbox of sensors can presumably be expanded to accommodate additional structurally diverse nicotinic agents. 

The approach used in the paper is rigorous, the data are high quality and support the conclusions. The use of a variety of complementary techniques to evaluate the novel sensors is a clear strength. The structural information may be useful to guide future efforts in design of related sensors. The kinetic analysis of sensor response provides clear information to guide use of these sensors in biological experiments, but also raises additional mechanistic questions for future studies.

Reviewer #1 (Public Review): 

The manuscript reports the effects that loss of Srf function has on different neural crest lineages in the mouse. SRF is an important transcription factor that regulates proliferation, migration and differentiation via discrete sets of co factors, such as TCFs and MRTFs. The authors generated a null mouse, as well as a mutant that inhibits Srf interactions with MRTF factors specifically. The manuscript describes a detailed and beautifully illustrated phenotypic analysis and comparison of the resulting phenotypes together with bulk RNASeq analyses. The authors find that Srf mice with point mutations that disrupt Srf interactions with MRTFs show the same gene expression changes as Srf flox/flox mice, but they lack the craniofacial defects. Therefore, they conclude that within neural crest, the main function of Srf is in the cardiac neural crest lineage where it regulates cytoskeletal genes. The study does not provide mechanistic insight into why the Srfα allele mostly affects the cardiac neural crest lineage and not the early embryo or other crest lineages, but the possible mechanisms are sufficiently discussed. The carefully executed study suggests new mechanisms by which Srf regulates transcription.

Reviewer #2 (Public Review): 

The SRF transcription factor regulates gene transcription through associating with ERK- and actin-regulated cofactors belonging to the TCF and Myocardin families. Each family has multiple members, which exhibit differing expression profiles, and which are to varying extents functionally redundant, which has complicated their functional analysis. Thus, while inactivation of SRF itself leads to failure of gastrulation, inactivation of individual TCF and myocardin-family genes results in much later developmental defects, or barely affects development. Nevertheless studies of this sort have established that myocardin is limiting for VSM development at e10.5, while MRTF-B and myocardin function become limiting in neural crest cells at e14. 

Vasudevan and Soriano previously presented evidence for a PDGF-SRF axis operating in neural crest cells during craniofacial development: NC-specific SRF inactivation caused facial clefting, and in embryonic palatal mesenchymal cells, PDGF signalling to SRF cytoskeletal target genes that are controlled through MRTF in fibroblasts was impaired. These results pointed to a role for MRTF signalling to SRF in craniofacial development. 

"Differential regulation...." by Dinsmore and Soriano revisits these findings. They take a novel approach to assessment of SRF cofactor function by analysing an SRF variant, SRF-alpha1. This SRF derivative was previously shown to be defective in recruitment of MRTF-A (and by extension other myocardin family members) but remained competent to recruit the TCFs. They show that: 

• Homozygous SRF-1 mutant mouse embryos survive to e10.5, when they succumb to defects similar to the myocd knockout.

• Anterior mesoderm (Mesp1-cre) SRF-alpha1 embryos last to e10.5, similar to the null, and phenocopy global Myocd mutants.

• Surprisingly, (Wnt1-cre) SRF-alpha1 embryos do not show facial clefting, and there is no genetic interaction with PDGFR, although their palatal MEPM cells are selectively defective in MRTF-SRF target gene expression

• Instead, (Wnt1-cre) SRF-alpha1animals survive to birth, and succumb to cyanosis resulting from highly penetrant cardiac outflow defects reminiscent of those seen in two other models: a hypomorphic MRTF-B genetrap mutation, and an NC-specific (Wnt1-cre) Myocd mutant.

The results raise two main questions:

1) What is the basis of the gastrulation defect seen in SRF-null embryos? The results suggest that in cannot not reflect a deficient MRTF signalling, but triple TCF-deficient embryos live beyond this point, so why is there a defect? 

2) What is the basis for the craniofacial defects in wnt1-Cre SRF-null embryos? The previous proposal that they result from defective PDGF-MRTF-SRF signalling was based on correlation with defective MRTF gene expression, but the SRF-alpha1 result suggests this is not in fact correct. 

The authors cannot answer these questions, but propose three possible explanations for their findings: (i) that the SRF-alpha1 allele is a hypomorph not a null for MRTF interaction; (ii) that the TCF cofactors execute some SRF pathways; and (iii) that other undefined SRF cofactors may be involved. 

I found this paper enjoyable to read, but hard to review, because it is a clever experiment that raises more questions than it answers. It is an interesting study for a specialist in the SRF field, but less conclusive in terms of clarifying SRF's biological roles. 

Unfortunately the paper does not directly answer the questions it raises - it does not directly test the role of MRTFs (and/or myocardin) in the processes analysed, and does not assess the requirement for SRF cofactors per se using appropriate SRF mutants. For example, global MRTF-A/B double knockouts (or A/Mycd, BMycd), which could give direct insights into potential early myocardin-family functions, have not been reported. In the view of this reviewer, however, to do such experiments as part of this study would be inappropriate, and the paper would be of value as a spur to the field. 

However, I do have concerns as to the way the data are presented and discussed. To a more general reader in development or transcription, the discussion does not pose the issues clearly, and would benefit from reworking. It would be clearer if the alternative explanations for variance from the simple MRTF-null view of SRF-alpha1 should be posed briefly, and then the basis for each of the phenotypes observed considered in turn. 

In addition, the authors leave some issues unaddressed in their discussion. For example, they do not consider: 

1) That the gastrulation defect of SRF-nulls may reflect a cofactor-independent SRF activity. This is plausible, since SRF does have a constitutive transcription activation function. One possible way to test it would be to introduce a mutation such as SRF V194E that blocks both TCF and myocardin-family interactions with SRF (Ling et al 1998 JBiolChem). This mutant phenocopies an SRF-null in immune cells (Mylona et al, 2011 MCB), and should it bypass the gastrulation defect, TCF/MRTF-independent SRF function would be highly likely. 

2) That the SRF-alpha1 allele is a hypomorph for MRTF interaction but a null (or stronger hypomorph) for Myocd interaction. On this model the SRF-alpha1 phenotypes might reflect Myocd recruitment - the lack of craniofacial phenotype might reflect residual MRTF-B interaction, but the later cardiac outflow phenotype would arise from limiting Mycd interaction. 

3) That for some functions the TCF and Myocardin families act through SRF in a functionally redundant manner, so inactivating one family would not impair function. 

4) That the MRTF-A and MRTF-B proteins act functionally redundantly with myocardin. 

5) Their previous paper identified Mrtfa as the only MRTF expressed above background level in MEPM cells. However, the MRTFa knockout mouse develops normally. Thus, if MRTFs are involved in the clefting phenotype, a substantial decrease in MRTF activity can be tolerated before the phenotype becomes manifest. The nonclefting phenotype of the SRF-alpha1 mutant would not be unexpected if this were the case. 

6) The TCFs and MRTFs seem to compete to some extent at most SRF targets - for example, loss of TCFs potentiates cytoskeletal contractility. Thus the effectiveness of the SRF-alpha1 mutation in blocking MRTF-SRF activation in a given setting will be dependent not only on MRTF level but also on TCF level.

Reviewer #3 (Public Review): 

The manuscript "Differential regulation of cranial and cardiac neural crest by Serum Response Factor" aims to illuminate the differential functional mechanism of Srf during cranial and cardiac neural crest development, especially with respect to the SRF-TCF and SRF-MRTF complexes. The early embryonic lethality of Wnt1-Cre;Srffl/fl mice suggests that Srf has a critical role during neural crest development. By in vivo analysis of marker genes involved in cranial neural crest patterning, the authors found that craniofacial patterning was largely normal in these mice. Further bulk-RNAseq analysis showed that the most differentially expressed genes in Wnt1-Cre;Srffl/fl compared to controls are those targeted by the SRF-MRTF complex, which led the authors to hypothesize that this complex may play a critical role in midfacial development. To test this hypothesis, the authors generated point mutation mice (SrfαI/αI mice) in which SRF-MRTF-DNA formation is disrupted, while the SRF-TCF-DNA complex is unaffected. SrfαI/αI mice were found to die at the early embryonic stage and morphological defects were observed at E9.5, including turning defects, delayed neural tube closure, missing/hypoplastic second pharyngeal arch, abnormal hematopoiesis, and more. Further in vivo analysis showed reduction in the F-actin level, the number of CD31-positive cells, and cell proliferation, along with significantly increased cell death. The authors then investigated the function of Srf in the anterior mesodermal lineage using Srfflox/flox;Mesp1Cre/+ and SrfαI/flox;Mesp1Cre/+ and found that the defects observed in these two models were similar. Surprisingly, no craniofacial defects were observed in SrfαI/flox; Wnt1-CreTg/+ mice, although the trends of gene and protein expression were very similar between SrfαI/flox;Wnt1-CreTg/+ and Srfflox/flox;Wnt1-CreTg/+ mice. Meanwhile, SrfαI/flox;Wnt1-CreTg/+ mice died at newborn stage and exhibited outflow tract defects. 

This study provides an excellent model to investigate the function of Srf in cardiovascular development. While the discovery of a differential response to the same regulator within different neural crest populations is novel and interesting, the detailed regulatory mechanism will need additional clarification.

Reviewer #1 (Public Review): 

The dependence of cell volume growth rate on cell size and cell cycle is a long-standing fundamental question that has traditionally been addressed by using unicellular model organisms with simple geometry, for which rough volume estimates can be obtained from bright field images. While it became soon apparent that the volume growth rate depends on cell volume, the experimental error associated with such measurements made it difficult to determine the exact dependencies. This challenge is even more significant for animal cells, whose complex and dynamic geometry makes accurate volume measurements extremely difficult. Other measures for cell size, including mass or fluorescent reporters for protein content, partially bypassed this problem. However, it becomes increasingly clear that cell mass and volume are not strictly coupled, making accurate volume measurements essential. In their previous work, Cadart and colleagues established a 'fluorescent exclusion method', which allows accurate volume measurements of cells with complex geometry. In the present manuscript, Cadart et al. now take the next step and measure the growth trajectories of 1700 HeLa cell cycles with further improved accuracy, providing new insights into animal cell growth. 

They convincingly demonstrate that throughout large parts of the cell cycle, individual cells exhibit exponential growth, with the volume-normalized specific growth rate moderately increasing after G1-phase. At the very early stages of the cell cycle, cells exhibit a more complex growth behavior. The authors then go on and analyze the growth rate fluctuations of individual cells, identifying a decrease of the variance of the specific growth rate with cell volume and observed time scale. The authors conclude that the observed growth fluctuations are consistent with additive noise of the absolute growth rate. 

The experiments and analysis presented by Cadart et al. are carefully and well executed, and the insights provided (as well as the method established) are an important contribution to our understanding of cell growth. My major concern is that the observed fluctuation pattern seems largely consistent with what would be expected if the fluctuations stem from experimental measurement noise. This fact is appropriately acknowledged, and the authors aim to address this issue by analyzing background noise. However, further controls may be necessary to unambiguously attribute the measured noise to biological fluctuations, rather than experimental error. 

Major points: 

1.) To address whether the observed fluctuations could be due to experimental error, the authors analyze the fluctuations recorded in a cell-sized area of the background, and find that the background fluctuations are small compared to the fluctuations of the volume measurements. I think this is a very important control that supports the interpretation of the authors. However, I am not convinced that the actual measurement error is necessarily of the same amplitude as the fluctuations of the background. The background control will control for example for variations of light intensity and fluctuations of the fluorophore intensity. But what about errors in the cell segmentation? Or movement of the cells in 3D, which could be relevant because the collected light might be dependent on the distance from the surface? Is cell autofluorescence relevant at all? I am aware that accurately estimating the experimental error is exceptionally difficult, and I am also not entirely sure what would be the perfect control (if it even exists). Nevertheless, I think more potential sources of error should be addressed before the measured noise can be confidently attributed to biological sources. Maybe the authors could measure objects with constant volume over time, for example vesicles? As long as the segmented area contains the complete cell, the measured volume should not change if the area is increased. Is this the case? 

2.) I am particularly puzzled by the fact that even at the timescale of the frame rate, fluctuations seem not to be correlated between 2 consecutive time points (Fig. 5-S2b). This seems plausible for (some) sources of experimental error. Maybe an experiment with fast time resolution would reveal the timescale over which the fluctuations persist - which could then give us a hint about the source? 

3.) The authors use automated smoothing of the measurement and removed outliers based on an IQR-criteria. While this seems reasonable if the aim is to get a robust measurement of the average behavior, I find it questionable with respect to the noise measurements. Since no minimum time scale has been associated with the fluctuations interpreted as biological in origin, what is the justification of removing 'outliers', i.e. the feature that the authors are actually interested in? Why would the largest fluctuations be of technical origin, and the smaller fluctuations exclusively biological? 

4.) If I understood correctly, each volume trajectory spans one complete cell cycle. If this is the case, does Fig. 1e imply that many cell cycles take less than 2-3 hours? Is this really the case, and if so, what are the implications for some of the interpretations (especially the early cell cycle part)?

Reviewer #2 (Public Review): 

In this paper, the authors use a volume exclusion-based measurements to quantify single cell trajectories of volume increase in HeLa cells. The study represents one of the most careful measurements on volume regulation in animal cells and presents evidence for feedback mechanisms that slow the growth of larger cells. This is an important demonstration of cell autonomous volume regulation. 

While the subject matter of the present study is important, the insights provided are significantly limited because the authors did not place their findings in the context of previous literature. The authors present what seems to be remarkably accurate single cell growth trajectories. In animal cells, a joint dependency of growth rate on cell size and cell cycle stage has been previously reported (see Elife 2018 PMID: 29889021 and Science 2009 PMID: 19589995). In Ginzberg et al, it is reported "Our data revealed that, twice during the cell cycle, growth rates are selectively increased in small cells and reduced in large cells". Nonetheless, these previous studies do not negate the novelty in Cadart et al. While both Cadart and Ginzberg investigate a dependency of cellular growth rate on cell size and cell cycle stage, the two studies are complimentary. This is because, while Ginzberg characterise the growth in cell mass, Cadart characterise the growth in cell volume. The authors should compare the findings from these previous studies with their own and draw conclusions from the similarities and differences. Are the cell cycle stage dependent growth rate similar or different when cell size is quantified as mass or volume? Does the faster growth of smaller cells (the negative correlation of growth rate and cell size) occur in different cell cycle stages when growth is quantified by volume as compared to mass?

Reviewer #1 (Public Review):

Transient receptor potential vanilloid 4 (TRPV4) is a mechanosensitive ion channel that mediates the influx of extracellular Ca2+, subsequently, activates the intracellular Ca2+-triggered signaling cascades. The expression of TRPV4 is increased in several cardiac pathologies such as pressure overload, aging, ischemia-reperfusion, and heart failure. TRPV4 inhibitor ameliorated pulmonary edema associated with heart failure in animal models. Early phase clinical trial for TRPV4 as a potential treatment for heart failure has been done with promising results.<br> In this study, the authors showed the upregulated TRPV4 in the hearts of pressure overload mice model and heart failure patients. Deletion and inhibition of TRPV4 improved cardiomyocyte hypertrophy both in-vivo and in-vitro in the TAC model and angiotensin II/PE treated NVRMs, respectively. Mechanistically, TRPV4 activation induced cardiac hypertrophy, inflammation, and fibrosis via Ca2+ influx which augmented CaMKII phosphorylation and subsequently activated the NFkB pathway. All of which were blocked by TRPV4 antagonism and thus cardiac function was improved.

Strength<br> This study is methodologically robust, combining both in vivo and in vitro experiments with convincing data and appropriate statistical analysis.

Weakness<br> The scientific merit of this study may be limited since TRPV4 has been studied for over two decades, and there are several studies have been done concerning heart failure.

Reviewer #2 (Public Review):

This study is well-performed, methods and controls are adequate, and the manuscript is well-written. The authors combined in vitro experimental approaches with a preclinical in vivo model and human samples. The data reveal that mechanosensitive calcium channel TRPV4 is a critical driver in a preclinical model of hypertrophy and its activation induces a pro-inflammatory changes on mRNA and protein level in myocytes through CAMKII/NFkB p65 signaling.

The conclusions of this manuscript are well supported by data in a comprehensive and translational study. My main concerns relate i) to the missing attempt to show a clinically relevant intervention, ii) to the choice of the genetic model, iii) to a lack to critically discuss the role of endothelial TRPV4 in this model, iii) and the missing clinical background information since the investigated cohort of patients is quite small.

Reviewer #1 (Public Review):

The authors report the generation of a mesoscale excitatory projectome from the ventrolateral prefrontal cortex (vlPFC) in the macaque brain by using AAV2/9-CaMKIIa-Tau-GFP labeling and imaging with high-throughput serial two-photon tomography. They present a novel data pipeline that integrates the STP data with macroscopic dMRI data from the same brain in a common 3D space, achieving a direct comparison of the two tracing methods. The analysis of the data revealed an interesting discrepancy between the high resolution STP data and the lower resolution dMRI data with respect to the extent of the frontal lobe projection through the inferior fronto-occipital fasciculus (IFOF) - the longest associative axon bundle in the human brain.

The authors report the generation of a mesoscale excitatory projectome from the ventrolateral prefrontal cortex (vlPFC) in the macaque brain by using AAV2/9-CaMKIIa-Tau-GFP labeling and imaging with high-throughput serial two-photon tomography. They also present a novel data pipeline that integrates the STP data with macroscopic dMRI data from the same brain in a common 3D space, achieving a direct comparison of the two tracing methods. Overall the paper can serve as a how to example for analyzing large non-human primate brain data, though some parts of the paper can be improved and the interpretation of the data should also be further strengthened.

The methodological part should include more detail on image acquisition - speed of imaging, pixel residence time, total time for data acquisition of a single brain and data sizes. Also the time and hardware needed for the computational analysis should be included, including the registration to the common reference and the running time for the machine learning predictions - this should also include the F score for the axon detection.

The discrepancy between the high resolution STP data and the lower resolution dMRI data with respect to the extent of the frontal lobe projection through the inferior fronto-occipital fasciculus seems puzzling. One would expect that the STP data would reveal more detail not less.. One possibility is that the Tau-GFP does not diffuse throughout the full axon arborization of the PFC neurons, resulting in a technical artifact. Can this be excluded to support the functional significance of the current data?

Reviewer #2 (Public Review):

The authors utilized viral vectors as neural tracers to delineate the connectivity map of the macaque vlPFC at the axonal level. There are three main goals of this study: 1) determine an effective viral vector for tract-tracing in the macaque brain, 2) delineate the detailed map of excitatory vlPFC projections to the rest of the brain, and 3) compare vlPFC connectivity between tracing and tractography results.

Accordingly, my comments are organized around each aim:

1) This study demonstrates the advantage of viral tracing technique in targeting neuron type-specific pathways. The authors conducted injection experiments with three types of viral vectors and found success of AAV in labeling long-distance connections without causing fatal neurotoxicity in the monkey. This success extends the application of AAV from rodents to nonhuman primates. The fact that AAV specifically targets glutamatergic neurons makes it advantageous for mapping excitatory projections.

Although the labeling efficacy of each viral vector type is described in the text, Fig. 2 does not present a clear comparison across viral vectors, despite such comparison for a thalamic injection in Fig. 2S. Without a comparable graph to Fig. 2E, it is unclear to what extent the VSV and lentivirus failed in labeling long-distance pathways.

2) The authors quantified connectivity strength by the GFP signal intensity using a machine-learning algorithm. Both the quantitative approach and the resulting excitatory projection map are important contributions to advancing our knowledge of vlPFC connectivity.

However, several issues with the analysis lead to concerns about the connectivity result. First, the strength measure is based on axonal patterns in the terminal fields (which the authors refer to as "axon clusters"), detected by a machine-learning algorithm (page 25, lines 11-13). However, the actual synaptic connections are the small dot-looking signals in the background. These "green dots" are boutons on the dendritic trees. The density of boutons rather than the passing fibers reflects the density of synapses. The brief method description does not mention how the boutons are quantified, and it is unclear whether the signal was treated as the background noise and filtered out. Second, it is difficult for the reader to assess the robustness of the vlPFC connectivity patterns, due to these issues: i) It is unclear how many injection cases were used to generate the result reported in the subsection "Brain-wide excitatory projectome of vlPFC in macaques". The text mentions a singular "injection site" (page 8, line 12) and Fig. 4 shows a single site. However, there are three cases listed in Table 1. Is the result an average of all three cases? ii) Relatedly, it is unclear in which anatomical area the injection was placed for each case. Table 1 lists the site as "vlPFC" for all three cases, while the vlPFC contains areas 44, 45 and 12l. These areas have different projection patterns documented in the tract tracing literature. If different areas were injected in the three cases, they should be reported separately. iii) It is hard to compare the projection patterns with those reported in the literature. Conventionally, tract tracing studies report terminal fields by showing original labeling patterns in both cortical and subcortical regions without averaging within divided areas (see e.g. Petrides & Pandya, 2007, J Neurosci). It is hard to compare Fig. 3 with previous tract tracing studies to assess its robustness.

3) Using the ground-truth from tract tracing to validate tractography results is a timely problem and this study showed promising consistency and discrepancy between the two modalities. Especially, the discrepancy between tracing and tractography data on the IFOF termination brings critical insights into a potential cross-species difference. The finding that IFOF does not reach the occipital cortex provides important support for the speculation that IFOF may not exist in monkeys (for a context of the IFOF debate see Schmahmann & Pandya, 2006, pp 445-446).

I have minor concerns regarding the statistical robustness of the tracing-tractography comparison. The authors compared the vlPFC-CC-contralateral tract instead of a global connectivity pattern without justification. Why omitting other major tracts that connect with vlPFC? In addition, the results are shown for only one monkey, while two monkeys went through both tracer injection and dMRI scans. It is unclear how the results were chosen or whether the data were averaged.

Reviewer #1 (Public Review): 

This manuscript by Silver, et al., details work investigating the relationship between season of conception and DNA methylation differences at sites across the genome, measured by widely-used arrays, in two cohorts of children using Fourier regression. They find that season of conception is associated with persistent methylation differences at several hundred CpG sites, and that these CpG are enriched for properties, compared to sets of control sites, that suggest that methylation at these sites is influenced very early in development/during conception and that these sites are positioned in genomic regions relevant for gene activation and regulation. Additional analyses investigated the effects of genetic variation of these sites, and found no evidence for single nucleotide polymorphisms nor child sex confounding the associations between season of conception and DNA methylation. As the number of sites measures by these arrays are a very small amount of total sites across the genome, the authors suggest that these findings indicate there may be many more sensitive methylation 'hotspots' in the genome that are not captured by these arrays but could impact on health/development. 

The key strengths of this manuscript include the use of two cohorts of children at different ages, providing evidence that these effects of season of conception appear to attenuate by 8-9 years of age; and comparison with control sites and additional analyses investigating confounding to build the evidence for these relationships reflecting true, biological associations rather than statistical artefacts or the result of confounding. 

However, the conclusions around the potential functional importance of these methylation differences are limited by a lack of evidence for a relationship between methylation of these season-of-conception-associated sites and child growth/development, so while this manuscript builds compelling evidence for the effects of season of conception on methylation, it's functional relevance is unclear. Additionally, there are some choices made in the analyses where the rationale for those choices should be made more clear, such as the use of CpG sites above or below a certain estimated effect size for different analyses. 

Overall, the approach taken here to demonstrate different levels of evidence for true relationships between early development exposures and differences in DNA methylation is a compelling one, and the manuscript delivers clear evidence for its primary conclusions.

Reviewer #2 (Public Review): 

This is a very interesting manuscript, which will be of interest for a broader readership. The authors have analysed an unique cohort, which is of importance to understand the impact of environmental factors on DNA methylation. 

The performed analysis is well balanced, and the conclusions are justified by the presented data. It is a strength of this study, that results from the initial ENID study have been re-evaluated in the EMPHASIS study. Unfortunately, DNA methylation has been analysed using HM450 and EPIC arrays. Both methods are providing only a limited view on methylome-wide DNA methylation. 

Another limitation (as already addressed by the authors) is the lack of longitudinal samples. This would potentially have helped to gain further knowledge about the identified attenuation of DNA methylation levels at SoC associated CpGs. 

Finally, I am not entirely sure, that one confounding factor has been completely ruled out: It is known, that blood composition may cause methylation variability. In general, the authors addressed this point and analysed blood compositions (supplementary Figure 16) of both cohorts. Here, no marked seasonal differences between and within both cohorts have been identified. However, the participants of the EMPHASIS cohort have a very similar age (8-9 years). For this reason, I am wondering if methylation variability/ differences and in addition the attenuation of methylation levels might be influenced by the younger age of ENID participants compared to EMPHASIS study individuals.

Reviewer #3 (Public Review): 

Silver et al. Investigate the influence of seasonal variation (nutrition, infection, environment) on blood DNA methylation in two cohorts of children (233 [2y] and 289 [8y-9y]) from the same sustenance farming communities in rural Gambia. One cohort (450K,233) was extensively studied before in multiple publications, the second dataset (850k,289) is unpublished. Using cosinor modeling they find 768 CpGs with a significant seasonal pattern(SoC-CpG, FDR<0.05) in the probes that overlap between the 450k and 850k arrays. Look-up of these 768 SoC-CpGs in the second sample showed 61 SoC-CpGs with FDR 0.05 (no mention is made if the direction of effect is consistent, but we assume it is so). The authors notice that most SoCs seem to be attenuated in the 8-9y sample. Then the authors select out of the 768 SoC-CpG the FDR<0.05 and >=4% seasonal amplitude in this discovery sample: 257 which they bring further in (enrichment) analyses. It is unclear if all 257 are (nominally) significant in the replication sample. These SoC-CpGs are enriched for imprinted and oocyte germline loci. Roughly 10% of SoC-CpGs overlap with so-called meta-stable epialleles (MEs), on which the authors have published greatly. This is a large fold enrichment, and subsequently the main focus of the Results and Discussion. Indeed, it skews the Discussion heavily and one wonders what could have been found in the other 90%? The Discussion is heavily geared to interpretation within their MEs focus and does little to discuss study weaknesses and strengths, to which the tail of the Results suggest there are multiple. For at the end of the Results and in the Methods we find additional sensitivity analyses and discussion points on a very strong enrichment for CpGs with a mean difference in methylation between the sexes (>1/3 of the 257), adjustments for genetic confounding and a high inflation factor in the discovery cohort. 

Indeed, despite the strong and good flow of the Result section and the impressive (albeit somewhat one-side) look-up of SoC-CpGs in published datasets; the tail and Methods section leaves this reader with a strong suspicion of possible methodological issues on the measurement level already identified prior. 

The authors reports that the discovery cohort is biased in the collection of conception months (figure 2A), has a strong inflation of 1.3 (no QQ-plot is shown to assess bias in addition to inflation), no adjustment for genetic background could be made (which is false, as the 450k array contains several dedicated SNP probes, even hundreds when extracted with the omicsPrint package) and > 1/3 of SoC-CpGs is a sex CpG. For the latter observation the authors regressed out sex and repeated the analysis, noting no difference. However, regressing out sex does not help if sex is heavily correlated with confounding biological/sampling/technical covariates. 

The authors reason that the inflation is nothing to worry about citing single cohort studies on global effects on DNAm of methyl donors. Global DNAm is indeed often association with methyl donor intake but generally these studies investigate ALU or SINES repetitive elements and the PACE consortium reported only modest effects on select 450K array loci for prenatal folate supplementation, showing that their reasoning might hold on the ME loci (in/close to repetitive elements) but not the genome-wide analysis per se. 

The authors should convince the reader that their (discovery) data is valid. The data they do show in Supplemental tables 16 and 17 show that after functional normalization a strong effect of batches remains, while from my own experience these are normally nicely mitigated via functional normalization. Normally only strong cell type correlations remain in the first PCAs of the normalized data. But for ENID we see a remainder of sentrix row, often the strongest batch effect, and slide and plate remaining. Also, the biological, season and cohort specific variables are not noted here. We just must assume that the blank correction for the first 6 PCAs, rather than the actual adjustment for the measured batch/confounding effects, does not remove (or over adjusts) for biological/study design (village, genetic ancestry) effects. In addition to these observations figure 2C seems to indicate that the controls CpGs (elegantly selected by the authors) also show seasonal variation, just not as much as the SoC-CpGs. This leaves the reader to wonder: is there bias in their sample randomization across plates, rows and slides? This feeling is amplified by the fact that almost all SoC-CpGs seem to show an increase in DNAm in jul-aug (Suppl Fig. S5 and Figure 1B). [An observation that is not given enough prominence in the Results]. Which might or might not hint to a correlation with a batch effect (like sentrix row?).

Reviewer #1 (Public Review):

The manuscript is of interest to those studying the biophysical rules of adhesion molecules, and those studying the molecular underpinnings of synaptic self-avoidance. The work adds much of the biophysical detail that existing model lacked, and therefore provides support to the prevailing mechanism. The data is of high quality and interpreted properly.

Protocadherins are the proposed receptors for self-avoidance in vertebrate neurons. The authors have put forth an "isoform-mismatch chain-termination model" in previous manuscripts with ample structural evidence to support it, and further advance this model in current paper by providing a large amount of data on interaction biophysics. The paper strengthens previous claims that trans interactions (between cells) are only homodimeric, but also show that cis interactions (on the same cell) are promiscuous. They present specific reasons to why and how a few cis pairs form in asymmetric patterns. The cis interactions (if they exist and the exact asymmetrical nature) do not appear to follow a strong general rule (there is a preference for pairs from non-matching classes) - but they do not have to.

In the clustered protocadherin system, one of the unknowns was relative affinities and promiscuities of the homophilic vs heterophilic interactions. The authors report SPR data for 100+ heterophilic (trans) cPcdh interactions, showing strict homodimerization but no interactions between non-identical pairs. The binary nature of these interactions (present or absent), as the authors highlight, is indeed very remarkable. The crucial biophysical measurement of these affinities are provided by AUC data, although the data are not provided - only calculated dissociation constants in a supplemental table are. They report crystal structures of a C-type protocadherin (gC4 domains 1 to 4), which generally agrees with previous reports of cPcdh structures. They also test which cPcdh forms which side of the cis dimer via MALS by mutating the interfaces looking for heterodimer formation.

Overall, the manuscript sets out to answer a specific, limited set of questions, and does that very well. The amount of biochemical work is immense, and the interpretation of the data is masterly. (There are technical limitations to measuring affinities in mixed homo and heterodimeric systems, which prevents the authors from drawing a complete energetic description of the system, but it is not clear if one is needed to understand the relevant biology, or even sufficient to model it.) The reported structures are of superior quality, and even the lower resolution, anisotropic dataset reported is, if anything, undersold. It is clear that very careful model building has been performed in real space. There are only minor issues that the authors may want to address.

Reviewer #2 (Public Review):

Overall, the data are solid, extensive, well-illustrated, and well-presented. This is the first extensive biophysical analysis, and provides very useful quantitative data in considering how cPcdhs work. The conclusions are overall well supported by the data. The paper can be divided into three important sets of results:

First, the authors use SPR to demonstrate that all tested trans dimerization interactions are strictly homophilic. The authors emphasize that this is the strictest homophilic preference observed thus far for a family of adhesion proteins. It remains unclear whether or how this strict homophilic specificity could be important for the proper function of the cPcdh protein family.

Second, the authors present two crystal structures of a C-type trans interface (that of γC41-4), the first structures of available for a C-type cPcdh. They find that it is similar to the trans interfaces observed for both clustered and non-clustered protocadherins. Therefore, even though C-type cPcdhs have distinct expression patterns and significant sequence divergence compared to other cPcdhs, their interaction architecture is effectively the same. This is important in considering how the C-type cPcdhs could be incorporated just like other isoforms in the large zipper assemblies that are postulated at cell-cell encounters by the chain termination model for self-recognition.

Third, the authors measure the binding affinity a significant number of both homophilic and heterophilic cis dimerization interactions and find them to be not indiscriminately promiscuous, but rather in many instances preferentially heterophilic over homophilic. This is surprising, and the available data can start to provide additional constraints on the chain termination model.

The first two advances provide support for a simple model for cPcdh assemblies: we can safely only consider trans homodimers (no trans heteromers), and we can consider all trans interfaces to be roughly equivalent in terms of protein architecture (within a few angstroms of RMSD). The third advance, in contrast, suggest that we should incorporate some additional constraints on possible cis interfaces, beyond the previously observed constraints that alpha isoforms can only form cis heterodimers. Although the authors provide several interesting discussion points about how to consider the new cis interface information, they do not go as far as to develop an updated chain termination model that incorporates the information - this will likely take some time and effort, and perhaps additional data and information.

Reviewer #3 (Public Review):

This paper by Goodman et al. is the latest in a series focusing on the structural determinants of clustered protocadherin (cPcdh) isoform cis- and trans-interactions. The goal of this particular paper is to garner further details in support of the "isoform-mismatch chain-termination model" of cPcdh interaction, which was developed by the group in 2015. The model is based on their landmark initial crystallographic structural analysis of particular cPcdh ectodomains, as well as on earlier work from other groups showing that (at least) some cPcdh proteins interact homophilically in trans but promiscuously in cis. The model predicts that cis-dimers of various cPcdh isoforms form via the 5th and 6th extracellular cadherin repeats (EC5/6), and that these dimers then interact in trans strictly heterophilically via EC1-4 to form "dimers of dimers" as an initial event. If cPcdh repertoire between two cells primarily matches, then a linear "zipper" of such dimers will expand, increasing interaction and presumably associated intracellular signaling. Mismatching isoforms expressed in one cell but not the other will terminate this zipper chain, and thus cPcdh repertoire matching between cells will determine self/non-self recognition. Other groups have shown that homophilic matching between neurons is-depending upon the neuronal subtype-important for driving neurite self-avoidance or growth and branching of dendritic arbors, so the mechanisms of interaction will be important to understanding events in neural development.

The present paper builds on others by the group (e.g., Rubinstein et al., 2015, Goodman et al., 2016, 2017, Brasch et al., 2019), and primarily extends these results to more isoforms, providing also more molecular detail. There are three main findings. First, the concept that cPcdh trans-interactions are strictly homophilic is supported by many new analyses using surface plasmon resonance (SPR) assays in which an ectodomain of one isoform is coupled to a chip and those of identical vs. distinct isoforms are flowed over it to measure interactions. The data are rigorous and nicely presented and demonstrate-unsurprisingly given many prior demonstrations-that trans interactions mediated by EC1-4 are strictly homophilic. A main advance here is in the methodology, which can quantitatively and directly measure such interactions, in contrast to the qualitative cell aggregation studies that were already published. The authors also present an informative mutagenesis series identifying 5 interfacial residues that, when mutated individually or in concert to match a different highly similar intra-family isoform quantitatively shift trans-interaction from homo- to heterophilic.

The second main finding is the presentation of a new antiparallel trans-dimer structure of the gC4 EC1-4 interaction. While structures of other gamma Pcdhs have been published by the group before, the addition of the C4 structure is important for several reasons: 1) this isoform is the only one of all the cPcdhs that is essential for postnatal viability and normal neuronal survival in mice; 2) this isoform is the only one of the gamma Pcdh family that does not make it to the plasma membrane without dimerizing with a "carrier" cPcdh of some kind, which had cast doubt on whether it would interact in the same way as other cPcdhs; 3) A recent publication (not cited by the authors yet as it came out coincident with their submission) demonstrated that truncating or structure-disrupting mutations in the human PCDHGC4 gene result in significant neurodevelopmental disorders. The authors show that the structure of the C4 trans-dimer is similar to that found for other cPcdh isoforms, though the interaction is weaker than observed for others. They suggest that particular residues in the EC1:EC4 and EC2:EC3 trans interface may be responsible for this, though they do not follow up with mutation experiments to confirm. Doing so (mutating the identified C4 residues to those of, say gB2 or a delta2 Pcdh) would contribute to the novelty of the paper, as it is unclear as of yet how strength of cPcdh interactions might be regulated or manipulated.

Finally, the authors extend and confirm that cPcdh cis-interactions are promiscuous between isoforms without being ubiquitous. The essential aspect of this finding has been known since 2010 and was confirmed in papers from this group in 2015, 2016, and 2017. The primary advance again is the use of SPR rather than cell aggregation or analytical ultracentrifugation, this time using EC3-6 constructs that contain the cis-interaction EC5/6 interfaces. The data support that cis-interactions are, at least, more promiscuous than are the strictly homophilic trans-interactions, but do reveal more about the "rules" governing which isoform can interact with which others. The key finding is that interfamily (e.g., beta Pcdhs with gBs, or C-type with betas, gA or gB) heterodimers are favored and that homodimers are disfavored (with intra-family heterodimers occurring rarely). The authors complete the study by demonstrating nicely that gA4 preferentially plays the "EC6 only" part of a cis-dimer with gC3, which plays the "EC5/6" part; this builds on prior results from Goodman et al., 2017 showing this slightly offset method of cis-dimerization whereby the EC6 of one partner interfaces with EC5 and 6 from the other. Evidence is presented that at least some isoforms exhibit "handedness" in terms of which role they will play in a dimer, which is important in that it would limit the dimers that could efficiently form and thus have implications for the model they have been building. Some nice mutational studies are presented confirming this by SPR (mutations that block an isoform from an EC6 position or an EC5/6 position but not vice-versa).

The data are generally well-presented and rigorously collected; though this reviewer is not a crystallographer the presentation of the structures is clear and all supportive data are present. They support the conclusions drawn and are logical. The main issue with the paper is that, as noted above, the advances are somewhat incremental in that the main points were known prior to this study: the cPcdhs interact in trans in a strictly homophilic manner (first shown in 2010 for some isoforms, then 2014 for all, and later in 2015 and 2016 in structures); individual isoforms interact in trans in an antiparallel manner involving EC1-4 (C4 structure is new but the main conclusion is unaltered from the other structures reported in 2015 and 2016); and cPcdh cis interactions are promiscuous but constrained by some partner preferences (first reported as promiscuous based on a few isoforms in 2010, 2014, and later confirmed but with exceptions and mechanistic detail in 2016 and 2017). What is new is the methodology, which in some cases is more direct and open to quantitative determinations; the type and number of examples collected; and the aforementioned mutational analyses that drive home the conclusions Still, the paper is largely confirmatory of the authors' many excellent prior papers. For example, in Goodman et al., 2017 (PNAS), the authors present structures yielding a putative EC5/6+EC6 dimer surface and state "the structure explains the known restrictions of cis-interactions of some Pcdh isoforms"-these restrictions were known due to Goodman et al., 2016 (eLife) where it was shown that gA's do not homo-dimerize in cis but can hetero-dimerize with alpha Pcdhs. Because of this, the impact of this latest work is likely to be greatest for researchers who directly work on the Pcdhs or other members of the cadherin superfamily, for whom the additional data will be welcome confirmation and extension. Due to the confirmatory nature of the paper the impact may be less apparent to the more general reader.

Reviewer #1 (Public Review):

The authors have previously developed a powerful time-lapse recording protocol that allows them to observe in real time the formation and degradation of phagosomes in specific phagocytic cells (C1-C3) in the developing C. elegans embryo. Using this protocol in combination with specific reporters, they find that LC3-positive vesicles fuse with phagosomes and that these vesicles are double-membrane vesicles. Taking advantage of different genetic requirements for the formation of LAPs and autophagosomes, they furthermore provide evidence that these LC3-positive vesicles are autophagosomes. Having established that autophagosomes fuse with phagosomes, the authors demonstrate that preventing this fusion genetically (by blocking autophagosome biogenesis) results in a general engulfment defect and that this is due to a defect in the degradation of phagosomal content. The authors identify RAB-7 and the HOPS complex as necessary for autophagosome-phagosome fusion and the CED-1, CED-6, DYN-1 pathway as necessary for recruiting autophagosomes to phagosomes. Finally, the authors find that preventing autophagosome-phagosome fusion does not affect lysosome-phagosome fusion thereby ruling out that the effects observed are indirect and a consequence of defects in lysosome-phagosome fusion rather than autophagosome-phagosome fusion.

This is a very rigorous and very convincing study that will have a big impact on our understanding of cell corpse engulfment and degradation. It also uncovers a novel function of autophagosomes (i.e. in cell corpse degradation). Its strength lies in the combination of time-lapse observation of specific organelles in specific cells in vivo and the use of mutations in genes that affect specific cell biological processes. There are no major weaknesses but it would have been nice to have evidence for autophagosome-phagosome fusion for example through EM images.

Reviewer #2 (Public Review):

This is an interesting manuscript from Zhou and colleagues which make several important steps forward in understanding how cell corpses are digested by phagocytes. First, they show that autophagosomes associate with nascent/growing phagosomes. Interesting, they find that many atg genes required to make autophagosomes are essential for efficient corpse clearance. These include homologs of atg13 and atg14, which in mammals are not required for the production of LAPs. The argument from the authors is that these non-LAP vesicles are double-membrane and fuse with phagosomes to contribute contents important for efficient corpse clearance. If correct, this would be a new way in which autophagosomes contribute to phagocytosis. This is a reasonable interpretation, but not the only interpretation of the data.

Novel points include the recruitment of autophagosomes (non-LAPs presumably based on genetics) to phagosomes and the requirement for many atg genes in corpse elimination. That Ced-1 and components of this pathway drive recruitment of these vesicles is interesting. The genetics are very strong, convincing, and well done. It seems clear that these many ATG genes are playing a role in efficient disposal of corpses in phagosomes, but I do think it remains unclear how. It is not clear to me that a double-membrane autophagosome actually fuses with the phagosome. Possibly they are fusing to aid in degradation, but another possibility is that they are interacting in a way that contributes lipids for phagosome growth/expansion. What if this is a mechanism that allows phagosomes to grow their lipid membranes, rather than fusion to digest what's in the autophagosome? Atg2/9 drive the transfer of lipids from the ER to autophagosomes, that's how autophagosomes grow. It is possible that autophagosomes are intermediates that serve as lipid sources for phagosomes. Is there an autophagosome target (inside) that one could track to show actual degradation? It would be useful to discriminate these possibilities.

Reviewer #3 (Public Review):

The manuscript by Peña-Ramos et al. describes a new role of autophagosomes in the maturation of phagosomes containing apoptotic corpses. The authors find that vesicles containing LGG-1 and or LGG-2 bind to and fuse with phagosomes and provide evidence that these structures are, ostensibly, double-membraned autophagosomes. They then proceeded to assess the role of such fusion events in the rate and extent of phagosome maturation, using mutants lacking components required for autophagosome formation. In addition, they provide evidence that fusion involves Rab-7 and the HOPS complex and is influenced by the presence of the phagocytic receptor CED-1. Lastly, they document that lysosome fusion with the phagosomes persists in the absence of autophagosomes.

The findings are novel, generally clear and convincing. On the other hand, some of the interpretations are not unambiguous and, importantly, the ultimate mechanism underlying the defective maturation is not resolved or even addressed.

-By the authors' own admission, at least a fraction of the autophagosomes have acquired Rab-7 and likely fused with lysosomes. In that event, the fusing structures are autolysosomes and not necessarily double-membraned autophagosomes, as claimed. Delivery into the phagosome of the contents of the digested inner bilayer of the original autophagosome is likely to have occurred. If so, appearance of labeled contents inside the phagosome is being misinterpreted (or at least over-interpreted) to mean that the fusing structures are double-membraned, bona fide autophagosomes. The resolution of the images seems insufficient to distinguish these two possibilities. If autolysosomes are in fact the organelles fusing, how is this different from fusion of regular lysosomes? The conceptual novelty of the paper would be greatly diminished if what is being reported is homotypic fusion of two maturing (auto)phagolysosomes or fusion of a lysosomal organelle with phagosomes. Are the LGG-1/2-positive structures acidic and do they contain NUC-1?

-The conclusion that lysosomal fusion with phagosomes is normal is based on the quantitation of NUC-1 fluorescence acquired by the phagosomes. However, the quantitation was made relative to the fluorescence of phagosomes at time 0, when no NUC-1 is expected to be present. The validity of these measurements and comparisons between wildtypes and mutants is therefore questionable.

-The preceding comment is critical because it is generally believed that degradation of phagosomal contents is solely dependent on fusion of lysosomes that deliver degradative enzymes and make the phagosomal lumen suitably acidic for optimal function of the hydrolases. If these parameters are normal, as argued by the authors, what is preventing normal degradation from occurring? The authors invoke a mysterious molecule(s) delivered by the autophagosome as an essential component for optimal degradation. While provocative, this notion seems unfounded unless it is shown that the delivery of hydrolases and the luminal pH of the phagosome are normal, yet degradation of proteins, lipids and nucleic acids and/or their resorption is affected by the absence of the mysterious molecule(s). It is unclear whether the authors propose that the contents of autophagosomes are somehow required for hydrolase activity.

Reviewer #1 (Public Review):

Fatty acids that are too long to be transported into mitochondria are instead transported into peroxisomes for their break down i.e., beta-oxidation. The ABC transporter ABCD1 (ALDP) translocates very-long-chain fatty acids (VLCFA) conjugates to coenzyme A across peroxisomal membranes. Here, the authors have determined the cryo EM structure of human ALDP in complex with the substrate to reveal interesting differences to other ABC transporters. They have mapped ALDP-associated disease mutants, and further assessed their impact on transport by following the uptake of C22:0-CoA and C24:0-CoA into a mammalian cell line.

Strengthens: The cryo EM structure of human ALDP is well resolved and the impact of disease-causing mutants by monitoring uptake of C22:0-CoA and C24:0-CoA using LC-MS/MS rather than ATP hydrolysis (as is typically done) is appreciated.

Weakness: The proposed coordination of fatty acid CoA lacks the highly-positive charges and one would expect to require for the coordination of this negatively-charged compound. Indeed, a similar structure deposited in bioRxiv has a different location for the substrate that makes more chemical sense. The proposed conformational differences based on AlphaFold models should be taken with caution. The paper is difficult to follow in places.

Reviewer #2 (Public Review):

The structures of human proteins provide an essential framework for addressing their biomedically functions, including their mechanism, therapeutic design, and understanding the molecular basis of genetic mutations. The adrenoleukodystrophy protein (ALDP), a member of the D sub-family of ATP-binding cassette transporters or ABCD1, participates in the transport of free very long-chain fatty acids and their CoA esters across the peroxisomal membrane. The biomedical relevance is highlighted by the identification of over 900 mutations on ALDP that can lead to the severe genetic disorder X-linked adrenoleukodystrophy.

The manuscript by Jia et al describes the structure determination of ALDP in the inward-facing conformation by single particle cryo-EM at a nominal resolution of 3.4 Å. Highlights of this work include:<br> -identification of a short helix at the peroxisomal side, which distinguishes ABCD1 from the other three members of the ABCD family<br> -identification of an extended alpha-helix at the C-terminus of ALDP that forms a coiled-coil with the symmetry related sequence, although it was not possible to build an accurate model of this region<br> -identification of two CoA esters by ALDP in the cavity formed by the two TMDs (Fig. 2)<br> -mapping of pathogenic mutations of ALDP on the structure (Fig 4a)<br> -results of a cell base transport assay to evaluate the influence of several ALD-associated mutations on substrate translocation and specificity (Fig 4B)

For the reasons noted above, the structures of human ABC transporters are important advances for providing the XYZ values for the atomic coordinates. That said, given the explosion of structural information on ABC transporters (for example, of the ~50 human ABC transporters, structures have been reported for the human A1, A4, B1, B2, B3, B4, B6, B8, B10, B11, C7, D4, G2, G5, G8 transporters in addition to D1), it is a challenge to interpret new results in the context of existing knowledge about ABC transporters. In this regard, the functional implications of the ALDP structure paper are more incremental, based on the following considerations:<br> 1. the locations of the mutations (Fig 4a) and the classification of mutations discussed on page 6 could have been deduced without determining the ALDP structure.<br> 2. the transport assays provide an approach to connecting structure and function, but too little information is provided about the assays to make those connections. Of the ~900 mutants, why were the variants in figure 4B selected? Where are they located in the structure? What do the results of the transport assay mean? For example, is the reduction of C22:S-CoA in the cytosol for the G343V mutant due to impaired transport or to reduced expression, misfolding or some other cause?

Reviewer #3 (Public Review):

The manuscript describes the cryo-EM structure of the human Adrenoleukodystrophy protein (ALDP) that transports very long-<br> chain fatty acids (VLCFAs) across the peroxisomal membrane. Disease mutations have been associated with ALDP dysfunction. ALDP belongs to the ABCD1 family of ABC transporters. The overall structure resembles other members of the family; it is a homodimer that consists of 12 TM helices and 2 NBDs. In the TMD they observe density for lipid like molecules that they have assigned as Co-A esters.

Based on the available mutations, they mapped them on their structure and tried to classify them. They further evaluated some of these mutations in whole cell transport assays.

Overall, the structure provides a good framework to understand the molecular basis of diseases mutations.

Reviewer #1 (Public Review):

The authors identified a new regulatory pathway involved in T cell development centred on a transcription factor called Zfp335. They demonstrate that Zfp335 deficiency resulted in reduced development, in vivo and in vitro, due to blockade in development at the DN3 stage. This was associated with reduced intracellular TCRβ and increased apoptosis. The authors undertook transcriptomic and ChIP-seq analyses and determined that Zfp335 interacts with Bcl6 and Rorc and retroviral expression of either of these transcription factors countered the effects of Zfp335 deficiency.

Strengths<br> The paper is easy to understand and the results are presented in clear and logical manner. The authors identify a novel transcription factor that had a clear impact on T cell development, which is exciting as it helps expand our understanding of this important process.

Weaknesses<br> I felt a weakness of the paper was that the authors identified a number of processes that were impacted by this transcription factor but did not look at the interaction between them. Furthermore, the discussion would benefit from more discussion of previous research to put the results into context, including other regulatory pathways involved in T cell development.

Reviewer #2 (Public Review):

In this paper the authors address the role of the Zinc finger protein Zfp335 in early T cell development. Earlier studies established that a Zfp335 hypomorph mutant mouse had defects in T cell development. Here the authors show that conditional ablation of Zfp335 using LckCre results in a dramatic reduction of thymic cellularity and a developmental block at the DN3 stage of development. Using OP9-DL1 cultures and mixed BM chimeras they establish that the developmental defect is cell intrinsic. Here it is not clear why the authors use a starting ratio of 1 to 4 WT to KO and whether they take into account this starting difference in the final calculations. Curiously LckCre mediated ablation of Zfp335 also impacts gdT cell development which branches independently of the preTCR assembly and signaling. Then the authors suggest that the developmental bock is in the transition from the DN3 to DN4 stage which is not surprising given that LckCre mainly targets DN3 cells. The authors do not provide further detail on the transition between the DN3a and DN3b stages. The knockout cells do not seem to have a defect in proliferation however they have increased apoptosis both at the DN3 and at the DN4 stage. The authors then show that a reduced fraction of the mutant DN4 cells expresses intracellular TCRb and that providing OT1, a mature abTCR, to the developing cells resolves the transition to the DN4 stage but not to the DP.

The authors then provide molecular analyses Using RNA seq in WT and KO DN4 cells they identified 566 downregulated, and 899 upregulated genes. They specifically focus only on the downregulated genes and show a heatmap with a number of hand peaked downregulated genes with roles in thymocyte development. They also perform ChIPseq of Zfp335 in FACS sorted mix of DN3 and DN4 WT thymocytes indicating that they isolated 5x107 cells of these rear cell populations for the ChIP experiment. They identify 2797 peaks. It is not clear how many independent ChIP-seq replicas were processed neither is it clear whether the identified peaks are enriched for the consensus Zfp335 binding motif as a means of validating the ChIP. The authors also indicate a GEO accession number that could answer these questions but access to this site is still blocked. Then the authors compare a select number of 97+22 differentially expressed genes and it is not clear how these genes were peaked from the 566 downregulated and 899 upregulated genes identified as differentially expressed. Using this methodology the authors narrow down to Bcl6 and Rorc which are downregulated and are marked by Zfp335 binding at TSS proximal regions of Bcl6 and Rorc.

The authors then perform functional assays in which they retrovirally overexpress Mock, Zfp335, Bcl6 or Rorc in KO DN3 thymocytes cultured on OP9-DL1 that results in a certain rescue of development and reduction in apoptosis. They have not assessed whether a similar retroviral overexpression of Zfp335, Bcl6 or Rorc in WT DN3 cells could have affected the progression of these cells.

Reviewer #3 (Public Review):

Wang et al. report the identification of Zfp335 as a novel regulator of T cell development and demonstrate that it is critical for the DN to DP transition. Their conclusions are supported by the analysis of Zfp335 deficient mice, by careful phenotyping of the blocked. By employing adoptive transfer experiments in vivo and OP-DL1 cultures in vitro, they confirm that the requirement of Zfp335 is intrinsic to developing thymocytes. A strength of the study was the application of ChIP analysis to identify targets of Zfp335 and validating the function of two targets, Bcl6 and Rorc, by transduction into Zfp335 deficient cells to rescue the developmental block. Overall, this is an interesting study and reveals new knowledge about the regulation of the DN to DP transition in T cell development.

Reviewer #1 (Public Review):

The biggest criticism is that the data is descriptive. The authors spend a great deal of time stating the genes in each cluster, but they don't make many inferences as to the biological function of these diverse populations. For example, do these subpopulations express unique axon guidance molecules or receptors that point to the developmental and functional heterogeneity in these genetic clusters.

Reviewer #2 (Public Review):

This work addresses the developmental origins of functionally distinct neuronal populations in the arcuate nucleus of the hypothalamus (ARH). During gestation, immature Pomc-expressing neurons differentiate into at least 3 subpopulations of mature POMC neurons, as well as non-POMC neuronal sub-types (eg., AgRP and KNDy neurons). The authors set out to address the issue of whether these diverse populations arise from a common progenitor or from multiple, molecularly distinct progenitor populations.

They performed single cell RNA-seq on Pomc-expressing neurons (FACS-purified on the basis of expression of a Pomc-driven reporter transgene) across embryonic and early postnatal stages (E11.5 to P12). They also compared these transcriptional profiles to translational profiles of Pomc-expressing neurons at P5 and P12 generated with the TRAP-Seq approach. Clustering and developmental trajectory analyses confirm reports by other groups that immature Pomc-expressing neurons give rise to non-POMC cell fates (including AgRP/NPY and KNDy neurons) and that terminal differentiation of POMC neurons is achieved after P12. While the data generated here will be a useful resource for the field, there are weaknesses in the analyses used to support the central claim that POMC neurons arise from heterogenous progenitor populations.

Strengths<br> • This is an interesting topic that would be of interest to scientists studying neural circuits regulating energy balance.

• The expression databases provide a valuable resource for the community. They can be mined to identify genes that can be used as markers and as the foundation for functional studies. They can also inform efforts to generate and stage specific ARH cell types using induced pluripotent stem cell technology.

Weaknesses<br> • My main concerns stem from the fact that all Pomc-expressing neurons in the developing ARH are considered as a single category of "progenitors" in these analyses. While they meet this strict definition because they do not express markers of terminally differentiated POMC neurons, the failure to distinguish between early and late progenitors limits the conclusions that can be drawn.<br> o The earliest stage analyzed here is E11.5, which represents the peak of POMC neuronal differentiation. To capture the precursors of these neurons (1.Pomc-hi/Prdm12 at E11.5), it is necessary to perform transcriptomic analyses at earlier stages.<br> o EBFs have been shown to regulate neuronal differentiation and migration out of ventricular layer into mantle layer. It is critical to determine whether EBF-expressing neurons in the ARH similarly represent an "early" progenitor stage that follows cell cycle exit migration out of the ventricular zone and precedes the expression of transcription factors that specify a particular cell fate. If so, EBF-expressing neurons in 2.Pomc-med-Ebf1 could represent progenitors of 1.Pomc-hi-Prdm12 neurons. In support of this idea, the transcriptome of 2.Pomc-med-Ebf1 subcluster 1 neurons map onto the two major subpopulations of POMC neurons (Supplemental Figure 15).

• Because key terminal markers of POMC neurons are not expressed at P12 (i.e. Ttr, Anxa2), it is hard to precisely map progenitor populations onto neuronal subpopulations in the adult.

• While the data support the idea that there are several molecularly distinct subpopulations of POMC progenitors, these analyses do not provide clear answers to the following key questions: 1) Do AgRP/NPY and KNDy neurons arise from molecularly distinct populations of Pomc-expressing progenitors? 2) At what point in the developmental trajectory are molecularly distinct subpopulations of POMC neurons specified?

Reviewer #3 (Public Review):

This study reports the transcriptional program of develop of multiple subtypes of POMC neurons, which are important for multiple physiological and behavioral functions, including appetite.

The paper uses cutting edge scRNA-Seq methods and analysis to reveal the develop of these neurons. This work is likely to be important for research studying the physiology of the arcuate nucleus and those interested in defining the underlying neuronal components based on their transcriptional similarity and differences.

Joint Public Review:

ATM is critical for double stranded DNA breaks response, yet, how ATM is activated for this response is poorly defined. Warren and Pavletich have used cryo-electron microscopy to structurally study the interaction between ATM and Nsb1, key proteins which sense DNA-double strand breaks and subsequently activate cell cycle checkpoints and homology-directed repair. They have solved cryo-EM structures of ATM with and without a peptide of Nbs1 bound, both to 2.5 A resolution. In the ATM alone dataset, they were able to subclassify the data to reveal an open and closed state of the protein. These structures are in line with those previously published, but the authors give a detailed account of the dimeric assembly. In the dataset which included the Nsb1 peptide the authors were able to solve a cryo-EM map of ATM with extra density corresponding to 10 residues of the C-terminal FxF/Y motif of Nsb1. They were able to model into this density and locate the peptide within a conserved hydrophobic cleft within the Spiral domain of ATM. They have also used mutagenesis and biochemical assays to assess the importance and specificity of this interaction. Kinetic assays reveal that MRN-dsDNA complex significantly stimulates ATM phosphorylation activity, consistent with the yeast Tel1/MRX-DNA results. Kinetic data also shows that the disruption of Nbs1 peptide binding site does not seem to greatly affect this activity. These data extend the current understanding of how the MRN (Mre11-Rad50-Nbs1) complex may recruit and activate ATM at sites of DNA-double strand breaks, but still leaves the questions of how Rad50 and Mre11 may interact and activate ATM.

Strengths:<br> This manuscript displays solid well-presented structural and biochemical data. Specifically, the cryo-EM data collection and analysis is complete with detailed workflows describing sub-classifications and data processing to obtain high-resolution maps of ATM with and without the Nsb1 peptide. Even without the novelty of the Nbs1 peptide binding, the map resolution is an improvement upon previously deposited ATM structures. This has improved accuracy and completeness of structural modelling of this protein.

Weaknesses:<br> Using cryo-EM to visualise small peptides binding to a large protein is challenging. Although the cryo-EM data is solid, the overall resolution of the map is 2.5 A but the density corresponding to the peptide is much lower. However, incorporating the mutagenesis biochemical data strengthens the conclusion that the peptide is binding in this region.

An overall strength of this paper are the high-quality ATM structures and identification of the Nbs1 FxF/Y peptide binding site in the Spiral domain. The authors also define key structural features of different domains and detail interactions along with the structure analyses. The manuscript could be strengthened by more data or discussion on the hypothesis for the mechanism of ATM activation. Yet, in its current form, the high-quality structure and expert structural analyses, which detail interactions and insights, result in high impact data and report that will be foundational for future studies.

Joint Public Review:

One of the most important questions in the field of genome stability is the heterogeneity of damage and repair kinetics for specific types of lesions within the genome due to chromatin structure and transcription factor binding sites. To this end several groups have begun to develop genome wide approaches to follow DNA damage and repair. This current study by an excellent team follows up an initial study published by this group in (Mao et al, Genome Res in 2017), which developed and validated methods for analyzing 7meG and 3meA formation and repair from the yeast genome after the alkylating agent methylmethane sulfonate (MMS). This original study showed that nucleosome free regions were repaired more rapidly than nucleosome containing regions. After looking at the distribution of lesions after one dose of MMS the authors studied the relatively levels of 7meG at the two transcription factors, Abf1 and Ref1 binding sites across the genome from cells versus naked DNA treatment. The authors note a 40% and 70% reduction in 7meG lesions using a 5 bp sliding window in Abf1 and Ref1, respectively. They also note that the high occupancy Ref1 sites showed this decrease in 7meG lesion formation, whereas the low occupancy sites did not. However, high resolution mapping of these two binding sites indicated as much as a 1.5 increase at specific positions within the binding sites, and Reb1A showed an increase in 3meA at a -3 position. The kinetics of repair were examined at 1 hr and 2 hr in both WT and Mag1 deficient strain. The authors confirm and validate their findings using two additional genomic approaches ORGANIC and ChIP-exo and an additional TF binding protein, Rap1. Using a robust approach involving anchor-away strains which rapidly deplete the levels of Abf1 and Reb1, indicated that the differences in the MMS-induced lesions noted in these binding sites were lost. Repair kinetic analysis suggest that removal of 7meG and 3meA, is suppressed within the Abf1 and Reb1 binding sites. High resolution analysis suggests that the +4 position of the Reb1 high occupancy binding site is refractory to repair. In order to test this hypothesis, the authors show in a biochemical assay that a substrate containing inosine is poorly removed by Mag1/Ape1 when bound by Ref1. In the final set of experiments the authors next nicely demonstrate that removal of UV-induced CPD is also influenced by TF binding extending further beyond the influence observed for alkylation repair.

The results presented in this study provide an important new information on the effects of two transcription factors on formation and repair of damage at hundreds of binding sites within the yeast genome. The results may contribute to explaining why cancer mutations are frequently elevated at TF binding sites. The manuscript is well-written, however, the way some of the experimental results were split in the main text and supplemental figures made the overall manuscript a bit cumbersome to work through and the authors miss an opportunity for a nice summary figure showing high resolution mapping of lesions and repair kinetics within the Reb1 binding sites. The data analysis would benefit from rigorous statistical analysis.

Reviewer #1 (Public Review):

This study is about the development of a novel assay based on droplet digital PCR (ddPCR) to detect malaria parasites capable of evading rapid diagnostic test (RDT) through the deletion of histidine-rich proteins (HRPs) commonly recognised by the RDTs. The HRP proteins are encoded by hrp2 and hrp3 genes. The ddPCR is probe-based assay targeting four regions in three genes: hrp2 (two regions), hrp3 and tRNA.

The strength of the assay lies on its capability to detect hrp2 and hrp3 deletions in samples with multiclonal (more than one clone) infections. The assay also is capable of measuring the DNA concentration using absolute quantification.

The study has several weaknesses. Firstly, the assay lack internal control to monitor variations caused by external factors such as pipetting and other handling steps such as during DNA extraction. Secondly, data for validation of the hrp3 assay is missing - HRP3 protein is known to give signal in RDT in the absence of HRP2, particularly when the parasite density if high. Therefore, it is equally important to validate the performance of the assay in detecting hrp3 deletion. Thirdly, the authors used "genome" copy to assess the performance of the assay and it is hard to compare its performance with other assays such as nested PCRs and quantitative PCRs that use parasite density for performance evaluation. Fourthly, the presence of more than one clone in the samples was not verified by other methods, such as by qPCR or genome coverage, and it is difficult to validate the accuracy of the ddPCR. Finally, no data is available to show that both hrp2 and tRNA assays have similar or the same amplification efficiency permitting the use of the assays for calculating the ratio DNA concentration to determine deletions in multiclonal infections.

The assay has the potential to be used as a tool to rapidly detect RDT-evading malaria parasites and contribute to control and elimination programs.

Reviewer #2 (Public Review):

Vera-Arias et al developed and evaluated a new molecular method using droplet digital PCR (ddPCR) for surveying hrp2/hrp3 deletions in Plasmodium falciparum parasites. These deletions are problematic because they produce false negative results when rapid diagnostic tests (RDT) are used in endemic areas. In this study, they compared the new ddPCR method with existing nested PCR (nPCR) method and found the new ddPCR method has increased sensitivity and accuracy and recommended the use of ddPCR. The conclusions of this paper are well supported by data.

The strength of this paper is that the authors have demonstrated the use of this new assay in mixed culture in experimental setting and extended this evaluation in field isolates as mixed infections are the common occurrence in the endemic areas. An assay with high sensitivity to detect these key deletions amongst mixed infection is particularly important in monitoring the true prevalence of these deletions in low-density transmission settings.

The weakness of this paper is that the authors did not evaluate against the existing real-time PCR methods. A good and extensively validated real-time PCR assay can sometimes work as effectively as a ddPCR assay with similar sensitivity. Some data from real-time PCR can be useful in term of comparison existing techniques to novel ones.

Reviewer #1 (Public Review):

In this study the authors use viral tracing and optogenetic manipulations to demonstrate the causal implication of a vHIP-AH circuit mediating both contextual fear memory retrieval and shelter-directed escape responses. The authors first show that photostimulation of AH neurons induces escape, aversion, and conditioned place avoidance and subsequently show that this effect is specifically mediated by a direct monosynaptic excitatory input derived from ventral hippocampal structures. By targeted optogenetic inhibition, they elegantly show that inhibition of a HPC to AH projection results in impaired shelter seeking similar to the one observed when an aversive stimulus (20 Hz sound) is delivered in the absence of a shelter. These findings are not unexpected as the AH has long been postulated to be the gateway structure within the medial hypothalamic defensive system for hippocampal inputs conveying contextual information. Here, the first causal evidence confirming this hypothesis is provided. The experiments are well designed, technically sound and methodological details well reported. Statistical analysis is mostly appropriate. However, in Fig. 6J and 8M TWO-WAY ANOVA should be performed instead of t-test. In particular, for Figure 8M, I don't believe that the obtained results support the conclusion and the results section should be changed accordingly. In conclusion, I believe that this manuscript, although not unexpected, substantially contributes to our understanding of how context encoding structures like the hippocampus contact the medial hypothalamic defensive system and guide innate and learned behavioral responses.

Reviewer #2 (Public Review):

I read with great interest the manuscript of Bang and colleagues and I recognize the importance of the work on addressing missing points in the field. The questions of how the brain computes the contextual information produced by different aversive innate scenarios and the contextual memory induced by innate threats are addressed with several behavioral approaches. The authors combined multiple behavioral approaches with optogenetics, tracing and slice recordings to reveal a new hippocampus-hypothalamus circuit (HPC-AHN) that is involved in escape response and seems to play a role in contextual information.<br> Although I recognized the hard work of the group, the importance of the questions raised as well as the quality of the experimental approach, there is a lack of clarity in the description of the methods, fundamental for interpretation of the data and reproducibility. Moreover, some interpretation of the data at the light of the results and the literature needs clarification.

Major general suggestions:

1) The study provides several interesting and innovative behavioral tasks that will definitely be of great interest to the behavioral neuroscience field. However, the description of the methodology for all experimental procedures requires an extensive revision. The way the methods are presented it is very unlikely other labs will be able to replicate with accuracy the procedures.

1a) Information such as, total task time, number of trials; time difference among the escapability tasks (min, hours, days?) are missing. This is important for interpretation of whether the stimulation in one situation could affect the other.

1b) It is not clear if the same animals were used among the different tasks and frequencies (6 or 20hz). I believe that clearly stating whether same or different animals were used is extremely important to the interpretation of the results. If the same animals were used for both frequencies and tasks, an extensive discussion about how the previous stimulation or tasks influences the effect of the second.

1c) The way some of the behavioral quantification was done it is not clear or even not mentioned in the methods. For example, Fig 1 missing explanation of speed increase "fold", not sure what the fold means, how was the comparison made? Also lacking a clear explanation for the real-time indexes. The paper will benefit a lot if all behavioral measures are explicitly and clearly stated preferentially with the formulas used for the calculation, even if they are simple math operations. In igure 6, it is not clear what was defined as a "run".

2) Another very important point that requires further clarification is the type of memory being tested. Sometimes the study has protocols that test for short-term memory whereas other protocols are testing long-term memory. It is not clear if the interpretation presented by the authors is related with memories that require consolidation or simply by spatial navigation.

3) In addition, the role of the circuit HPC-AHN seems to be general for escape responses and does not depend on memory consolidation. The activation of the pathway produces escape responses even in absence of memory consolidation (Fig 4, Fig 7 and Fig 8). Authors need to clarify why their interpretation for the role of the circuit is experience-dependent, once the manipulation of the circuit also promotes escape behavior in a novel environment. It is interesting that the circuit can induce escape response in a novel environment, but when the memory is consolidated the escape response is more organized. The authors discussed the gap between the present study and previous ones that suggest PMd is fundamental for innate escape response and also contextual fear induced by innate threats. However, I do suggest for the authors to also review Wang and Schuette, et al (Neuron 2021 - PMID: 33861942) and try to reconcile their data with the Wang and Schuette study where they show the important role of PMd, but not other hypothalamic nuclei, on context-specific and panic-related escapes. In this paper the group discuss organized versus non-organized escape response. I believe the HPC-AHN circuit could be contributing to this kind of organized context-dependent space response, but also play a role in an unknown context inducing disorganized avoidance. This circuit might get potentiated with experience and therefore produce goal-directed (organized) escape behavior. However, it is not clear if the animals had a mental schema of the Fig 7 and 8 mazes. If animals were not exposed to the mazes, how do the authors reconcile the context-dependent experience interpretation with the fact the environment was new to the animals?

4) Authors must include the individual data in all the plots (e.g. dots for each one of the animals), individual variability is extremely informative and can be explored further by the field, and brings more transparency for the data.

Reviewer #1 (Public Review):

In the present study, the authors first analyzed simultaneously recorded human EEG-fMRI data and found the fMRI signatures of burst-suppression. Then, they reported such burst-suppression fMRI signatures in the other three species examined: macaques, marmosets, and rats. Interestingly, their results indicated an inter-species difference: the entire neocortex engaged in burst-suppression in rats, whereas most of the sensory cortices were excluded in primates. The fMRI signatures of burst-suppression were confirmed in several species, suggesting that such signature is a robust phenomenon across animals. These findings warrant further investigation into its neural mechanisms and functional implications.

Major Issues:

1) One of the major findings is that burst-suppression in primates appeared to largely spare sensory cortices, especially V1. However, as seen in the tSNR map for macaques and marmosets (Figure 3 &4 -figure supplement 4), the tSNR around the primary visual cortex was much weaker than other cortices. Moreover, in marmosets, the EPI slices did not cover the entire brain and actually left most of the V1 uncovered as seen in Figure 4. If so, the authors should draw their conclusions very carefully when talking about the differences in V1 across species. It would be better to analyze and discuss how the tSNR differences affect their findings. For example, the author may consider including the tSNR as covariance in their map analysis.

2) To confirm their findings, it would be great to look into the EEG signals around the sensory cortex (e.g., V1) to see whether the findings in fMRI could be also confirmed with EEG.

3) As seen in Figure 2-figure supplement 2, there was a significant anticorrelation with burst-suppression at the ventricular borders. It is unclear whether the authors have done physiological or white matter/CSF/global nuisance regression as most of the rest-fMRI studies did. Please make it clear. If not, please explain why and discuss whether it would affect their results.

4) Three different concentrations of the anesthetic sevoflurane were chosen for human participants. The authors found that the high concentration (3.9-4.6%) induced burst-suppression much better than the other two lower concentrations as expected. However, in rats, almost all asymmetric PCs were found at an intermediate concentration (2%) of isoflurane less at the low (1.5%) or high (2.5%) concentration in Rat 1. At the same time, all fMRI runs from Rat 2 with a 1.3% concentration of isoflurane had a prominent asymmetric PC. That is, it seems that only the high concentration of isoflurane could not induce burst-suppression well in rats, which was opposite to those findings in humans. The authors may explain what reasons may cause such differences and whether such differences may affect the major findings in differences between primates and rodents.

Reviewer #2 (Public Review):

The strong point in their manuscript is the originality of their results. Using the fMRI's spatial resolution, they can successfully reveal that not all brain areas are synchronized during the burst suppression. Furthermore, they can find that the difference is the most obvious when comparing primates with rats, which makes sense considering the distance on the phylogenetic tree. As far as I know, this manuscript first reports these points.

On the other hand, there is a weak point in their method. As they've already discussed this point, they needed to use arbitrary thresholds to evaluate whether there is burst suppression or not. Furthermore, this study cannot reject the possibility of spatial inhomogeneity and/or anesthesia-specific modulation in hemodynamic response. If there is such a mechanism, one can find different results from those obtained through electrical measurements.

1) The authors found that some sensory areas in primates are excluded from those highly synchronized during the burst suppression. While it is true, I wonder if each voxel in such areas shows burst suppression-like activity that is not synchronized with others. If this is the case, burst suppression can still be a global phenomenon. Though authors seem to investigate this point, they used in-ROI averaged time-series so that it cannot reject the possibility that each voxel inside the ROI is not synchronized but shows burst suppression in its manner. I recommend the authors look into each voxel if this is the case or not.

2) The other but similar point is about their way to detect burst suppression. Why did they use the principal component? By definition, burst suppression should be defined by the existence of burst and suppressed periods. I cannot understand why they did not simply use this definition to check whether each voxel shows such an intermittent activity to evaluate whether it is a global phenomenon or not.

3) Why is there no synchronization during the slow-wave states under light anesthesia? During the slow-wave sleep, it is shown that the entire cortical network is decomposed into a modular-like network structure. Is there synchronization inside each module while no synchrony between modules?

Reviewer #3 (Public Review):

The authors present a multicenter, multimodal rs-fMRI study of the spatial signature of burst suppression in the brain of humans, non-human primates and rats. They have used EEG to identify burst suppression activity in human data from simultaneous EEG-rs-fMRI measurements of subjects under servoflurane anesthesia. After having identified a (neurovascular) rs-fMRI representation of burst activity, the authors show that bursts can equally be identified from MR data alone. After a principal component analysis, bursts and their spatial signature were identified by an asymmetry of the correlation coefficients. Across species the authors identified similar spatial signatures, which were conserved for all (investigated) primates, but differed for rats. While rats showed a pan-cortical involvement, signatures in primates were more complex, e.g., not including the visual cortex.

In this study, the authors have presented a novel purely MR-based method to identify burst suppression and its spatial signature. Their method may be used to readily identify burst suppression in fMRI data. However, no general threshold for the median of the cortex-wide correlation could be identified. The authors also establish a conserved signature of burst suppression for primates and reveal subtle but important differences to rodents. Both achievements are novel and represent a major advance in the field of neuroimaging.

The study was well designed, including important control data to rule out artefacts as source of the observed burst suppression patterns. The particular strengths of this study are: (1) including multicentre data (although only rats were scanned at two different sites); and (2) including four species from humans to rats.

The manuscript was very carefully and well written (I did not even notice a single typo) and the figures were carefully devised, comprehensively illustrating the large amount of data. The authors further provide a comprehensive account of the relevant literature. Towards the end of their discussion they also clarify the difference in terminology used for burst suppression in some recent rodent studies.

The only (and in my opinion notable) weakness, is the lack of a general threshold for the asymmetry of the median of the cortex-wide correlation coefficients. With such a threshold, rs-fMRI could be readily used to automatically detect burst suppression across species. However, the authors clearly state this shortcoming and openly discuss its implications. I do not think that an altered experimental design or additional data could provide further remedy.

To conclude: This very comprehensive study was very well designed, extremely carefully performed, presents a novel tool for identification of burst suppression, and provides insight across species. It has clearly translational potential, which however, is limited by the lack of a general threshold for burst suppression detection.

I congratulate the authors for this very nice piece of work, and the most typo-free manuscript I have ever read.

Reviewer #1 (Public Review):

This paper tackles an important problem for cognition and neuroscience. While episodic, semantic, and working memory clearly all interact together in behavior, they have typically been studied independently, and I am unaware of any prior modeling work considering how they might work together so that controlled retrieval of episodes together with semantic processing can support sequential inference/behavior. The proposed model integrates insights from much prior work in a comparatively simple framework; simply getting the systems to work together in a plausible and intuitive way is an impressive accomplishment, and the model's connection to recent imaging results from studies of encoding and retrieval is also promising. For the most part the paper is clearly written, anticipates and addresses important questions as it unfolds, and situates both the novel contributions and open questions fairly with respect to the rest of the literature. I think this is an important contribution that will help move the field toward more integrated models of memory and behavior. I offer the following thoughts in hopes of further strengthening a very interesting paper.

1) I am still not 100% clear on how the sequences were constructed and how these relate to the structure of real schematic events. From the methods I think I understand that the inputs to the model contain a one-hot label indicating the feature observed (e.g. weather), a second one-hot vector indicating the feature value (e.g. rainy), and a third for the query, which indicates what prediction must be made (e.g. "what will the mood be"?). Targets then consist of potential answers to the various queries plus the "I don't know" unit. If this is correct, I was unsure of two things. First, how does the target output (ie the correct prediction) relate to the prior and current features of the event? Like, is the mood always "angry" when the feature is "rainy," or is it chosen randomly for each event, or does it depend on prior states? Does the drink ordered depend on whether the day is a weekday or weekend---in which case the LSTM is obviously critical since this occurs much earlier in the sequence---or does it only depend on the observed features of the current moment (e.g. the homework is a paper), in which case it's less clear why an LSTM is needed. Second, the details of the cover story (going to a coffee shop) didn't help me to resolve these queries; for instance, Figure 2 seems to suggest that the kind of drink ordered depends on the nature of the homework assigned, which doesn't really make sense and makes me question my understanding of Figure 2. In general I think this figure, example, and explanation of the model training/testing sequences could be substantially clarified.

2) The authors show that their model does a better job of using episodic traces to reinstate the event representation when such traces are stored at the end of an event, rather than when they are stored at both the middle and the end. In explaining why, they show that the model tends to organize its internal representations mainly by time (ie events occurring at a similar point along the sequence are represented as similar). If episodes for the middle of the event are stored, these are preferentially reinstated later as the event continues following distraction, since the start of the "continuing" event looks more like an early-in-time event than a late-in-time event. This analysis is interesting and provides a clear explanation of why the model behaves as it does. However, I wonder if it is mainly due to the highly sequential nature of the events the model is trained on, where prior observed features/queries don't repeat in an event. I wonder if the same phenomenon would be observed richer sequences such as the coffee/tea-making examples from Botvinick et al, where one stirs the coffee twice (once after sugar, once after cream) and so must remember, for each stirring event, whether it is the first or the second. Does the "encode only at the end" phenomenon still persist in such cases? The result is less plausible as an account of the empirical phenomena if it only "works" for strictly linear sequences.

3) It would be helpful to understand how/why reinforcement learning is preferable to error-correcting learning in this framework. Since the task is simply to predict the upcoming answer to a query given a prior sequence of observed states, it seems likely that plain old backprop could work fine (indeed, the cortical model is pre-trained with backprop already), and that the model could still learn the critical episodic memory gating. Is the reinforcement learning approached used mainly so the model can be connected to the over-arching resource-rational perspective on cognition, or is there some other reason why this approach is preferred?

Reviewer #2 (Public Review):

This article presents a network model with multiple regions trained with deep reinforcement learning techniques (Mnih et al. 2016) to address the timing of encoding and retrieval for episodic memory. The authors relate the model results to both behavioral data and functional imaging of the timing of hippocampal-cortical interactions at the transition between events. In many early memory experiments human participants were told when they should encode or retrieve information, but neural models have previously demonstrated the importance of regulating the timing of encoding and retrieval into episodic memory (Hasselmo and Wyble, 1997; Zilli and Hasselmo, 2008). The model presented here focuses on learning of actions that gate encoding of information into episodic memory (EM) in a range of different tasks to be used by a neocortical network that has semantic memory in terms of its connectivity and working memory in terms of its ongoing activation, but does not have episodic memory represented in the neocortical circuits. Thus, the neocortical circuit requires gating of episodic memories for effective performance in certain components of behavior tasks. The behavioral tasks being simulated include the task from a recently published task (Chen et al., 2016) in which subjects viewed an episode of the twilight zone in two parts, either on the same day (Recent memory RM) or on two different days (distant memory DM) or only viewed part 2 (no memory). The activation patterns in the model demonstrate that retrieval is effectively used more in the distant memory condition when episodic retrieval from the previous day (rather than working memory from the same day) is necessary for performance of the task. The authors also show that unnecessary episodic retrieval can impede performance by recalling irrelevant information. The results also show interesting effects demonstrating that performance is better when encoding is delayed until the end of an episode, because mid-episode encoding can overshadow the later encoded memory. They effectively link these results on the timing of encoding to data showing activation of hippocampal circuits for episodic encoding at transitions between episodes (Ben-Yakov et al., Reagh et al. 2020). The model addresses important aspects of the control of episodic memory function, and some of its shortcomings are already addressed in the discussion, but there could be further discussion of these points.

Major comments:<br> 1. Line 566 - "new frontier in the cognitive modeling of memory." They should qualify this statement with discussion of the shortcomings of these types of models. This model is useful for illustrating the potential functional utility of controlling the timing of episodic memory encoding and retrieval. However, the neural mechanisms for this control of gating is not made clear in these types of models. The authors suggest a portion of a potential mechanism when they mention the potential influence of the pattern of activity in the decision layer on the episodic memory gating (i.e. depending on the level of certainty - line 325), but they should mention that detailed neural circuit mechanisms are not made clear by this type of continuous firing rate model trained by gradient descent. More discussion of the difficulties of relating these types of neural network models to real biological networks is warranted. This is touched upon in lines 713-728, but the shortcomings of the current model in addressing biological mechanisms are not sufficiently highlighted. In particular, more biologically detailed models have addressed how network dynamics could regulate the levels of acetylcholine to shift dynamics between encoding and retrieval (Hasselmo et al., 1995; Hasselmo and Wyble, 1997). There should be more discussion of this prior work.<br> 2. Figure 1 - "the cortical part of the model" - Line 117 - "cortical weights" and many other references to cortex. The hippocampus is a cortical structure (allocortex), so it is very confusing to see references to cortex used in contrast to hippocampus. The confusing use of references to cortex could be corrected by changing all the uses of "cortex" or "cortical" in this paper to "neocortex" or "neocortical."<br> 3. "the encoding policy is not learned" - This came as a surprise in the discussion, so it indicates that this is not made sufficiently clear earlier in the text. This statement should be made in a couple of points earlier where the encoding policy is discussed.

Reviewer #1 (Public Review):

Strengths: This is an elegant study on the systematic assessment of divergence of duplicated genes encoding transcription factors that originated from the whole genome duplication in yeast. It provides novel insight into the evolution of DNA binding preferences, the contribution of different domains in the promoter as well as protein interactions between paralogs. The experimental ChEC-seq datasets are a valuable resource to the community.

Weaknesses: Transcription factor binding domains do not directly translate into transcriptional output. This represents a limitation of this study as it does not provide any immediate insight into divergence of transcription factors based on the actual transcriptional output.

Reviewer #2 (Public Review):

Gera, Jonas and colleagues used paralog pairs that derived from a whole genome duplication to investigate how newly evolved transcription factors gain new binding specificity using CRISPR/Cas genome editing and DNA binding profiling using high-throughput sequencing. This approach proved to be useful. They find that paralogous transcription factor pairs typically retained very similar binding profiles and occupied the same or similar sites, with few notable exceptions. Surprisingly, divergence in binding preference was driven mostly by sequence variation outside of the proteins' DNA binding domains. Only one transcription factor family (zinc cluster factors) repeatedly showed some influence of the DNA binding domain on binding preference. They also showed if and how transcription factor paralogs influence DNA binding of their respective partner, additively, cooperatively or competitively. For paralog pairs that did diverge in binding specificity, the authors show that divergence is typically a mix of sub- and neo-functionalization, in a model where pairs often retain some ancestral specificity and in addition gain new binding sites.

The claims and conclusions of the paper are well supported by a wealth of data. Experiments were targeted strategically at addressing the authors' key questions. The probably most surprising result comes from the observation that binding site divergence is mostly due to variation in sequences outside of the proteins' DNA binding domains. This is all the more surprising as some of the pairs show quite a few binding domain sequence differences. The authors provide strong evidence in favor of this observation by swapping DNA binding domains between paralogs. This shows that the backbone of the protein, and not the DNA binding domain, is mainly responsible for determining where a factor binds. One drawback of this observation is that it is not clear what exactly causes the divergence in binding specificity. It could be possible that it is due to residues in the immediate proximity of the DNA binding domain, which could mean that they influence properties of the binding domain. This would mean that the DNA binding domains play a major role after all, albeit through the work of other residues. Another possibility is that binding specificity is determined by interaction with co-factors. The authors raise the possibility that other protein domains directly access DNA which confers binding specificity. The data presented cannot resolve this and further experiments are needed. However, the observation that swapping the DNA binding domain between paralogs does not lead to a change in binding (in most cases) is a major, unexpected and interesting finding.

The authors then address the big question of the relative importance of sub- vs neo-functionalization of protein duplicates. They compare the paralogs' binding profiles to orthologous factors from a species that did not undergo the genome duplication, providing an "ancestral" baseline. This comparison shows that one of the paralogs often retained a more similar binding profile to the ancestral state than to the other paralog. In most cases, the paralog that had undergone more protein sequence evolution showed the more divergent binding profile. While paralogs with little diverged binding profiles generally bound to the very same promoters as the "ancestral" ortholog, paralogs with more divergent profiles showed increasingly new binding sites, pointing to neo-functionalization to be the main outcome at high levels of divergence. The caveat here is that many proteins did not diverge much at all with the consequence that the sample size for this analysis is not large. Paralog pairs that did not diverge much at all must be considered "non-divergent" and hence uninformative for this question. This analysis appears to lack a clear definition of when a paralog pair is considered divergent enough to be assessed, which causes the potential issue of conflating non-divergence with sub-functionalization. This potential issue needs to be clarified.

The authors present another exciting result, where the deletion of one of the paralogs leads to reduction (in two cases ablation) or expansion of the binding profile of the other paralog. The data presented is clear and immediately offers models of how the paralogs interact with each other to result in these binding profile changes. This experiment is insightful but there are some caveats to keep in mind. Few of the paralogs had any effects of deletion of one of the pair, meaning that the sample size is rather small and it is unclear how general this pattern is. In addition, some of the factors might show similar effects upon deletion of other interaction partners. This is irrelevant regarding divergence and interaction of paralog pairs, but it means that this data cannot be used to learn about the importance of binding partners for binding specificity in general terms.

Overall this paper provides exciting insights based on extensive experimental data that advance our understanding of transcription factor binding preference evolution in both expected and unexpected ways.

Reviewer #3 (Public Review):

In their manuscript entitled "Evolution of binding preferences among whole-genome duplicated transcription factors" the authors investigate the evolution of transcription factors following genome duplication including the implication for the evolution of their binding sites.

The hypothesis tested in this study is very much of interest to a large readership, furthermore the authors have carefully selected approaches to allow them to assess the different aspects of the implications of gene duplication. This includes reciprocal knockouts of paralogues, DNA binding domains swapping.

One of my major criticism stems with the writing and the figure preparation. The writing is extremely dense making difficult to understand arguments and results. Similarly the figures have many panels at the cost of readability and ease of comprehension. Some of the supplementary figures are better suited for the main manuscript for instance

The methods need attention, as for instance the parameters used to identify cut sites and avoid noise are not provided are not explained (NGS data processing).

"Promoter binding quantification" The authors summed the normalized genome coverage over the promoter region of each gene, by doing so the authors remove potential inherent variability between replicates, the authors need to keep replicate separate and thereby gain confidence in their results.

"Relative, gene-specific binding changes upon paralog deletion or DBD swapping" the authors select signal from the 100 strongest bound promoters, why choosing those top 100 and not all promoters?

Need more info on 2A the authors need to indicate conservative and radical replacement. It is hard to see the impact of the changes highlighted

"Consistent with our sequence analysis, DBD swapping perturbed binding for three of the four zinc-cluster TFs tested, although in none of these cases was DBD swapping sufficient for swapping promoter preferences (Fig. 2D,E). The fourth TF (Yrr1) remained largely invariant (Pearson's r>0.8) to DBD swapping, as did the eleven additionally tested TFs, taken from six different families. Of note, this invariance to DBD swapping characterizing most TFs was observed not only when comparing promoter preferences, but also when comparing in-vivo preferences to DNA 7-mers (Fig. 2D and fig. S2). We conclude that, for most duplicate pairs, the variations driving divergence". This section is very difficult to read

Page 6 "The fourth TF (Yrr1)", this transcription factor is not shown on the figures

Page 7: "Two TFs completely lost binding signals (Pip2, Hms2)," how much is this impacted by the number of binding sites

Page 7: "Cooperative interactions were generally minor (e.g. Stp2), as were compensatory interactions (e.g. Pdr3 or Ecm22; Fig. 3C,D). Therefore, strong interactions between TFs paralogs are rare and existing ones tend to increase mutation fragility." Figure 3D highlights gene specific variation upon domain swapping, yet the authors do not explore this

Page 11: "Second, mutations within Rph1's DBD prevented its binding to Gis1-specialized sites, thereby reducing paralog interference (Fig. 6E)." Figure 6D suggest Rph1 to bind upon Gis1 ko, the difference in binding affinity upon swapping is not massive, more variants should also intervene

Reviewer #1 (Public Review):

LIS1 is a key dynein regulator and details about its mechanism of action have captivated many in the field. In this manuscript, the authors reported the first high-resolution (3.1Å) structure of the dynein-Lis1 complex using yeast proteins including the dynein motor domain and LIS1 dimer. This high-resolution structure allows the authors to reveal new details of interaction interfaces, specifically the LIS1-binding site on AAA5, the LIS1-binding site on the stalk as well as the contact site between the two LIS1 propellers (one binds the ring and the other binds stalk). The authors made specific mutations (in LIS1 or dynein) to investigate the significance of these interactions. Finally, they also made mutations in human dynein that affect LIS1 binding to the same sites (ring and stalk) and investigated the functional significance of the involved amino acids in cargo-adapter- and dynactin-mediated dynein activation/motility assay. Their results suggest that these sites are important for the formation of active human dynein complexes.

Major Strengths of the methods and results:

Overall, the authors have done a very impressive and beautiful series of experiments. Solving such a high-resolution structure of the dynein-LIS1 complex is a great accomplishment and will be highly impactful. Another major strength is the combination of in vivo and in vitro functional analyses on the significance of the amino acids at the contact sites. The functional study was done by using multiple assays both in vitro (binding curve and effect on in vitro velocity) and in vivo (nuclear positioning and dynein localization in budding yeast). This combination is very powerful, as it allows the authors to reveal separation-of-function mutations. One mutation affecting the interaction between LIS1 and the dynein stalk is particularly significant (please see my detailed comments).

Major weaknesses of the methods and results:

In vitro assays were done to test whether the mutations affect LIS1's ability to induce a tight dynein-microtubule binding or an inhibition of dynein motility in the presence of ATP, while Marzo, et al., 2020 NCB has shown that yeast LIS1 enhances dynein processivity if salt concentration is increased for the in vitro motility assay.

More detailed comments:

1. Currently, the model that the authors proposed emphasizes the idea that the LIS1-induced tight dynein-microtubule binding enhances the microtubule-plus-end accumulation of dynein, which in turn promotes active complex formation. This is inconsistent with experimental findings: In the budding yeast, the microtubule-binding domain is not required for the plus-end accumulation of dynein (Lammer and Markus 2015). In mammalian cells (as well as filamentous fungi), dynactin is needed for the plus-end accumulation of dynein (Xiang et al., 2000; Zhang et al., 2003, 2008, Lenz et al., 2006; Egan et al., 2012, Splinter et al., 2012., Yao et al., 2012) and this is also true in vitro in reconstituted systems (Duellburg et al., 2014; Baumbach et al., 2017; Jha et al., 2017). In addition, it is unclear how LIS1 would affect the mechanochemistry of dynein to induce a tight binding of dynein to microtubules (aside from inducing the open state).

2. Data were presented to argue for an inconsistency with a "tethering" model in which LIS1 tethers dynein to microtubules non-specifically. This was done to argue against the idea proposed by Marzo et al., 2020 that the Pac1/LIS1-caused speed reduction of dynein is due to the non-specific binding of Pac1/LIS1 to microtubules. My problems with the argument are as follows: Marzo et al showed that Pac1/LIS1 does not need to bind dynein to cause a speed reduction, since dynein complexed with Pac1 or not complexed with Pac1 gets the same kind of speed reduction. Aside from that, the dimeric LIS1 (LIS1 WT) does seem to lower the velocity more dramatically (Fig 3C), and the argument that dynein-binding and microtubule-binding cannot possibly use the same site of LIS1 only works when a monomer is considered. In this context, while the current study clearly shows the importance of propeller-propeller interaction, LIS1 monomer was shown to be effective in promoting active human dynein complex formation in vitro, suggesting that monomers at a high concentration may still lead to propeller interaction (Htet et al., 2020).

3. Lis1S248Q and Lis1F185D, I189D, R494A are capable of inducing tight microtubule binding in the presence of ATP but mutants carrying these mutations show a clear nuclear-positioning phenotype in yeast. This at least argues against the tight MT-binding being the only key effect of LIS1 in vivo. What would be the other key effects? "The weakening of dynein's interactions with microtubules when AAA3 is bound to ATP" does not seem to agree with the genetic data from Aspergillus that the wB-AAA3 mutation allows LIS1 function to be bypassed (Qiu et al., 2021 bioRxiv). It seems more likely that LIS1's effect on promoting the open dynein state is relevant in vivo as this was shown in both yeast and Aspergillus (Qiu et al., 2019; Marzo et al., 2020). These two mutants that the authors made are significant, and it would be very interesting to see if the open dynein (phi mutant) would suppress their nuclear-positioning defect.

4. The detailed analysis performed by the authors has revealed Lis1S248Q (affecting the stalk interaction) as a mutation that affects dynein function (cortical interaction and nuclear distribution) but does not affect the plus-end dynein accumulation. I would like to emphasize more explicitly the importance of this separation-of-function mutant of Pac1/LIS1 (to my knowledge, this is the first such separation-of-function mutant of Pac1/LIS1). The cortical interaction of dynein-dynactin with Num1 depends on the plus-end accumulation of dynein in budding yeast, and thus, it depends on Pac1/LIS1 (Lee et al., 2003; Sheeman et al., 2003; Markus and Lee 2011). Pac1/LIS1 recruits dynein to the plus end most likely because Pac1/LIS1 binds directly to the microtubule plus-end-tracking protein Bik1/Clip170 (Sheeman et al., 2003; Coquelle et al., 2002; Lin et al., 2001; Perez et al 1999; Markus et al., 2012). Due to the requirement of Pac1/LIS1 for the plus-end dynein accumulation, it was not easy to dissect the specific role of Pac1/LIS1 in cargo adapter-mediated dynein activation in budding yeast, although the coiled-coil domains of the cortical adapter Num1 also activates dynein to cause its relocation from the plus end to the minus ends (Lammers and Markus 2015). In Aspergillus nidulans and ustilago maydis, LIS1 is not required for the plus-end accumulation of dynein (Zhang et al., 2002, 2003, Lenz et al., 2006, Egan et al., 2012), which makes it more straightforward to find a role of LIS1 in cargo adaptor-mediated dynein activation in A. nidulans (Qiu et al., 2019). In this context, I believe that this new Lis1S248Q mutant will become a wonderful tool for the field because it will allow new assays (both in vivo and in vitro) to be performed to further study the function of LIS1.

Reviewer #2 (Public Review):

The main goal of the current study was to better define interaction sites between LIS1 and the microtubule (MT) motor cytoplasmic dynein 1 (dynein). Dynein exists in a closed, autoinhibited form and an open form capable of binding to activating factors dynactin and various cargo adaptors. These "DDA complexes" are processive, while mammalian dynein on its own is not. Cell-based studies indicated that LIS1 stimulates dynein, but work from this group and others indicated that LIS1 inhibits dynein by inducing a tight MT bound state. New data shows LIS1 binds to the open form of dynein and prevents it from switching back to the closed form, catalyzing the formation of processive DDA complexes and preventing conversion to the autoinhibited state. LIS1 also recruits 2 dynein motors into the complex, leading to more force generation and increased speeds and run lengths. The current thinking about LIS1's ability to stall dynein by inducing a tight MT-bound state is that binding at MT plus ends allows it to recruit plus-end-localized dynactin to the open dynein complex .

In previous work this group identified two PAC1 (budding yeast LIS1) interaction sites in dynein heavy chain which they call site "ring" and site "stalk". However, the resolution was not sufficient to build models to predict the interaction sites at finer resolution. In the current manuscript the resolution of Cryo-EM data was increased by biotinylating dynein and using streptavidin affinity grids, which was able to reduce dynein's strong preferred orientation on the EM grids and allowed it to be tethered to the grid in random orientations. They identified sites of contact between LIS1 and dynein and between the two LIS1 b-propellers that, when mutated, prevent LIS1 from being able to induce the formation of activated dynein complexes in yeast, and reduced LIS1's capacity to stimulate human DDA complexes. This paper clearly adds to our understanding of the precise sites of dynein - LIS1 interactions and has several strengths, but there are a few concepts that need clarification.

Strengths

• This paper represents a huge body of work and the sites identified by their approach will guide future studies from many labs.<br> • The work revealed a previously unknown additional contact site for LIS1 in dynein's AAA5 and used a yeast assay for dynein activity to show this contact site is important for LIS1 to regulate dynein in vivo.<br> • They provide evidence that the interaction between LIS1 b-propellers is required for LIS1 regulation at site "stalk".<br> • The work supports a model in which the NUM1 cortical protein in yeast can act as a cargo adaptor for dynein/dynactin complex.<br> • They show that mutations in residues they identified as important in regulating yeast dynein are also important for human dynein.

Weaknesses

The following points were not clear to me as a general reader and it would have been helpful if they were clarified

• Is it possible that the mechanism of driving orientation of the sample on the grid changes potential interactions or selects for particular interactions?<br> • It is not clear why the decision was made to generate a model with ATPs in AAA1, AAA2, and AAA3. Also, ADP was apparently bound to AAA4 in their complexes. How was this determined, and why was this the case? Does LIS1 impact ATP hydrolysis at AAA4?<br> • Binding affinities in Figure 2 , 3 and 7 all appear to utilize monomeric dynein. Is there any data regarding if or how these values would change using the dynein holoenzyme?<br> • In FIG 3 PAC1/LIS1 is predicted not to interact directly with MTs, in part because monomeric PAC1/LIS1 also slowed dynein. If it did so by crosslinking dynein and MTs, then dynein's stalk angle would need to be ~ 15˚, which is rarely seen, at least with tail-truncated monomeric dynein (Can et al, 2019). Have stalk angles on MTs been measured in the presence of either monomeric or dimeric LIS1?

Reviewer #3 (Public Review):

This manuscript reports a high resolution (~3.1 Å) cryo-EM structure of the yeast dynein-Lis1 complex. The resolution of the EM map allowed for the determination of additional dynein-Lis1 contacts that are important for Lis1's regulation of dynein. Specifically, the molecular interactions between the Lis1 ß-propellers and its two binding sites on dynein were described and probed. The authors propose a model in which Lis1 does not tether dynein to microtubules. Taken together, the authors leverage their high-resolution structure of the dynein-Lis1 complex to provide insights into how Lis1 modulates yeast dynein function. Mutations in human dynein at the same sites disrupt Lis1-mediated dynein activation.

This manuscript is extremely well written and we appreciated the effort made by the authors in laying out the manuscript and figures, as well as the introduction. The authors have also incorporated previously published data and possible models into their analyses. Technically, this is a very difficult sample for cryo EM, with severe preferred orientation, and the use of specialized streptavidin affinity grids was key in facilitating the high resolution structure. The conclusions are supported by the cryo EM and functional data presented, and the model is consistent with the data.

Reviewer #3 (Public Review):

In this study, Sims et al. perform a phosphoproteomic analysis of the ATR signaling pathway in mouse testis. By studying the different phosphorylated peptides found in testis samples from ATR inhibited mice and from mutant mice for the member of the ATR-activating 9-1-1 complex, RAD1, authors defined a comprehensive map of the ATR signaling pathway in the mouse testis. In general, the methodological approach performed is appropriate to accomplish the desired goal and the results obtained are well explained and properly discussed. The conclusions raised by the authors are supported by the results obtained and the manuscript reads easily. Thus, overall the manuscript is of high quality. Furthermore, the information provided in this study is novel since to my knowledge this is the first attempt to characterize the ATR signaling pathway in the testis. In my opinion, these data will be very relevant to better understand the role of the ATR in mouse spermatogenesis, and in meiosis in particular, in the future. 

Nonetheless, I have a few major concerns about this manuscript. Firstly, I think an important part of the description of the results is placed in a related preprint by the authors (Pereira et al. https://www.biorxiv.org/content/10.1101/2021.04.09.439198v1). In my opinion, this manuscript lacks a more detailed analysis of the ATR signaling on DNA repair and chromosome axis structure, which are fundamental to understand the meiotic prophase. Secondly, the manuscript falls short of providing novel insights about ATR roles during the meiotic prophase. As ATR function on the meiotic prophase has been extensively studied, the ATR phosphoproteome should provide either some clues about possible novel functions ATR may do during the meiotic prophase or spermatogenesis, or provide a mechanistic explanation of how ATR performs its meiotic functions (e.g., meiotic sex chromosome inactivation or meiotic recombination). The final section of the results is an attempt at doing sol, but to me, the data provided only suppose a small incremental advance in our knowledge of how ATR promotes MSCI. I would have liked the authors to expand this section to prove the utility of the data.

Reviewer #1 (Public Review):

Are enzymes found in organisms that optimally grow at lower temperatures more active than the same enzymes found in organisms that grow at higher temperatures? Herschlag and colleagues have obtained the catalytic constants for a large number of enzymes from public databases and compared the relative magnitude of these constants with the optimal growth temperature. For a group of approximately 2200 enzymes they found no correlation between the ratio of activities for the enzymes from the cold adapted organisms relative to those from organisms that optimally grow at the higher temperatures (Figure 2C). The distribution exhibits an approximately equal number of enzymes that were more active from the higher temperature organisms and those that were less active from the higher temperature organisms, relative to the lower temperature organisms. Further support for their conclusion was obtained from the measurement of the catalytic constants from a selection of ketosteroid isomerases (KSI) from organisms that optimally grow between 15 and 46{degree sign}C. No correlation was apparent (Figure 3D). Overall, this is a nice contribution that directly addresses whether or not there appears to be any support for the notion that enzymes from cold adapted organisms are more catalytically active than those enzymes from organisms that grow at higher temperatures.

Reviewer #2 (Public Review):

The authors are trying to understand how enzymes evolve to best enable organisms to adjust to changes in the temperature of their environment. The paper reports an analysis of 2223 values of kcat from the BRENDA database, for 815 organisms with known optimal growth temperatures, and for which there are at least two variants per reaction. This analysis fails to show the expected preference for values of [(kcat)cold/(kcat)warm] > 1 observed in earlier studies.

This is a useful attempt to use one large databases to gain insight into how enzymes evolve to enable organisms to adapt to changes in temperature. They have done a good job in curating the BRENDA database to identify data that meets their criteria for analysis.

There are deficiencies that should be corrected.

(1) The first concerns the reported values of [(kcat)cold/kcat)warm]. Figure 1D shows "Rate comparisons of warm-adapted and cold-adapted enzyme variants made at identical temperatures." I think that it is important that these kinetic parameters be reported for catalysis at a common temperature, but it is not clear to me that is the case for the author's analysis. For example, they write beginning on line 234 that "The rate ratio kcold/kwarm per reaction was determined by dividing rate of the enzyme from the organism with the minimum TGrowth by the rate of the enzyme from organism with the maximum TGrowth." My reading of this sentence is that these rate constants kcat [not rates] were determined individually at the organisms optimal growth temperatures, and not at identical temperatures as reported in Figure 1D. This will complicate the author's interpretation of the two sets of results.

(2) The author's fail to present a clear physical model to use in analyzing these results.

For example, they write on line 35 that: "According to the rate compensation model of temperature adaptation, this challenge is met by cold-adapted enzyme variants providing more rate enhancement than the corresponding warm-adapted variants (Figure 1A)"

I cannot recall hearing the term rate compensation model, but am familiar with discussions on the differences in properties of enzymes isolated from organisms that have adapted to warm and cold environments. The term cold adapted enzymes is not appropriate, because it is the organism not the enzyme, that adapts to the change to a cold environment. This is accomplished through the natural selection of enzymes with kinetic parameters, stability, etc. that optimize the organisms chances of survival in a cold climate. The kinetic parameters for essentially all enzymes will decrease with decreasing temperature. The most highly evolved metabolic enzymes have kinetic parameters kcat/Km close to the diffusion controlled limit, because this optimizes energy production from metabolism. A decrease in temperature will cause the values of kcat and therefore kcat/Km for these enzymes to decrease, to the detriment of the organism. This may be overcome by selection of enzymes with values of kcat/Km close to that observed for the parent [unevolved] organism. The result is that larger kinetic parameters kcat, for catalysis at a common temperature, will be observed for enzymes isolated from the cold-adapted, compared to the unevolved parent organism. This simple application of Darwin's principals of natural selection is strongly supported by the data reported in Figure 1D.

(3) The paper alludes to, but does not clearly explain extensions of these ideas that are based on one model for how enzymes work. Enzymes often undergo large conformational changes during their catalytic cycle, and so must have sufficient flexibility for these changes to occur with rate constants that support catalysis. This predicts that the enhancement for catalysis observed for enzymes from cold-adapted organisms, might best be achieved through mutations that favor an increase in protein flexibility. There will also be natural selection of enzymes for thermophilic organisms that optimize the organisms chances of survival in a hot climate, where heat denaturation of the protein catalyst is minimized through the selection of stiffer protein catalysts. This analysis predicts a decrease in enzyme flexibility with increasing preferred growth temperature, that might give rise to an increase in protein stability with increasing optimal growth temperature.

(4) The authors should consider the possibility that the pressure to compensate for the cold-induced decrease in kcat for enzymes from cold-adapted organism will be strongest for highly evolved metabolic enzymes with values of kcat/Km close to the diffusion controlled limit. In cases where the enzyme starts out as less than perfect, an organism adapting to the cold might derive smaller, or even negligible advantages, from natural-selection of enzymes with enhanced kinetic parameters. For example, the organism might also minimize the effect of this change in kinetic parameter, by an adjustment or diversion of flux through the networks of metabolic pathways in which the enzyme functions. One possible explanation for the weak correlation observed between kcat and Tgrowth for ketosteroid isomerase is that the organisms studied gain little from optimization of the activity of this enzyme in cold-adapted organisms. One risk in the use of the larger BRENDA database may be the failure to account for differences in the pressure for enzymes to evolve to enable organisms adapt to cold environments.

Reviewer #3 (Public Review):

Enzyme catalysis underlies all living processes. Understanding the effects of temperature on enzymes is important in understanding how they are adapted to particular environmental conditions, and also relates to the response of organisms and even ecosystems to changes in temperature. The essential question is: what determines optimal growth rates of organisms, and the optimal temperature of other biological processes? Two potentially important factors are enzyme stability and catalytic activity.

This manuscript collates data from previous investigations and presents new results on KSI variants, aiming to look at the interesting question of what factors are important in relating enzyme activity and stability to optimum growth temperatures of organisms. It presents a useful survey of published data, particularly focusing on the enzyme ketosteroid isomerase (KSI) for which new resluts for a number of variants are presented, building on nice recent work by this group. The main finding in this manuscript is that enzyme optimum temperatures do not correlate well with enzyme activity. This has been found also previously. The manuscript provides quite an extensive analysis and is consistent with previous results and findings. There is useful information in this manuscript, and the compilation of data will be useful to the community, but some crucial aspects and recent relevant work are not covered, and the discussion is limited. The analysis does not identify any relevant determinant of optimum temperature, and the focus on a single temperature in each case may be misleading. Previous analyses have shown that optimum rates of enzymes do not correlate with optimal growth temperatures (e.g. Elias et al (2014) Trends in Biochemical Sciences 39, 299; Peterson (2004) Journal of Biological Chemistry 279, 20717; Thomas & Scopes (1998) Biochemical Journal 330, 1087; Lee et al (2007) FASEB Journal 21, 1934). This is particularly notable for psychrophilic (cold adapted) enzymes, but is also apparent from the fact that enzymes from the same organism often have quite different optimum temperatures. The data collected in the current manuscript are consistent with previous analyses and so are usefully confirming of this. The authors note that optimal growth temperatures may not correlate with activity for a number of reasons, including that the individual enzyme rate may not be under evolutionary pressure. Also, obviously, as noted by the authors, factors other than temperature are also important in enzyme evolution.

There is somewhat better correlation of enzyme stability with optimum growth temperatures, but it is not strong. Therefore, other factors must be important in determining optimum growth temperatures. The authors briefly mention some possibilities. One factor is that a given enzyme may not be a bottleneck in a metabolic pathway. It is not clear that KSI is in fact a metabolic limiter. Also, for many metabolic pathways, it may be essential to consider the kinetics of the pathway as a whole, which may not be determined by a single enzyme. Directly relevant here is the recent proposal of the 'inflection point hypothesis', which provides an explanation of these observations (Prentice et al. Biochemistry (2020) 59, 3562), which the authors do not mention, and may not be aware of. This hypothesis proposes that, rather than alignment of optimum temperatures or stabilities, rather the inflection points of enzymes in a metabolic pathways are aligned at the mean environmental temperature for the organism. This has the effect of coordinating relative enzyme rates and preventing metabolic disruption as temperature fluctuates. Also relevant here is that the response of metabolic pathways in general is not determined solely by a single enzyme. Prentice et al. show that, in general, the temperature-dependent properties of each enzyme in the pathway is important in determining the temperature dependence of the whole pathway.

It is certainly important to understand what molecular features determine the temperature dependence of enzyme activity and its relationship to stability. Some previous proposals are mentioned in the manuscript. One important factor at the molecular level, mentioned by the authors, is work of Åqvist, Brandsval and coworkers, who have convincingly shown that activation entropy and enthalpy differ significantly between psychrophilic enzymes and their mesophilic and thermophilic counterparts. For small soluble enzymes, this is particularly due to changes at the enzyme surface, which may also affect stability. As mentioned by the authors, there have been many proposals over the years that suggest a relationship between stability and activity, though there is not a simple general relationship. Also directly relevant for the discussion here is what factors limit enzyme activity as temperature increases. The traditional view is that loss of activity is due to protein unfolding at high temperatures (the poor correlation of stability with growth temperatures found here indicates that this cannot be a general explanation). There is increasing evidence that this simple picture is wrong (see e.g. Daniel & Danson. (2010) Trends in Biochemical Sciences 35, 584). This behavior may be accounted for by conformational (e.g. two state) effects as proposed by Danson et al, distinct from the 'flexibility' proposals mentioned in the supporting information here. The introduction of the manuscript here states that "reaction rates are reduced at lower temperatures" , which might naively seem obvious but actually is not universally true, many reactions do not display simple Arrhenius-type behavior (see e.g. Kohen and Truhlar PNAS 2001 98 848). Many enzymes show a temperature of optimum activity, i.e. activity drops above the optimum temperature but before unfolding occurs. As the authors note, Arcus et al. show that this can be accounted for by an activation heat capacity, significantly larger in psychrophiles. Signatures of this behavior are apparent at the large scale (e.g. Schipper et al Global Change Biol. 2014 20 3578; Alster et al (2016) Front. Microbiol. 7:1821) and it appears to be generally important.

Reviewer #1 (Public Review):

This study produces conservative estimates of the rates of SARS-CoV-2 importation into Canada through February 2021. The study also estimates the relative rates of intra-provincial, inter-provincial, and international transmission by province. Because these rates are investigated over time periods with varying types of non-pharmaceutical interventions, the results provide foundational information on the impact of NPIs and rates of spread to and within Canada. These rates provide useful benchmarks for other regions and deepen our understanding of the natural history of SARS-CoV-2.

Aside from a few places where speculation is unexpectedly mixed with careful data interpretation, the main limitation of the paper appears to be the unclear impact of sampling biases on the results. These biases occur inside and outside Canada. As the authors note, sequences are missing entirely from many countries and time periods where there was surely transmission. The analysis takes steps to mitigate this problem, but it is not clear how much distortion might remain. It is also unclear whether preferential testing or sequencing of specimens from recent travelers occurred and how strong this preference was (relative to sampling "random" community cases) in different places and times. These limitations are shared by many other phylogeographical analyses, but they raise the question of how literally the quantitative estimates and confidence intervals should be interpreted. My intuition is that some are much more robust than others, but this is left as an exercise.

Reviewer #2 (Public Review):

In this article entitled "early introductions of SARS-CoV-2 sublineages into Canada drove the 2020 epidemic", McLaughlin et al analyze genetic patterns in a large set of publicly-available SARS-CoV-2 sequences to characterize COVID-19 introductions and spread throughout Canada early in the pandemic. The authors conclude a majority of viral introductions into Canada can be traced to the United States via Quebec and Ontario. In addition, they report a reduction in viral importation into Canada following implementation of travel restrictions and other public health measures to reduce spread. The authors speculate that more rapid implementation of border controls and quarantine might have significantly reduced COVID-19 disease burden in Canada, at least early in the pandemic.

Although many similar genomic epidemiology studies using SARS-CoV-2 data have been published, this is the first major study focused on Canada at a national scale. The authors download a large dataset from GISAID and use appropriate tools and methods to clean and subsample this dataset. They appropriately acknowledge the limitations of their dataset as a small subset of the total Canadian case counts. Although the work is largely retrospective, the authors argue and I agree that this work can be valuable in evaluating the effectiveness of public health interventions to reduce viral importation and spread and therefore can be informative of ongoing public health measures and useful in comparing viral dynamics to the present time (future work).

This article can be strengthened in a few key areas. Primarily, the authors do not assess the robustness of their results against alternative subsampling schemes. They subsample their global sequences proportionally to case counts, but retain all Canadian sequences. As a result, their dataset is skewed heavily to sequences collected during the winter and spring of 2020, which is not representative of case counts or of case distribution. Additionally, the study focuses primarily on international importations with very limited analyses and perspective on the role of person-to-person spread within Canada.

Overall, this study deploys a set of tools used by many others in a new and important geographic region of Northern America. They make important, although, retrospective conclusions about the drivers of the COVID-19 pandemic in 2020 in Canada and conclude a reduction of international travel and quarantine requirements were important measures to reduce spread.

Reviewer #3 (Public Review):

The authors present a comprehensive description of the early importation and transmission dynamics of SARS-CoV-2 during the early stages of the COVID-19 epidemic in Canada. They implement phylodynamic analyses on a rich genomic data set generated within the country, contrasted to a vast collection of publicly available SARS-CoV-2 sequences from across the globe. Due to the vast quantities of genomic data available for this virus, they apply a downsampling scheme to generate a computationally manageable set of sequences on which analyses are run: this set includes all of the (high-quality) available sequences generated within Canada and a selection of sequences from other countries, which is proportional to the monthly reported COVID-19 cases in each of those countries. Following this step, the authors use a series of phylogenetic and phylogeographic methods to explore the number of importations of the virus to the country, the sources of these importations and the recipient provinces in Canada. They also characterise the sublineages that result from these importations (i.e., importations that result in onwards transmission), particularly regarding their size, duration and circulation between provinces.

The authors make good use of an abundant collection of SARS-CoV-2 genome sequences collected across all of Canada, providing one of the most in-depth panoramas of the spatiotemporal spread of the virus in the country during 2020. While not all Canadian provinces are represented within the data set, it is evident that the ones that contain the largest urban areas and represent the main international travel hubs within the country are included. The characterisation of the sublineages that emerge from the inferred importation events are very comprehensive and highlight how the largest importation peak of 2020 was preceded by the implementation of non-pharmaceutical interventions, while also showing that overall introductions continued at considerably lower levels during the months where these interventions remained stringent. They also show how most of these earlier sublineages became 'inactive' (i.e., extinct or no longer represented in the country's genomic surveillance) while a small proportion of the earlier introductions did remain active for longer timespans. The exploration of the main hubs where importations were detected (Quebec and Ontario) and the role that these provinces had in seeding transmission lineages across other Canadian provinces provides an interesting picture of the domestic transmission dynamics for SARS-CoV-2.

The attempt by the authors to identify the international sources of importation faces some challenges which arise from the vastly heterogeneous sequencing efforts by different countries across time. Phylogeographic methods have been long known to be sensitive to sampling bias; this is particularly the case for the COVID-19 pandemic where key territories presented well-documented underreporting of both new cases and viral genome sequences, likely introducing gaps in the available genomic data. The authors choose an interesting approach to address this bias, informing their downsampling by the monthly COVID-19 cases reported by the Johns Hopkins University Center for Systems Science and Engineering (through the 'coronavirus' R package). It is likely that this approach manages to account for some of the sampling bias between countries, but the lack of validation tests for the method and the lack of external confirmation of these results through complementary data sources warrants some careful interpretation of these findings and the uncertainty associated to them. Beyond the available sequence data, case reporting (e.g., data collected by the JHU-CSSE) has also been found to be heterogeneous across countries, particularly where diagnostic scale-up did not keep up with the local epidemic trends. These biases are less likely to affect some of the main identified sources of importation like the USA, but the possible effects for other locations will probably vary.

While individual reports of the early epidemics in specific provinces have been published, this is the first nation-wide analysis of the early COVID-19 epidemic in Canada. Given the geographical location and size of the country, these findings are key in understanding the early phases of the COVID-19 pandemic; they also add to the growing body of evidence describing the effects of multiple seeding events on the persistence of an epidemic caused by a respiratory pathogen, the speed at which such a pathogen can spread across large distances and the changes in transmission dynamics that accompany behavioural changes in human populations (in this case, derived from public health interventions). It is also important to highlight that the downsampling approach used by the authors to generate a computationally manageable data set could potentially be useful and applied to other contexts, following deeper exploration and validation.

Reviewer #1 (Public Review):

In this manuscript, Yang et al. trained monkeys to play the classic video game Pac-Man and fit their behavior with a hierarchical decision making model. Adapting a complex behavior paradigm, like Pac-Man, in the testing of NHP is novel. The task was well-designed to help the monkeys understand the task elements step-by-step, which was confirmed by the monkeys' behavior. The authors reported that the monkeys adopted different strategies in different situations, and their decisions can be described by the model. The model predicted their behavior with over 90% accuracy for both monkeys. Hence, the conclusions are mostly supported by the data. As the authors claimed, the model can help quantify the complex behavior paradigm, providing a new approach to understanding advanced cognition in non-human primates. However, several aspects deserve clarification or modification.

1. The results showed that the monkeys adopted different strategies in different situations, which is also well described by the model. However, the authors haven't tested whether the strategy was optimal in a given situation. According to the results, the monkeys didn't always perform the task in an optimal way, as well. Most of the time, the monkeys didn't actively adopt strategies in a long-term view. They were "passively" foraging in the task: chasing benefit and avoiding harm when they were approached. This "benefit-tending, harm-avoiding" instinct belongs to most of the creatures in the world, even in single-cell organisms. When a Paramecium is placed in a complex environment with multiple attractants and repellents, it may also behave dynamically by adopting a linear combination of basic tending/avoiding strategies, although in a simpler way. In other words, the monkeys were responding to the change of environment but not actively optimizing their strategy to achieve larger benefits with fewer efforts. The only exception is the suicides. Monkeys were proactively taking short-term harms to achieve large benefits in the future.

One possible reason is that the monkeys didn't have enough pressure to optimize their choices since they will eventually get all the rewards no matter how many attempts they make. The only variable is the ghosts. Most of the time, the monkeys didn't really choose between different targets/ strategies. They were making choices between the chasing order of the options, but not the options themselves. It is similar to asking a monkey to choose either to eat a piece of grape or cucumber first, but not to choose one and give up the other one. A possible way to avoid this is to stop the game once the ghost catches the Pac-Man or limit each game's time.

2. It is well known that the value of an element is discounted by time and distance. However, in the model, the authors didn't consider it. A relevant problem will be the utility of the bonus elements, including the fruits and scared ghosts. Their utilities were affected not only by their value defined by the authors but also by effects, including their novelty and sense of achievement when they were captured, as the ghosts attracted relatively much more attention than the other elements (considering the number is 2 for them, see in figure 3E).

3. The strategies are not independent. They are somehow correlated to each other. It may result in, in some conditions, false alarming of more strategies than the real, as shown in figure 2A. It is hard to believe that a monkey can maintain several strategies simultaneously since it is out of our working memory/attention capacity.

Reviewer #2 (Public Review):

In this intriguing paper, Yang et al. examine the behaviors of two rhesus monkeys playing a modified version of the well-known Pac-Man video game. The game poses an interesting challenge, since it requires flexible, context-dependent decisions in an environment with adversaries that change in real time. Using a modeling framework in which simple "basic" strategies are ensembled in a time-dependent fashion, the authors show that the animals' choices follow some sensible rules, including some counterintuitive strategies (running into ghosts for a teleport when most remaining pellets are far away).

I like the motivation and findings of this study, which are likely to be interesting to many researchers in decision neuroscience and animal behavior. Many of the conclusions seem reasonable, and the results are detailed clearly. The key weakness of the paper is that it is primarily descriptive: it's hard to tell what new generalizable knowledge we take away from this model or these particular findings. In some ways, the paper reads as a promissory note for future studies (neural or behavioral or both) that might make use of this paradigm.

I have two broad concerns, one mostly technical, one conceptual:

First, the modeling framework, while adequate, is a bit ad hoc and seems to rely on many decisions that are specific to exactly this task. While I like the idea of modeling monkeys' choices using ensembling, the particular approach taken to segment time and the two-pass strategy for smoothing ensemble weights is only one of many possible approaches, and these decisions aren't particularly well-motivated. They appear to be reasonable and successful, but there is not much in the paper to connect them with better-known approaches in reinforcement learning (or, perhaps surprisingly, hierarchical reinforcement learning) that could link this work to other modeling approaches. In some ways, however, this is a question of taste, and nothing here is unreasonable.

Second, there is an elision here of the distinction between how one models monkeys' behavior and what monkeys can be said to be "doing." That is, a model may be successful at making predictions while not being in any way a good description of the underlying cognitive or neuroscientific operations. More concretely: when we claim that a particular model of behavior is what agents "actually do," what we are usually saying is that (a) novel predictions from this model are born out by the data in ways that predictions from competing models are not (b) this model gives a better quantitative account of existing data than competitors. Since the present study is not designed as a test of the ensembling model (a), then it needs to demonstrate better quantitative predictions (b).

But the baselines used in this study are both limited and weak. A model crafted by the authors to use only a single, fixed ensemble strategy correctly predicts 80% of choices, while the model with time-varying ensembling predicts roughly 90%. This is a clear improvement and some evidence that *if* the animals are ensembling strategies, they are changing the ensemble weights in time. But there is little here in the way of non-ensemble competitors. What about a standard Q-learning model with an inferred reward function (that is, trained to replicate monkeys' data, not optimal performance). The perceptron baseline as detailed seems very poor as a control given how shallow it is. That is, I'm not convinced that the authors have successfully ruled out "flat" models as explanations of this behavior, only found that an ensembled model offers a reasonable explanation.

Reviewer #3 (Public Review):

Yang and colleagues present a tour de force paper demonstrating non-human primates playing a full on pac-man video game. The authors reason that using a highly complex, yet semi controlled video game allows for the analysis of heuristic strategies in an animal model. The authors perform a set of well motivated computational modeling approaches to demonstrate the utility of the experimental model.

First, I would like to congratulate the authors on training non-human primates to perform such a complex and demanding task and demonstrating that NHP perform this task well. From previous papers we know that even complex AI systems have difficulty with this task and extrapolating from my own failings in playing pac-man it is a difficult game to play.

Overall the analysis approach used in the paper is extremely well reasoned and executed but what I am missing (and I must add is not needed for the paper to be impactful on its own) is a more exhaustive model search. The deduction the authors follow is logically sound but builds very much on assumptions of the basic strategy stratification performed first. This means that part of the hierarchical aspect of the behavioral strategies used can be attributed to the heuristic stratification nature of the approach. I am not trying to imply that I do not think that the behavior is hierarchically organized but I am implying that there is a missed opportunity to characterize that hierchical'ness (maybe in a graph theoretical way, think Dasgupta scores) further.

All in all this paper is wonderful. Congratulations to the authors.

Reviewer #1 (Public Review):

Welzel and Schuster propose an interesting hypothesis that connexins replaced innexins in chordate gap junctions due to an evolutionary bottle neck. The majority of the animal phyla possess multiple innexins, and some of them form gap junctions while others do not. Chordates have only three innexins which are unable to form gap junctions due to glycosylation of extracellular loops; chordate gap junctions are built of connexins. The authors analysed innexin sequences from multiple chordate and non-chordate phyla and discovered that non-chordates possess both types of innexins: with and without N-glycosylation sites in their extracellular loops. Because in chordates there are only glycosylated innexins, the authors conclude that non-glycosylated innexins were lost in the last common ancestor of chordates (bottleneck effect). While in Lancelets there are no connexins, they are present in Tunicates evidencing that they evolved in the last common ancestor of Tunicates and Vertebrates.

Strengths<br> The strength of this work is that the authors analysed over 2000 innexins from multiple animal groups: seven non-chordate phyla and nine groups of chordates; both non-bilaterian and bilaterian phyla were included. This comprehensive dataset allowed the authors to propose a general scenario of gap junction evolution. In my opinion, the bioinformatics analysis reported in this work convincingly support the bottleneck mechanism in the innexin evolution.

Weaknesses<br> The weaknesses of the work are rather minor and do not affect the main conclusions:<br> (1) There is no experimental proof that the glycosylation occurs on the same motives in all the animal phyla. However, such experiments would be quite challenging therefore bioinformatics prediction is an acceptable approach.<br> (2) There is no experimental proof that the glycosylated innexins of Lancelets and Tunicates don't form a gap junction. In some animals glycosylated innexins still do form gap junctions therefore it is possible that it is the case in Lancelets and Tunicates too, especially considering that the glycosylation sites are not conserved between Vertebrates, Tunicates and Lancelets. In this case it would mean that after the loss of multiple innexins in the last common ancestor of chordates, innexins lost the ability to form gap junctions only in vertebrates.<br> (3) It is not clear if the authors aimed to use only genomic data or both genomic and transcriptomic. While it is stated that "The taxonomic groups that we have analyzed in this study were constrained by the availability of publicly available genomic data", the majority of datasets available on the Neurobase (used in this study) are ctenophore transcriptomes; the only ctenophore genome dataset is from Pleurobrachia bachei. Currently there are two other ctenophore genomes available: from Mnemiopsis leidyi (Ryan JF et al, 2013) and Hormiphora calfornensis (Schultz D et al, 2021). Additionally, genomic data are available for more cnidarian species too (e.g. sea anemones Nematostella vectensis, Exaiptasia pallida, Actinia tenebrosa and multiple corals (Shinzato C et al, 2020)).<br> (4) Line 210: «innexins were recruited as gap junction proteins in the common cnidarian/bilaterian ancestor» - gap junctions have been reported in ctenophores as well (Satterlie RA & Case JF, 1978) therefore it probably happened much earlier (in the last common ancestor of animals).

Reviewer #2 (Public Review):

1) Understanding exactly the situation in chordates and non-chordate deuterostomes is key to accurately reconstructing the evolutionary steps at the base of chordates. The authors should increase their sampling in these important groups and include hemichordates and other xenambulacrarians.<br> In Fig. 2. the alignments could include the non-vertebrate chordates (tunicate, lancelet) and lampreys to show whether the NGS sites are conserved in these taxa.<br> Tunicates have both innexins and connexins, does the NGS in innexin align to that of vertebrates?<br> Please also show the situation with hemichordates in Fig 3.

2) The authors should discuss the genomic patterns also in light of the ultrastructural evidence from the literature. For example, their data suggest that cephalochordates lack gap junctions.

"The most important finding is that the sequence of the only innexin of lancelets, which do not yet express connexins<br> (Mikalsen et al., 2021; Slivko-Koltchik et al., 2019) (Figure 3D), contains a NGS in its extracellular loop 1. This suggests that the most basal chordates not only had a limited number of innexins but might also not be able to form functional gap junctions"

Does this mean that lancelets have no gap junctions? The authors in particular should check and discuss these studies:

Tissue and Cell Volume 19, Issue 3, 1987, Pages 399-411<br> Cell junctions in amphioxus (Cephalochordata): A thin section and freeze-fracture study<br> https://doi.org/10.1016/0040-8166(87)90035-8

This study finds no gap junctions in amphioxus epidermis, alimentary tract and notochord.

Primary Sensory Cells in the Skin of Amphioxus (Branchiostoma lanceolatum (P))<br> Erik Baatrup, 1981 https://doi.org/10.1111/j.1463-6395.1981.tb00624.x

In particular:<br> "This agrees with the description (Baskin 1975) of the epidermal junctional complex of Branchiostoma californiense, but in addition this author found a membrane apposition resembling a gap junction. This was not observed in the present investigation of Branchiostoma lanceolatum.""

but some authors described gap junction like structures<br> https://europepmc.org/article/med/2628486

Gap junctions are common in tunicates, this should also be mentioned:

Georges, 1979 D. Georges<br> Gap and tight junctions in tunicates. Study in conventional and freeze-fracture techniques<br> Tissue & Cell, 11 (1979), pp. 781-792

In echinoderms, there are gap junctions but no connexins<br> BMC Evol Biol. 2019 Feb 26;19(Suppl 1):46. doi: 10.1186/s12862-019-1369-4.<br> Are there gap junctions without connexins or pannexins?

Georgy A Slivko-Koltchik 1 , Victor P Kuznetsov 1 , Yuri V Panchin 2 3<br> PMID: 30813901 PMCID: PMC6391747 DOI: 10.1186/s12862-019-1369-4

Reviewer #3 (Public Review):

The gap junction-forming proteins, vertebrate connexins and invertebrate innexins, are two distinct protein families with very similar structures and functions. In the process of evolution, innexins first arose in invertebrates, followed by connexins in vertebrates.

The authors focused on the extracellular glycosylation site in innexins,　that inhibit channel coupling between two cells, and analyzed available innexin sequences using the genomic database and reported sequences.<br> The results showed, as phylogenetic evolution progresses, innexins lose their diversity and converge only on innexins that undergo glycosylation. And connexins without glycosylation sites arose as new gap junction-forming proteins. The authors proposed a new evolutionary scenario in which the switching of gap junction protein from invertebrate innexins to vertebrate connexins is due to the loss of diversity (especially glycosylation) of innexins.

Strengths:<br> This study, which focuses on the molecular evolution involved in the biologically important mechanism of gap junctions, is significant, and will influence many future studies. Overall, the data were properly analyzed, and the visible diagrams have been created based on a vast amount of analysis.　

Weaknesses:<br> 1) The authors discussed the decrease or appearance of specific genes based on the results obtained from comprehensive sequence analysis. However, in order to discuss the number of specific genes in each animal species, especially to prove that a particular gene does not exist, the quantity and quality of the genome database greatly affect the results. It is unlikely that no gap junction proteins present at all in Echinoderms. For animal phyla for which accurate sequence data are scarce, an additional search that includes TSA will yield better results.

2) The authors proposed a scenario in which connexins emerged due to the loss of gap junction forming ability of innexins during evolution. However, this study focused only on the presence or absence of glycosylation modifications and did not consider the number of proteins in the innexin and connexin families per each animal species. Normally, gap junction-forming proteins have multiple family proteins in each animal species, and these proteins are combined to regulate channel function. The authors' scenario does not explain the small number of variety of innexin and connexin family proteins found in each phylum of Echinoderms and lancelets, and this needs to be discussed.

Reviewer #1 (Public Review):

Radke et al. used scRNA-seq to characterize Th2 cells responding to infection with Nippostrongylus bradiliensis. By comparing IL-4+ T cells from lung and mesenteric lymph they describe tissue specific signatures as well subsets of cells that share migration and other transcriptional features, suggesting a continuum of differentiation between the two organs. Further analysis of the TCR repertoire highlights significant overlap between lung and distal lymphoid T cells, although highly expanded clones tend to be more abundant in one organ. The manuscript is well written and informative with a circumspect interpretation of the results and their relationship to the current literature. The dataset in the manuscript will be particularly useful for other groups investigating Th2 responses across a variety of infection and disease models.

Reviewer #2 (Public Review):

In this study, Radtke et al. use a model of helminth infection in IL-4-IRES-eGFP (4get) mice, in which transcription at the Il4 locus is reported by eGFP, in order to define the transcriptional signatures and clonal relatedness between Il4-licensed, CD4+ T cells in the mesenteric lymph nodes (mLN) and lungs. By infecting 4get mice with the hookworm Nippostrongylus brasiliensis, which is well described to induce a robust type 2 immune response, the authors isolated and sorted eGFP+CD4+ T cells from the mLN and lungs at 10-day post infection and performed single cell RNA-seq analysis using the 10X Chromium platform. Transcriptional profiling of activated CD4+ T cells with scRNA-seq has been performed in a murine model of allergic asthma, including the lung and lung-draining lymph nodes, but this study involved unbiased capture of all activated CD4+ T cells (Tibbitt et al., Immunity, 2019). Radtke et al. have used a distinct model with Nippostrongylus brasiliensis and have focused on sorting Il4-licensed, CD4+ T cells, allowing for a greater number of captured CD4+ T cells with a "type 2" lymphocyte program for single cell analysis. Furthermore, this study sought to identify distinct and overlapping transcriptional signatures and clonal relatedness between Il4-licensed, CD4+ T cells in two "distant" tissues. In support of such an approach, there is growing evidence for tissue-specific and model-specific features of CD4+ T cell differentiation (Poholek, Immunohorizons, 2021; Hiltensperger et al., Nature Immunol, 2021; Kiner et al., Nature Immunol, 2021).

Upon dimension reduction, the authors found mLN- and lung-specific clusters, including two juxtaposed clusters that form a "bridge" between the mLN and lung compartments, suggesting immigrating and/or emigrating cells. Consistent with previous studies, the dominant lung cluster (L2) exhibited unique expression of Il5 and Il13, enhanced IL-33 and IL-2 signaling, and exhibited an effector/resident memory profile. The authors did find a small cluster in the mLN (ML4) with an effector/resident memory signature that also expressed CCR9, suggesting the potential for homing to the gut mucosa. Whether this population is specific to the mLN or would also be found in the lung-draining lymph nodes remains unclear. In the mLN, the authors also describe an iNKT cell cluster with CCR9 expression and a CD4+ T cell cluster with a myeloid gene signature, but the significance of these populations remains unclear.

The authors then use RNA velocity analysis to infer the developmental trajectory of Il4-licensed, CD4+ T cells from the two tissue sites. Consistent with previous studies, the authors found that T cell proliferation was associated with fate decisions. Furthermore, among the two lung CD4+ T cell clusters, L1 represents highly differentiated, effector Th2 cells while L2, which is juxtaposed to the mLN clusters, represents a population likely entering the lung with the potential to differentiate into L1 cells.

Next, the authors perform TCR repertoire analysis. The authors identified a broad TCR repertoire with the majority of distinct TCRs being found in only one cell. Among the TCRs found in more than one cell, a substantial number of clones can be found in both tissue sites, which is consistent with the findings that individual CD4+ T cells clones can produce different types of effector cells (Tubo et al., Cell, 2013). The authors find significant overlap of clones between the mLN and lung. In addition, they also identify clones enriched in a particular site and suggest that this represents local expansion. However, an alternative possibility is that certain CD4+ T cell clones are expanded at a particular site because the specific TCR preferentially instructs a particular cell fate. For example, fate-mapping of individual naïve CD8+ T cells suggests that certain T cell clones exhibit a greatly heightened capacity to form tissue-resident memory T cells over other cell fates (Kok et al., J Exp Med, 2020). Lastly, the authors analyze CDR3 sequences, finding the most abundant CDR3 motif belonging to the invariant TCRa chain of iNKTs. Among conventional CD4+ T cells, the abundant CDR3 motifs were not restricted to an exact TCRa/TCRb combination beyond a slight preferential usage of the Trbv1 gene. While TCR repertoire analysis allows for defining clonal relatedness among Il4-licensed, CD4+ T cells, the importance and relevance of the above findings to the in vivo type 2 immune response remain unclear.

There are several limitations of the study:<br> (1) The authors use the term "Th2 cells" to describe all Il4-licensed, CD4+ T cells. While CD4+ T helper cell nomenclature has evolved, Th2 cells and Tfh2 cells are generally used to describe distinct subsets driven by unique transcriptional programs (Ruterbusch et al., Annu Rev Immunol, 2020). While previous data suggested that Tfh2 cells are precursors to effector Th2 cells, subsequent studies support a model in which Tfh2 and Th2 cells represent distinct developmental pathways and should be designated as distinct subsets (Ballesteros-Tato et al., Immunity, 2016; Tibbitt et al., Immunity, 2019). Consequently, the authors' broad use of "Th2 cells" and a description of "Th2 cell heterogeneity" includes CD4+ T cell subsets with distinct developmental pathways that includes canonical Th2 cells as well as Tfh2 and iNKT cells. The clarity of the manuscript would be improved by describing eGFP+CD4+ cells as Il4-licensed, CD4+ T cells rather than Th2 cells.

(2) The authors used perfused lungs to isolate Il4-licensed, CD4+ T cells for scRNA-seq of "Th2 cells" in the lung tissue. However, previous studies indicate that leukocytes, including CD4+ T cells, in lung vasculature are not completely removed by perfusion, which confounds the interpretation of a tissue cell profile due to contaminating circulating cells (Galkina, E et al., J Clin Invest, 2005; Anderson, KG et al., Nat Protoc, 2014). This is particularly true in the lung and relevant as the authors found a lung cluster (L2) with a circulating signature and suggested that L2 may represent a recent immigrant "Th2 cells". Thus, it is unclear whether L2 cluster identifies immigrant Th2 cells or simply reflect the circulating Th2 cells trapped in the lung vasculature. The study would benefit of using the intravascular staining to discriminate cells within the lungs from those in the circulation (Anderson, KG et al., Nat Protoc, 2014) for the proper isolation of Il4-licensed lung CD4+ T cells to truly define immigrant "Th2 cells" within the lung parenchyma.

(3) The authors describe T cell exchange/trafficking across organs. However, in general, inter-organ trafficking refers to lymphocyte trafficking between distinct non-lymphoid tissues, rather than trafficking between lymph nodes and peripheral tissues (Huang et al., Science, 2018). Rather than inter-organ trafficking, the authors have described shared and distinct features of Il4-licensed, CD4+ T cells from a draining lymph node of one organ (gut) and a distant non-lymphoid organ (lung). The experimental approach used makes interpretation of some of the findings challenging. Specifically, canonical effector Th2 cell differentiation is well described to occur via two checkpoints, including the draining lymph node and the peripheral (non-lymphoid) tissue (Liang et al., Nature Immunol, 2011; Van Dyken et al., Nature Immunol, 2016; Tibbitt et al., Immunity, 2019). In the draining lymph node, Th2 cells acquire the capacity to express IL-4 alone, but do not complete effector Th2 cell differentiation until trafficking to the inflamed peripheral tissues and receiving additional inflammatory signals. Consequently, it is unclear whether the differences identified in the mesenteric lymph node and lungs simply reflect well-described differences between the two Th2 cell checkpoints or organ-specific differences (gut vs lung). Il4-licensed, CD4+ T cells from the intestinal mucosa and lung-draining lymph node would also be needed to truly define organ-specific differences during helminth infection.

(4) The study includes a single time point (day 10) whereas Tibbitt et al. performed scRNA-seq in the lung and lung-draining lymph node at multiple time points during type 2 immunity (Tibbitt et al., Immunity, 2019). As a result, it remains unclear how similarities or differences between the mesenteric lymph node and lung response would change over the duration of helminth infection, especially given the helminth life cycle involves multiple infection stages.

(5) The study analyzed one scRNA-seq experiment that included two mice without validation via flow cytometry or other method to infer a role of a particular finding to the type 2 immune response in vivo.

Reviewer #3 (Public Review):

The authors aimed to study in detail the process of Th2 differentiation during helminth infection, using droplet-based, single-cell RNA-seq and TCR sequencing combined with an established IL-4-reporter mouse.

A strength of this study is its focus on helminth infection. High-dimensional studies of Th2 cells have been published previously, but these have mostly examined allergic disease: https://www.jci.org/articles/view/125917 https://pubmed.ncbi.nlm.nih.gov/31231035/ https://pubmed.ncbi.nlm.nih.gov/31745340/. There are no such helminth papers or pre-prints that I am aware of; hence scRNA-seq examination of Th2 biology in worm infections offers a unique angle for better defining Th2 heterogeneity and differentiation in vivo.

The authors use IL-4-eGFP reporter mouse to study Th2 responses. This is a certainly an established, conventional approach, carrying with it the assumption that IL-4 mRNA expression defines Th2-like cells. However, this approach also restricts the study, reducing the chances of detecting Th2 cells that perhaps don't strictly adhere to this definition.

The authors examine mesenteric lymph nodes (MLN) and lung tissue, which given that the helminth infection proceeds via the GI tract and lung is appropriate. However, direct comparison between a secondary lymphoid organ (SLO) and non-lymphoid tissue is difficult and may be confounded by the different cell isolation methods needed. It was notable that assessment of lung draining lymph nodes or the gut epithelium/lamina propria was not included.

One weakness of this study was its apparent reliance on data from only two mice, and the lack of biological "validation" studies using, for example, flow cytometric detection of novel proteins or cellular states, or examination of Th2 biology in gene knockout mice.

Reviewer #1 (Public Review):

The subject of this review can be interesting and in principal helpful to the researchers who works on germline mutations. The authors have summarised all the work done in this area over the past 10 years. However, I found parts of the review unsurprising and also I am not sure if the reviewer have convinced the readers what is the best practice for calling gremlin de novo mutations.

Bergeron et al present an interesting paper about the biases and implications of the different methods used for the identification of de novo mutations from pedigree dataset. The first part is a review of the different methods and criteria used for the mutation identification (coverage, mapping quality, measure of callable genome among other) in previous studies. The second part of the study is an original approach named "mutationathon" where 5 teams received the same dataset from a pedigree to estimate the mutation rate. The objective here, that is very interesting for the field, is to understand how the different approaches from several teams impacts the mutation rate estimation. The 5 estimates are on the same order, but with one value significantly higher than the others. Moreover, the candidate mutations identified by each team is very variable: about 20-30 mutations candidates but with only 7 true positives in common despite strong criteria during identification steps. This study is very interesting and shows the importance of standard mutation identification method in mutation identification, and the difficulties or biases in the comparison of mutation rates from different publications using different approaches.

Reviewer #2 (Public Review):

I commend the authors for their extensive work on summarising large number of germline de novo mutation (DNMs) studies of Human and non-Human trios. They outlined all the methods used by different studies in order to call DNMs. They pointed out different stages which may affect DNMs calling, including; samples size, sample size, library preparation, alignment, variant calling, and post-filter. Finally, by analysing DNMs in macaques pedigree across five groups, they have demonstrated how different strategies in variant calling, may lead to reporting different mutation rates.

The authors are correct that identification of true DNMs are affected by experimental and analytical strategies. This is a long-time known issue in the field. However, as the authors also mentioned, despite all the variations yet the reported DNMs across different studies are very much in agreement. Indeed, in their Mutationathon exercise on calling DNMs in pedigree of three generations of rhesus macaques they have demonstrated that although all the five groups reported variation in number of DNMs yet the difference in mutation rate is insignificant. Moreover, I am not convinced variability in terms of calling DNMs is a major issue in this field at least not in recent years and specially for Human germline mutation. More recent studies with large number of trios such as analysis of ~12k trios by Kaplanis et al., bioRxiv, 2021 eliminates most of the issues due to systematic noise.

Having said above it is very helpful to have some guidelines to take different factors into consideration for future experiments. However, there are few issues that I am not sure if the authors have addressed in their review:

The authors have not address issues with relatedness: the strategy for calling DNMs in multi-sibling families. This is also very important in non-human studies.

The best practice suggested here is certainly not applicable for different species. Due to differences in selective pressures the number of DNMs in different spices is different. This directly affect the detection method. Moreover, cellular processes causing germline de novo mutations may vary between species. Hence, our mutation calling strategies cannot be generalised across species.

Issues with somatic mutation contamination, as the authors correctly mentioned, can vary depending on the tissue of choice. However, the authors do not suggest a solution. For example, in case of clonal hem what is the solution to overcome this issue and call DNMs? Perhaps, the authors can explore parameters such as cell fraction or purity of the tissue which, can guide the downstream analysis for DNM calling.

Another aspect that may affect DNM calling, is clinical history of the parents and/or child. What would be the strategy for these cases?

How about introducing a site-specific error rate? Given the high number of trios publicly now available it would be extremely useful to compute site specific error rate per nucleic acid.

Overall, as the authors also mentioned, DNMs calling require study- or species-specific thresholds. Therefore, I am not convinced if their suggested best practice is really applicable to all types of trio-studies.

Reviewer #3 (Public Review):

This study is motivated by the variable germline mutation rates that are estimated from numerous genome sequencing studies of primate pedigrees. The authors argue that this variability is the result of methodological differences in both the molecular and computational strategies employed. Therefore, the authors launch the "Mutationathon" as an effort to isolate the effect of computational differences employed by different research groups that form the authorship of this manuscript. Using PCR validation, they are able to assess the specificity and sensitivity of each approach, recommend some high-level guidelines, and conclude that all future studies should provide detailed reporting of all computational details.

Whole-genome DNA sequencing of pedigrees consisting of at least the mother, father, and at least one offspring has become the gold standard for estimating the rate of germline mutation in humans and other primates. While the estimated mutation rates are broadly consistent, they can vary by up to a fact of two. The authors make a strong argument that the primary reason for the variance in estimated rates is fundamental differences in the computational methods employed from study to study. Specifically, since germline mutations are rare, therefore most studies make substantial efforts to eliminate false positive predictions. The computational "filtering" approaches differ yielding variability in the specificity and sensitivity of each study. Furthermore, studies account take different approaches to account for the specificity and sensitivity of their approach. As a result, the final estimated germline mutation rates vary.

The authors seek to assess the impact of differing computational filtering approaches on estimated germline mutation rates by launching the Mutationathon. The study design is to provide DNA sequencing data from a single pedigree of rhesus macaques to five different research labs, who each apply their internal "best practice" computational approaches to estimate a germline mutation rate. The authors then use PCR validation of the union of all mutation predictions to directly measure specificity and sensitivity with an orthogonal molecular strategy. Finally, they provide recommendations for future studies to thoroughly document the methodologies for reproducibility and comparability.

A key strength of this study is the fact that the authors were able to isolate the impact of computational differences on estimated rates by providing identical sequencing data to each group. Another key strength is the fact that PCR validation of all predicted mutations was performed (or at least attempted), providing an independent assessment of errors. However, these strengths are balanced by the fact that the study was conducted on a single macaque pedigree, thus preventing an assessment of the variance in rates estimated across pedigrees for the same computational approach. Similarly, neither multiple tissues nor a multi-generational pedigree was used, thereby preventing the assessment of the degree to which tissue-specific mutations or early post-zygotic mutations masquerade as germline mutations given their observed allele ratios. Lastly, the basic conclusion is that methods are sufficiently variable that providing a gold standard approach was not possible and the paper concludes that each study should simply thoroughly detail the methods so that differences in reported rates can be better understood.

While the motivation behind the study is clear and the detailed treatment of the variability of computational approaches is fantastic, the basic conclusions of the study largely reflect the understanding of expert researchers conducting such work. However, the efforts to document the largest computational drivers of variability germline mutation rate estimation are laudable and will likely inform future efforts in this area.

Reviewer #1 (Public Review):

Small animal models are very useful to study the pathogenesis of human diseases and for preclinical studies of potential therapeutic agents. In the present study, Gawish et al. passaged SARS-CoV-2 in mice to generate a mouse-adapted virus strain (maVie16) that replicates efficiently in mice and induces inflammation-driven pulmonary disease. They show that the mouse-adapted virus remains strictly dependent on ACE2 and demonstrate protective effects of inhibition of proinflammatory cytokines and inhaled recombinant ACE2. The authors conclude that they established a novel mouse model for COVID-19 that recapitulates many features of human disease and allows the testing of antiviral agents.

As mentioned in the manuscript, several other groups have previously reported the development of mouse models for COVID-19. Initially, transgenic mice expressing the human ACE2 receptors were generated. However, these are extremely susceptible to SARS-CoV-2 infection and show a distinct pathology. Others have also generated mouse-adapted SARS-CoV-2 strains which share some features of the present maVie16 strain. Thus, not all aspects of the present study are entirely novel. Strength of the present study are that the authors show that, similarly to COVID-19 in humans, disease severity is to some extent determined by cytokine-driven immunopathology. Using ACE2 knockout animals, the authors also show that SARS-CoV-2 replication remains strictly dependent on ACE2. While this is as expected definitive experimental evidence is nonetheless important. Additional strengths are the evidence the pDCs play are key role in the response to infection and the demonstration that maVie16 strain in duces mCOVID-19 in both BALB/c and C57BL/6 mice. Altogether, the results suggest that this novel mouse-adapted SARS-CoV-2 strain will be useful for studies on viral pathogenesis and testing of antiviral agents.

Reviewer #2 (Public Review):

A SARS-CoV-2 isolate was passaged in Balb/c mice and it is demonstrated that viral load and cytokine induction increased with passage number while body weight decreased. Further, mutations acquired during passaging are reported and evidence obtained in silico is provided that these mutations increased binding of the viral spike protein to murine ACE2. The virus is shown to cause lethal disease in Balb/c but not C57/BL6 mice, although the latter show body weight loss and transient pneumonia. The impact of infection on blood cell populations and cytokine release as well as adaptive immune responses is reported. In addition, elevated levels of IFNgamma and TNF were measured and antibodies targeting these factors are shown to protect from body weight loss. Finally, evidence is provided that ACE2 is required for viral spread in the lung and induction of disease and these processes were blocked by soluble ACE2. The findings are solid and the model will be helpful to the field.

Reviewer #1 (Public Review):

In this computational study, the authors conducted extensive coarse-grained molecular dynamics simulations to study the location and kinetics of electroporation of complex lipid membranes. It was evident that electroporation does not occur uniformly in the membrane, but preferentially at certain locations. Analysis of the results using several machine learning models found that the local lipid composition is essential to the poration probability. While this observation is somewhat expected, the machine learning approach leads to a quantitative and predictive model, which is valuable. On the other hand, poration kinetics depend on additional lipid properties, such as the local area compressibility modulus. Again, this is somewhat expected qualitatively, but the Bayesian inference approach leads to a quantitative model that contains physically meaningful parameters (Eq. 3). Therefore, as the authors pointed out, by combining extensive coarse-grained simulations and these analyses, one is able to establish quantitative models for the poration kinetics at the whole-cell level, which are valuable for guiding the design/optimization of experimental protocols for electroporation in different applications.

Reviewer #2 (Public Review):

In this work the authors illustrate some of the molecular mechanisms of membrane electroporation, using coarse-grained molecular dynamics simulations of complex, ~60 lipid type, membrane mimics of both a brain plasma membrane (BPM) and an average mammalian plasma membrane (APM). They repeatedly simulated electroporation with both hyperpolarized and depolarized applied electric fields and carefully analyzed the resulting simulations: mapping the time and location of pore formation and the local lipid composition and membrane properties. They successfully described the pore kinetics using Bayesian survival analysis, characterized the local bilayer environment at preferred pore formation sites, and trained machine learning models to accurately predict pore forming sites from the local bilayer environment. Together that illustrates the importance of both global and local bilayer properties and the role of different lipid types such as polyunsaturated lipids in electroporation.

Strengths of this work include the appropriate use of coarse-grained simulations to access larger length and time scales as well as complex plasma membrane mimics. Combined with clever and thorough analysis, that allowed them to show the strong spatial dependence of poration sites on local lipid composition. The use of coarse-grained simulations is also the main weakness of this work because, as pointed out by the authors, they have significant limitations. Having further reduced degrees of freedom compared to all-atom simulations, coarse-grained force fields average over or implicitly include shorter and faster motion(s). The Martini force field used here has successfully captured a wide range of membrane dynamics and functions, including several demonstrations of electroporation in simple systems, but its lack of directional hydrogen bonding and the mapping of four water molecules to a single coarse-grained bead limits its ability to capture the finer details of pore formation.

These predictions are still of significant interest and hopefully will motivate further work, such as: 1) validation - using finer resolution modeling as well as experiments, and 2) extensions - such as those proposed in the manuscript, as well as combining local lipid neighborhood analysis with pore initiation rate analysis which would allow for finer decomposition of relative importance of different local vs global properties.

Reviewer #3 (Public Review):

The open questions targeted by this study refer to the proneness of membranes to electroporation as a function of membrane composition. The authors use two previously setup coarse grained membrane systems with a complex composition mimicking average mammal and brain plasma membranes, and expose them to nanosecond electroporating pulses. Pore formation and kinetics is studied using coarse grained molecular dynamics simulations, machine learning and Bayesian survival analysis. The authors demonstrate that this combination of approaches is a powerful tool to investigate the individual role of a variety of components of such complex heterogeneous systems in the electroporation phenomena.

A very valuable outcome of this study is the finding that pores colocalize with specific compositional regions in the membrane. From molecular dynamics studies, the authors find that, among other factors, in regions of higher local concentration of polyunsaturated lipids the membrane is easier to porate. Gangliosides appear to also substantially affect poration. Furthermore, the work shows dependence of the poration kinetics on the stretching elasticity of the membrane. It also puts forward the potential of the combination of the employed methods to help better understand and infer the kinetics of pore formation under short (sub)nanosecond electric pulses from numerical analysis.

Reviewer #1 (Public Review):

The manuscript: "Cavefish adapt to hypoxia by constitutive overexpression of hypoxia-inducible factor genes and increased erythrocyte development" by Corine van der Weele and William Jeffery addresses an interesting topic. The authors find that cavefish of the species Astyanax mexicanus develop more red blood cells than surface fish of the same species when raised in the laboratory under normoxic conditions. The authors perform a detailed analysis of the developmental origins of hematopoiesis and find an expansion of the hematopoietic domains in cavefish embryos. Further the authors test how cavefish respond to chemically induced hemolitic anemia and how growth is affected under artificial hypoxic conditions. Finally, they look at transcriptional regulation of known hypoxia response genes.

Major comments:

While the authors are performing a careful and conclusive developmental analysis of the hematopoiesis in these fish,there are a few inconsistences. It would be nice to see comparable timepoints in both the insitu and the qPCR analysis. For example, the authors show in Figure 1G/H 36 and 84 hpf timepoints while the qPCR is performed at different stages. This is especially relevant as the authors make quantitative statements from whole mount in situ analysis which are not necessarily suited to do quantitative comparisons. Furthermore, they are not using the same genes as readout as they use hbb2 in the insitu and hbbe2 in the qPCR analysis.

While the growth study is conceptually interesting, it is unclear why the authors average all 12 larvae even though they keep them individually. They mention that this is to avoid "pseudoreplication", however I am not sure why that would be the case. It would be important to see all the data. Also, in Figure 4B the statistics for the comparison between surface normoxic and hypoxic are not shown, even though the text mentions it as significant.

Reviewer #2 (Public Review):

The authors aim to gain an understanding of how cavefish have evolved to thrive in an environment where the oxygen content of the water is low. They provide strong evidence that cave-adapted fish, compared to their river-adapted counterparts, have more blood cells and higher constitutive expression of hypoxia-inducible genes. Evidence that these traits provide cavefish with an advantage in low oxygen is weak however. The authors achieve their aim of identifying possible mechanisms that would allow cavefish to survive low oxygen, but do not go far enough in my opinion to show the traits they identified are adaptive. The work will be of great interest to researchers in the fields of evolution, physiology, and developmental biology. Details about each claim and the data in support are provided below.

The authors claim that cavefish (CF) have enlarged embryonic hematopoietic domains compared to surface fish (SF). This is supported by in situ hybridization using several markers of hematopoiesis. It is further supported by quantitative PCR that shows elevated expression levels of the same markers. The authors also report that compared to other teleost, both SF and CF have expanded hematopoietic domains. This is an interesting finding of more broad relevance as it could provide insight into how A. mexicanus SF, and not other fish, were able to invade caves initially.

The authors claim that CF develop more erythrocytes than surface fish and this is a maternally controlled trait. This is supported by quantification of blood cell number in live images from CF, SF, and reciprocal hybrid crosses. The authors also show the qualitative difference in blood cell number with o-dianisidine staining.

The authors use a method to reduce oxygen concentration in the laboratory water to levels that they claim are closer to those observed in the cave (5.0-5.4 mg/L). It should be noted that lower (4.43 mg/L) and much lower (2.97 mg/L) levels of oxygen have been recorded in the rivers and caves, respectively. The authors claim that SF respond to the reduction by increasing insulin growth factor binding protein 1a (igfbp1a) expression while cavefish do not. This is supported by the qPCR data.

The authors claim that SF respond to hypoxic environments (5.0-5.4 mg/L) by reducing growth. While some of the SF exhibited lower axial growth rate compared to the control group, the results are not significant and the number of biological replicates in low (N = 6). The claim is therefore not strongly supported by the data. Repeating this experiment with additional biological replicates would strengthen the claims. Extending the amount of time the fish are exposed to hypoxia to measure a larger change in growth could also strengthen the claims.

The authors claim that CF exposure to hypoxia in the laboratory does not affect igfbp1a expression or retard growth. These results are supported by the qPCR data and the finding that there is not a significant difference in growth between CF in "normoxic" vs "hypoxic" conditions. The results would be strengthened by increasing the number of fish as suggested for the SF samples.

The authors claim that CF show elevated expression of Hypoxia Inducible Factor (HIF) regulatory genes relative to surface fish. This claim is strongly supported by quantitative PCR.

The authors carry out one functional experiment; they expose SF and CF to increasing concentrations of phenylhydrazine (PHZ) to reduce red blood cell number and examine the impact on development and growth. They claim that "cavefish are less sensitive to hemolytic anemia". I don't agree with this interpretation of the data. Instead, cavefish are less sensitive to developing hemolytic anemia with increasing concentrations of PHZ because they have more blood cells. This experiment can be used to ask what happens to CF when the number of red blood cells they contain is reduced to levels that are similar to SF. It appears that when CF are experimentally manipulated to have the same amount of red blood cells as SF, they develop abnormally. The results would be easier for the reader to interpret if the actual number of red blood cells in the treatment conditions were reported (instead of as a percentage of treatment) and if statistical analysis were performed on treatment conditions where the number of red blood cells in SF and CF are roughly equal.

Based on the data presented the authors conclude that the developmental and expression differences they observed in CF represent adaptations to the hypoxic cave environment. The evidence that these traits provide CF with an advantage in low oxygen waters is weak; the results showing that SF growth is restricted in low oxygen is not significant and the amount of time and oxygen concentration the CF are challenged with is minimal.

The existence of a trait in CF is not strong evidence that the trait is adaptive. I think the authors would need to provide more experimental evidence to support the major claim (the title of the paper) that "Cavefish adapt to hypoxia by constitutive overexpression of hypoxia-inducible factor genes and increased erythrocyte development."

Finally, there are several independent A. mexicanus CF populations and many different species of cavefish around the world; the authors would need to discuss whether the traits they observed are present in other cavefish or other cave animals before applying their conclusions more broadly.

Reviewer #3 (Public Review):

In this manuscript van der Weele and Jeffery propose two evolutionary mechanisms, via which cavefish (Mexican tetra Astyanax mexicanus) have adapted to a hypoxic environment based on direct comparison with hypoxia-sensitive surface dwellers of the same species. These adaptations are increased erythrocyte production and heightened transcription of hypoxia-inducible factor 1 (hif1). Using a combination of time lapse imaging, bright field microscopy and in situ hybridization with a battery of hematopoietic markers, they provide clear and compelling evidence that cavefish have a larger number of erythrocytes than surface dwellers and that this increase stems from an expansion of two erythropoiesis domains during early development. To address the functional relevance of increased erythrocytes they induced hemolytic anemia using the drug phenylhydrazine (PHT) and showed that cavefish are less sensitive to this disorder than surface fish, possibly as a result of the increase in erythrocytes. Nevertheless, both types of fish exhibited developmental anomalies post-treatment, including reduced tail length. They further propose that the larger number of erythrocytes may promote adaptation to hypoxia by countering the negative impact of this environmental stressor on growth. Indeed, surface fish appear to be more susceptible to hypoxia-induced stunted growth of the post-anal tail than cavefish, although the numbers are quite variable. Lastly, the authors go on to show that cave dwellers express constitutively high transcript levels of hif1 genes and downstream targets of Hif1. One of these targets is the growth suppressor IGFBP1a, which could explain growth restriction of surface fish under hypoxia. Overall, even though increased erythrocyte production is a well-documented response to life in hypoxic environments, this study provides an interesting perspective on this adaptation seen through the lens of evolutionary biology.

The data support the main conclusions that erythropoiesis and hif1 transcription are enhanced in cavefish but do not convincingly identify the functional relevance of these traits to hypoxia adaptation in cavefish. In this regard:

1) The induction of hemolytic lysis using PHT and associated developmental defects raises concerns about lack of drug specificity.

2) The connection between erythrocyte number and organismal growth is not clearly established. The reduced tail length defect that is observed following PHT treatment, even if specific, is likely to be an indirect consequence of abnormal notochord morphology rather than arrested tail growth.

3) Because the above connection was not clearly established, the lack of growth inhibition in cavefish following exposure to hypoxia (thought to be offset by the elevated number of erythrocytes in cavefish) is not readily explainable. The data itself showing a difference in tail growth between cavefish and surface dwellers exposed to hypoxia is not strong due to high variability and low numbers.

4) Increased hif1 transcript levels may contribute to enhanced hypoxia adaptation in cavefish, however Hif1 is known to be primarily regulated at the post-translational level, resulting in its enhanced stability and activity. It is unclear whether the activity of Hif1 is elevated in cavefish relative to surface fish as several known Hif1 targets are not up-regulated in cavefish relative to surface fish.

In addition, there are a few technical concerns<br> 5) The manner in which hypoxic water was generated (bubbling of nitrogen gas) is unlikely to maintain a constant value over time. Furthermore, covering the tank with foil will not prevent gas exchange. Hence there is a concern that the hypoxic values may be variable between trials and even within the same trial.

6) It is important that transcript levels for the q-PCR reference gene be stable across normoxic and hypoxic conditions. A discussion about the considerations that went into selecting RPL13a as a reference gene was not provided.

Reviewer #1 (Public Review): 

Rodríguez-Lorenzo et al combine multidimensional CyTOF analysis with multiplexed IHC to determine the presence of immune cells in the septum and choroid plexus of control and MS patients. The authors collect a cohort of human brains samples and perform extensive characterization with a 37-antibody panel. Interestingly, they found increased CD8 T-cells, CD4 T-cells and increased NK cells. The NK cells are of special interest as these cells have not been well characterized in MS and previous reports have shown a decline in the periphery. The authors further refine the T-cell and NK cell populations in both the choroid and septum, and the role of B-cells in the blood. Finally the authors use multiplex IHC to localize specific NK subsets to the lesion and border of MS lesions. 

Overall the work is an impressive analysis of an understudied cell-type in human MS, and represents an important finding. The paper is well presented and the figures very clear. However, the manuscript is descriptive and, although this is not a problem by itself, the depth and limitations of the Cytof (only 37 markers) leaves the reader without a clear idea of what these cells could be doing. 

Some single-cell RNAseq and other ways to interrogate potential mechanisms and function would be particularly helpful here, but is perhaps beyond the scope of the paper. At minimum more immunohistochemical and smFish or in situ hybridization to validate key findings (using the markers identified by CyTOF) and add to the spatial relationships of Nk Cells with other border and brain cells would be informative. 

A major weakness of the study is that is is underpowered and thus not clear how robust or representative these findings are in MS given the heterogeneity of the disease and also potential differences in Sex, Age and lack of healthy controls. (AD samples labelled as control.) 

It is also important to show the NK cells are actually in the parenchyma and interacting with other cells (e.g., microglia) of the lesion. If the authors have this tissue and antibodies to do that, this would add to the study. Moreover, the details on samples and controls should be more clearly communicated in the text and legends as well as the caveats and limitations of the study in the Discussion.

Reviewer #2 (Public Review): 

The data are extensive, valuable, convincing, and entirely descriptive (as studies using human post-mortem material must be, of necessity). What emerges is a detailed account of NK cells in specific regions of the MS brain (although here the authors slightly overplay how little is known about NK cells in MS). The study provides a very comprehensive resource. The authors speculate on what their data might mean in terms of disease dynamics is a reasonable and informed way, but much of what is concluded is inference not backed up by experiment studies that would allow this to be more than a resource paper.

Reviewer #3 (Public Review): 

The authors introduce their work in the context of the prevailing uncertainties about the pathogenesis of multiple sclerosis (MS) and, in particular, seem to reference the initiation of immune lesions in early MS. However, the work itself addresses end-stage MS situations, which is quite possibly an entirely different landscape altogether, and may not be informative about MS initiation. 

As a textual point, the manuscript makes far too many speculations about possible cell trafficking between compartments than is justified by a cross-section study. 

That said, the work itself is a carefully done descriptive characterisation of the leucocyte landscape found in the periventricular septum, choroid plexus (and peripheral blood) post-mortem from cases of multiple sclerosis (MS), non-MS neurological disease (dementia), and non-neurological controls (8-12 each). The material is rare, the post-mortem delays are quite short, the cell lineage characterisation is fairly extensive and some of the data are well supported by immunohistochemistry.

Reviewer #1 (Public Review):

ERK1/2 are two important MAP kinases that play important roles in regulating cell growth, proliferation, and stress responses. Their activities are well known to be regulated by phosphorylation on the activation loops. Here the authors report that ERK1/2 are also regulated by cysteine palmitoylation. This is a new finding and could potentially lead to new ways to control ERK1/2 signaling as new anticancer strategies.

The strengths of the work include: (1) convincing demonstration that ERK1/2 contain cysteine palmitoylation by several complementary methods; (2) identification of the sites of palmitoylation by mutagenesis; (3) showing that cysteine palmitoylation is dynamic and varies with EGF stimulation; (4) showing that the C254A mutation (which lacks palmitoylation on Cys254) in ERK2 promotes pERK2 (Thr185/Tyr187) but inhibits serine phosphorylation of ERK2; (5) Finding that several DHHC proteins as potential writers for this modification; (6) Showing that palmitoylation of ERK is regulated by high-fat diet in vivo in different tissues. These observations are interesting and further understanding of them could provide useful insights to the field.

A major weakness of the work is that even though a lot of interesting observations are made, there is not much mechanistic understanding about these observations. The lack of understanding makes reading the manuscript a little difficult, especially that some observations are contrary to each other. For example, even though the small molecule inhibitor data suggest that APT1 and APT2 are the enzymes that remove the palmitoylation, knockdown of these two enzymes failed to produce the expected effect. In particular, how the palmitoylation on Cys254 affects the two types of phosphorylation differently and how in turn that causes changes in the downstream signaling effects, which DHHCs are the physiological writers and why so many DHHCs can work on ERK1/2 under over-expression conditions. Addressing these questions will significantly enhance the manuscript.

Reviewer #2 (Public Review):

Azizi et al. examined the regulation of signaling by ERK1/2 via palmitoylation, a post-translational lipid modification by employing a series of cell-based studies and biochemical and molecular studies of ERK1/2 palmitoylation, activation, and downstream gene expression. The authors also investigated ERK1/2 palmitoylation in a mouse model of metabolic syndrome. The authors found that ERK1/2 palmitoylation is induced in response to stimulation with EGF (that activates ERK1/2 downstream of EGFR) and attempted to uncover the kinetics and enzymatic regulation of ERK1/2 palmitoylation by writers (DHHC acyltransferases) and erasers (thioesterases/depalmitoylases). Overall, the data presented clearly establish palmitoylation as a novel regulatory mechanism governing ERK1/2 signaling activity, which is a significant finding. The precise mechanism by which palmitoylation modulates ERK1/2 activity (i.e. opposing effects on TEY vs. Ser phosphorylation; impact on ERK1/2 localization, stability, etc) and writer/eraser enzymes that definitively control this process are less clear but confounded by the presence of multiple palmtioylation sites on ERK1/2 that may have different and even opposing effects on ERK function and the presence many DHHC writers with different cell-type expression patterns and intracellular membrane localization that could manipulate ERK activity at disparate cellular locations. A major strength of the manuscript is the use of several techniques to decode the enzymatic regulation of ERK2 palmitoylation including metabolic labeling/click chemistry and acyl biotin exchange (ABE) to detect palmitoylation, immunoprecipitation and proximity labeling (TurboID) to assess association of ERK2 with palmitoylases/depalmitoylases, and assessment of the impact of these enzymes on ERK1/2 phosphorylation. Overall, this manuscript provides novel insight on regulation of ERK1/2 by palmitoylation that could be of significance for many diseases including cancer and metabolic syndrome and provides a framework for elucidation of enzymatic regulation of dynamically palmitoylated signaling proteins.

Reviewer #3 (Public Review):

Azizi and coworkers report an expansive investigation of protein palmitoylation of ERK2 and resulting impact of cell signaling pathways impacted by this kinase. Using several chemical biology approaches, the authors determine the time- and external stimulus-dependence of ERK2 acylation in a number of cell lines. Mutagenesis studies are employed to probe the interdependence of ERK2 acylation and phosphorylation, providing a functional linkage for protein palmitoylation impact on downstream signaling. Pharmacological perturbation of ERK2 palmitoylation is shown to impact gene transcription controlled by ERK2-dependent pathways. The authors attribute ERK2 acylation to a subset of DHHC PATs and deacylation to APT2 based on overexpression studies and detection of protein-protein interactions in cell-based experiments. Finally, the authors demonstrate ERK1/2 acylation is affected by the cellular and organismal metabolic state. This work illuminates a novel regulatory mechanism for ERK1/2 and supports interplay of complementary posttranslational modifications controlling kinase signaling through this protein.

Overall, the data presented in the paper strongly support the conclusions of the authors with exception of the APT-related experiments. The studies of APT inhibition and associated impact of ERK1/2 acylation and signaling are somewhat problematic. Contrary to the authors assertions in the text, the data presented in Figure 4 and Figure S4C do not compellingly support an increase in ERK1/2 palmitoylation when cells are treated with APT inhibitors. In the absence of further quantitation, the data are not sufficient to indicate APT inhibition increases ERK1/2 palmitoylation. This concern is compounded by the lack of any change in ERK1/2 acylation in presence of APT overexpression or siRNA knockdown. The protein-protein interaction of ERK2 with APT2 demonstrated by biotin labeling does not provide sufficient foundation for annotating APT2 as the eraser for ERK1/2 palmitoylation based on the totality of data presented. Given the promiscuous nature of PalmB, it's also unclear the perturbations in downstream gene transcription in Figure 4C can be directly attributed to changes in ERK1/2 palmitoylation. Additional experiments would be needed to support the proposal of APT2 regulation of ERK1/2 palmitoylation.

Reviewer #1 (Public Review):

Studies of antibiotic tolerance at the single-cell level can tell us a great deal about bacterial physiology and hint at routes for the emergence of antibiotic resistance in clinical contexts. Here, Koganezawa et al. develop and use a light-activated recombinase system to excise the gene for chloramphenicol acetyltransferase at a precise time. They show that when E. coli cells experience this deletion after exposure to chloramphenicol, they are more likely to survive and resume growth under chloramphenicol treatment. First, the authors show that their system can cause the loss of the chloramphenicol gene in some cells, which they monitor via the loss of an attached fluorescent reporter. They then show that a fraction of cells that lose the chloramphenicol genes are still able to slowly divide while experiencing continued chloramphenicol, and that this fraction decreases and disappears when the chloramphenicol gene is lost at increasing time intervals before chloramphenicol exposure. They also show that cells with a growth restoration phenotype have a partially-restored ratio of 50S/30S ribosomal subunit expression. Overall, the authors present an interesting system to explore the effects of history-dependence on bacterial antibiotic survival, and use a set of carefully-designed experiments and controls to show how gene loss can lead to diverse phenotypes within a bacterial population. I commend the authors on conducting a challenging study with many relevant controls. The ability to precisely delete genes and monitor cells in response is very interesting and the authors present interesting data about a window of time in which cells with the resistance genes deleted are able to survive. However, to fully support the author's conclusions there are several aspects of the manuscript that would benefit from additional data, alternative presentation formats, or further explanation, as detailed below.

Major points:

1) I found the data on ribosomal protein stoichiometry to be somewhat unclear and had some questions about whether the results were statistically significant. Specific points: (a) In Fig. 5 it appears that growth-restored and growth-halted cells essentially have the same behavior, and pre-deleted cells are very similar. How are the growth-halted and pre-deleted cells growing in the presence of chloramphenicol? Also, why are the distributions essentially the same for all? (b) In Fig. 4E-F the behavior of both the growth-restored and growth-halted cells fluctuates a great deal. Are the differences between the two strains actually significant? It appears that the later timepoints (e.g. 65h) may start to diverge, but the experiments stop here so it is difficult conclude whether this is representative of the future or not.

2) Growth is quantified in different ways in the manuscript, which make it difficult to compare different data and potentially masks information about cell division. In some cases, generation time is presented in minutes (e.g. Fig. 1E), in others generation time is presented in hours (e.g. Fig. 2I), and then the authors switch to elongation rates (e.g. Fig. 3B). The generation time vs. elongation rate could potentially mask behavior where cells are filamenting but not dividing. The differences in units makes it difficult to understand the growth impact on growth-restored cells. I gather from Fig. 4B that these growth-restored cells are barely growing?

3) Do the growth-restored cells, which are very slow growing in chloramphenicol, return to normal growth after chloramphenicol has been removed?

4) In the Discussion the authors describe this as a barely-tolerable state. This coupled with the use of a relatively modest antibiotic concentration (15 ug/ml) makes me wonder about how sensitive the findings are to antibiotic concentration. It would be interesting to see if the key effect observed in Fig. 2 is maintained at higher antibiotic concentrations.

5) Single-cell resolution measurements are elegant and show the source of the survival, but cells growing in the mother machine do not compete with neighboring cells for resources. It could be interesting to repeat a key experiment in bulk cultures to show this or to speculate on how these results would look.

Reviewer #2 (Public Review):

The authors addressed the question of whether bacteria can adapt physiologically to the deletion of an essential gene using an innovative combination of light inducible recombination, single-cell time-lapse microscopy, and bulk genetic analysis. The authors grew chloramphenicol (Cp) resistant E. coli cells in a mother machine microfluidic device. At a precisely controlled time recombination was triggered causing the loss of the resistance cassette, together with a linked fluorescent marker, in a fraction of the cells. As expected, cell division stopped after the loss of the resistance cassette, but remarkably, a sizable fraction of cells (~40%) could gradually resume growth, albeit at a reduced rate. The authors recovered offspring of these cells and used batch assays (MIC measurements and PCR) to confirm that they had lost their resistance cassette and where genetically susceptible to Cp; moreover, whole-genome sequencing confirmed that no other mutations had occurred, suggesting that the observed growth in Cp was due to physiological adaptation.

The authors subsequently showed that the timing of gene deletion was essential: if the deletion happens too long before Cp treatment cells cannot adapt anymore. They thus hypothesized that cells need at least a few copies of the Cat resistance protein to be able to physiologically adapt. Finally, the authors propose that the mechanism of adaptation could be related to the stoichiometric balance of ribosomal subunits. They used single cell reporters to show that cell growth correlates with the stoichiometric balance of RpsS and RpsB (part of 50S and 30S subunit respectively); cells that lose the resistance cassette become stoichiometric unbalanced; cells that can recover growth also recover their stoichiometric balance, suggesting that these two factors are at least correlated (though a causal relation was not shown).

Overall, the manuscript is clearly written, and most conclusions are well supported by the data (however, I have some concerns regarding the sample size of one of the essential control experiments, see below). I believe this paper makes an important conceptual and methodological contribution to the field: combining light inducible recombination with single cell microscopy opens promising avenues to explore the interplay of genetic and physiological adaptation in bacteria. This can give both insight in fundamental question regarding evolutionary dynamics, as well as more practical questions regarding e.g., antibiotic tolerance and resistance.

The authors finding that cells can keep growing in normally lethal concentration of Cp, despite being genetically susceptible to this antibiotic, is also very intriguing. However, it also raises many questions that are not addressed within the manuscript. Most importantly, the question remains open what the mechanism is behind the physiological adaptation (the link to stoichiometric balance of ribosomal subunits is purely correlation based and further experiments are needed to show a causative link). Moreover, many physiological questions remain unanswered: e.g., for how long can the adapted cells keep on growing in Cp? And how quickly is the physiological adaptation lost after Cp is removed? The manuscript thus raises many new questions that remain unanswered; however, I do not see this as a major limitation as the presented work is novel and interesting as it is, and it paves the way for follow-up work by the wider community.

In my opinion, the manuscript has only one main weakness: the authors conclusions critically depend on the analysis of cells recovered from the microfluidic devices: they use this data to conclude that all mCherry negative cells have lost the cat resistance cassette (Fig. S6), however I am a bit concerned with the small number (n=5) of cells on which this conclusion is based. The cells were recovered after growing them for 6h without Cp. As Cp is a bacteriostatic drug it is conceivable that during this period also some of the growth-halted cells resume growth. The recovered mCherry negative cells could thus come both from growth-halted and growth-recovered populations. In fact, if both groups recover at the same rate, there would be about a 10% chance (62.7%^5) that all 5 mCherry negative cells would be the offspring of growth-halted lineages. Potentially there could be a difference in genotype between the growth-halted and growth recovered populations (e.g. maybe the growth recovered only lost mCherry, not cat, while growth halted lost both). Without additional information there is thus not sufficient evidence to support the authors conclusion that all mCherry negative cells observed in the microfluidic device are also cat negative.

Reviewer #3 (Public Review):

This study attempts to determine whether bacteria can "adapt to detrimental genetic modification." A E. coli strain with the CAT gene, which confers resistance to the bacteriostatic drug Chloramphenicol (Cp) was used. However, the gene was placed in such a way that the CAT gene can be removed from the genome on exposure to blue light. This is a creative way to alter the resistance levels. Although it appears that it does not work as well as chemical induction systems, there are many cases where chemical induction systems do not work. This optogenetic based method could be valuable to the field.

The conclusions reached in this paper, regarding the specific case of CAT gene loss shortly before Cp treatment, are well-supported by the data. But there are some issues to the relevance of these findings. It is already known that Cp is associated with adaptive resistance in wild-type bacteria - that is to say, the MIC is higher when the cells are exposed to a gradually-increasing concentration of the drug. This experimental control is a fancy way of probing adaptive resistance. From the perspective of the existing knowledge (i.e., Cp is associated with adaptive resistance), the findings are not particularly novel. As such, the statement "new insights into the emergence of drug-resistant bacterial and cancer cells" is not convincing. Some of these issues were mentioned (lines 222-227) but were not discussed in detail.

Reviewer #1 (Public Review):

The authors trained rats to self-initiated a trial by poking into a nose poke, and to make a sequence of 8 licks in the nose poke after a visual cue. Trials were considered valid (called "timely") only if rats waited for more than 2.5 sec after the end of the previous trial. An attempt to initiate a trial (nose poking) before the 2.5 sec criterion was regarded as "premature". The authors recorded from the dorsal striatum while rats performed in this task. The authors first show that some neurons exhibited a phasic activation around the time of port entry detected using an infrared detector ("Entry cell"), as well as port exit ("Exit cell). Some neurons showed activation at both entry and exit ("Entry and Exit cell") or between these two events ("Inside-port cell"). Fractions of neurons that fall into these four categories are roughly the same (Fig. 3C). The main conclusions drawn from this study are that (1) the activity preceding a port entry was positively correlated with the latency to initiate a trial (or "waiting time"; Fig. 4E), which appear to reflect the value upcoming reward, and that (2) in adolescent rats, the activity rose more steeply with the latency to trial initiation (Fig. 7J).

These observations are potentially interesting, in particular, the possible difference between adult and adolescent rats is intriguing. However, this study does not examine whether this brain region actually plays a role in the task. Some of the conclusions appear to be premature.

1. Previous studies have found correlations between the activity of neurons in the striatum and the latency to trial initiation (e.g. Wang et al., Nat. Neurosci., 2013) or action initiation more generally (e.g. Kunimatsu et al., eLife, 2018). In the former study, the trial initiation was self-generated, similar to the present study, and was modulated by the overall reward value (state value). In the latter study, the latency was instructed by a cue. Furthermore, there are many studies that showed correlations between striatal activity and future rewards (e.g. Samejima et al., Science, 2005; Lau and Glimcher, 2008). Many of these studies varied the value of upcoming reward (e.g. amount or probability). Although some details are different, the basic concepts have been demonstrated in previous studies.

2. The authors conclude that "in this task, the firing rate modulation preceding trial initiation discriminates between premature and timely trials and does not predict the speed, regularity, structure, value or vigor of the subsequently released action sequence". This conclusion is based on the observation that premature and timely trials did not differ in terms of kinematic parameters as measured using accelerometer. Although the result supports that the difference in activity between premature and timely cannot be explained by the kinematic variables, it does not exclude the possibility that the activity is modulated by some kinematic variables in a way orthogonal to these trial types.

3. The firing rate plot shown in Figure 4D should be replotted by aligning trials by movement initiation (presumably available from accelometer or video recording). Is it possible that the activity rise similarly between trials types but the activity is cut off depending on when the animal enters the port at different latency from the movement initiation? In any case, the port entry is a little indirect measure of "trial initiation".

4. The difference between adult and adolescent rats are not particularly big, with the data from the adolescent rats showing a noisy trace.

Reviewer #2 (Public Review):

The authors conduct an ambitious set of experiments to study how neural activity in the dorsal striatum relates to how animals can wait to perform an action sequence for reward. There are a lot of interesting studies on striatal encoding of actions/skills, and additionally evidence that striatal activity can help control response timing and time-related response selection. The authors bridge these issues here in an impressive effort. Recordings were made in the dorsal striatum on several tasks, and activity was assessed with respect to action initiation, completion, and outcome processing with respect to whether animals could wait appropriately or could not wait and responded prematurely. Conducting recordings of this sort in this task, particularly in some adolescent animals, is technically advanced. I think there is a very timely and potentially very interesting set of results here. However, I have some concerns that I hope can be addressed:

It seems like the recordings were made throughout the dorsal striatum (histology map), including some recordings near/in the DLS. Is this accurate? The manuscript is written as though only the DMS was recorded.

If I understand correctly, the rats must lick 8 times to get the water. If this is true, one strategy is to just keep licking until the water comes. Therefore, the rats may not have learned an 8-lick action sequence. The authors should clarify this possibility, and if it is, to consider avoiding using phrases like "automatized action sequence" since no real action sequence might have been learned. In short, I am not convinced the animals have learned an action pattern rather than to just keep licking once a waiting period has elapsed.

The number of subjects per group is very low. This is fine for analysis of within-animal neural activity. However, comparing the behavior between these groups of animals does not seem appropriate unless the Ns are substantially increased.

I found the manuscript difficult to decipher. There are many groups. If I understand correctly, there are the following:<br> -ITI 2.5s experiment<br> -ITI 5 s experiment<br> -ITI2.5-5s experiment<br> -ITI 2.5 s experiment (adolescent)<br> -Two accelerometer animals (unclear which experiment)<br> -Two animals in ITI 2.5 sec without recordings (unclear how incorporated into analyses)

Within each group, there are multiple categories of behavioral performance. This produces a large list of variables. In some parts of the results, these groups are separated and compared, but not all groups are compared in those such sections. In other sections the different groups (all or just some?) appear to be combined for analysis, but it is not clearly described. Another consequence of mixing the groups and conditions together in analysis as they do is that some of the statements in the results are very hard to follow (E.g., line 305 "...similar behavior observed in 8-lick prematurely released and timely unrewarded trials...").

Generally, it is difficult to understand the results without first understanding the details of the different tasks, the different groups of animals, and the different epochs of comparison for neural analysis. It took me a long time to work through the methods and I am still not sure I completely understand it. On this point, some sentences are very long and should be broken up into smaller, clearer sentences. There are a lot of phrases that only someone familiar with the cited articles might understand what they mean (e.g., even one paragraph starting with line 39 includes all of the following terms: automaticity in behavior; behavioral unit or chunk; reward expectancy; reward prediction errors and trial outcomes; explore-exploit; cost-benefit; speed-accuracy tradeoffs; tolerance to delayed rewards; internal urgency states). It is very hard to follow how each of these processes are to be understood in terms of behavioral measures used to study them and how they do or do not relate to the hypothesis of the present study. The discussion similarly uses a lot of different phrases to discuss the task and neural responses in a way that makes it hard to understand exactly what the author's interpretation of the data are. Is there maybe a 'most likely' interpretation that can be stated for some of the responses?

The data set is extremely rich; there are lot of data here. As a result it can be hard to understand how all of the data relate to the main hypothesis of the article. It often reads as an exploratory set of results section rather than a series of hypothesis tests.

Reviewer #3 (Public Review):

Cecilia-Martinez et al., implement a task that allows the study of premature versus timely actions in rats. First, they show that rats can learn this task. Next, they record the activity in the DMS showing start/stop signals in the cells recorded, next they propose that the activity detected before the release of actions sequences discriminate the premature vs the timely initiations showing a relationship between the waiting time and the activity of cells recorded, furthermore they show that it could be the expectancy of reward what could be encoded in the activity before entering the port. Last they show that adolescent rats show more premature starts than adult rats documenting a difference in activity modulation of DMS cells in the relation between waiting time and firing rate (although above the premature threshold, see comments below). 

Overall the paper is well presented describing a well-developed set of experiments.

1) I understand rats learn to execute sequences of <8licks or 8 licks, although diagrams are presented, no examples of the individual trials with 8 licks, neither distributions of bouts of these licks are presented. 

2) Relevant to the statement: "in this task, the firing rate modulation preceding trial initiation discriminates between premature and timely trials and does not predict the speed, regularity, structure, value or vigor of the subsequently released action sequence"... It is not clear if the latency to first lick (plot 2D) and the inter-lick interval (2E) is only from the 8Lick sequences or not. If that is not the case, it is important to compare only the ones with 8Licks.

3) Related to the implications of the previous statement, there seems to be a tendency for longer latency to first lick in timely vs premature trials in Figure 2D (timely-trials-Late vs premature-trials-late)? Again here it is important to compare the 8licks sequences only.

4) I could not find in the main text whether the individual points in Fig.2 (e.g. 2B-E) are individual animals. Please specify that.

5) Although very elegant the argument presented in Figure 4C and 6C, I wonder if the head acceleration may lose differences in movements outside the head in the two kinds of trials. If that is the case please acknowledge it.

6) Also in 4C, small separations between timely vs premature signals are seen before 0. Is there a way to know if animals in timely vs premature trials approached the entry port in the same way? This request is pertinent in order to rule out motor contribution to the differences in Figure 4A-B.

7) when saying: "Similar results were obtained in rats trained with a longer waiting interval (Supplementary Figure 5)", "is hard to see the similarity in the premature range, while in the 2.5 seconds task there is a positive relationship in the 5 seconds task it is not.

8) The data showing that the waiting modulation of reward anticipation grows at a faster rate in adolescent rats is clear, however, it is not clear how it could be related to the data showing that the adolescent rats were more impulsive.

9) Related to the sentence: "the strength of anticipatory activity increased with the time waited before response release and was higher in the more impulsive adolescent rats"....One may expect to see a difference in the range of the premature time however the differences were observed in the range >2.5 seconds. Please explain how to reconcile this finding with the fact that the adolescent rats were more impulsive.

Reviewer #1 (Public Review):

Using the combinatorial interpretation of single cell-based technology, the authors attempted to characterize differentiation dynamism of heterogenous mTEC clusters.

By analyzing the data, they found that Aire+ transit-amplifying TECs (TA-TECs) using Fucci system. In addition, they showed that TA-TECs could potentially differentiate into TSA-expressing functional mTEC. TA-mTECs appear to be maintained in the thymus from the youth to the adulthood.

Although they produced many datasets at single-cell resolution, the conclusion did not fully exploit the strength of integrative analysis of the transcriptome and chromatin accessibility. It is possible that the authors could obtain the same conclusion using the already existing datasets or using the data from either RNA-seq or ATAC-seq because their dataset itself was not different from the ones in many previous publications, as they mentioned in the manuscript.

Nevertheless, the high-quality data of single cell-based technology in the present study must be valuable for researchers in the field. Their finding of "transit-amplifying Aire+ TSA-low mTECs as a potential precursor population of functional mTECs" can contribute to our understanding of the mTEC development mechanism that plays a key role in immune tolerance.

Reviewer #2 (Public Review):

The paper is a careful and solid analysis to identify the cycling precursor cell for Aire-positive medullary thymic epithelial cells (mTECs) in the postnatal thymus. The study uses an extensive set of single-cell analyses and importantly, the Fucci2a mouse model, which enables the identification of exact transit-amplifying cell type. Overall this is an important study as it shows the differentiation dynamics and heterogeneity of mTECs in the adult thymus.

1. The scATACseq experiment included a relatively low number (n=2) of mice. Understandably this is a complicated experiment, however, the variation between individual animals remains unknown and may influence the interpretation of the results. The authors are advised to show the clustering analysis for the two animals separately in a supplement to confirm that the changes are seen in both animals, not just in one. Similarly, the individual scRNA-seq UMAPs of the three animals should be included in the supplement. The authors should discuss the limitations of small groups in single-cell experiments in the Discussion. Technically it is positive to see the integration of scATAC and scRNA seq results.

2. Genes expressed in clusters should be provided in the supplement. Without this information, it is difficult to decide about the cluster annotations and to compare them with other similar studies. The authors should give an unbiased list of all genes expressed in mTECs, in particular those genes that are expressed in TACs. To harmonize the findings in the field, it would be useful for the readers if the authors would compare their cluster-specific genes for the overlap with TEC single-cell clusters from other studies (Bornstein et al 2018, Dhalla et al 2020, Baran-Gale et al 2020, Wells et al 2020).

3. As a proliferating cell type, TACs should express multiple genes associated with cell cycling. The authors focus on Mki67 only but did R1 TAC express other proliferating cell markers, which would support the claim that these are indeed actively dividing cells?

4. The candidate cell population for TACs, cluster R1 expresses Aire and the proliferating cell marker Mki67. R1 also expresses Ccl21a. The authors did subclustering of R1 and found that R1A-D express Aire whereas R1E has a higher expression of Ccl21a. The authors note that "Thus, it is possible that TECs expressing cell-cycle-related genes, proposed by scRNA-seq analysis, contain at least two proliferating TECs subsets having different chromatin accessibilities and gene expression profiles." To confirm that both R1A-D and R1E subsets are proliferating TACs, the authors might show the proliferating gene markers in these subsets. Was Mki67 expressed among all R1 subpopulations? Would this argue for the presence of TAC among both Aire+ and Aire- cell populations?

5. Wells et al 2020 (ELife) recently reported mTEC TACs to be mTEClo preceding Aire expression and giving rise to both Aire+ and Ccl21a+ cells. How do the authors reconcile their results (TAC as Aire+) with Wells et al paper (TACs as Aire-)? How to interpret the finding that Wells et al find Ki67 significantly higher in mTEClo than in mTEChi, implying that cell division precedes high the expression of Aire?

6. In RTOC experiment adult mCherry-lo cells are transferred into embryonic thymic organ culture where they turn mCherry-hi (Figure 6A). The authors show this to confirm the differentiation of mCherry-lo to mCherry-hi. This is indeed a strong support for their conclusion, but would the authors agree that transferring adult mTECs into the embryonic cell environment may trigger other signaling pathways that induce the upregulation of the cell cycle and mCherry expression? Would the same happen if they would transfer these cells to the adult thymus?

Reviewer #3 (Public Review):

This study carefully addresses the questions raised in the introduction and provides convincing evidence:<br> - confirming the existence of a cycling mTEC population expressing Aire and clustering between/ next to mTEChi amd mTEClo.<br> - showing that this cycling population is sustained by a pattern of chromatin accessibility different from those corresponding to mTEChi or mTEClo.<br> - demonstrating that the cycling CD80high mTECs that the authors isolate are TACs giving rise to mature mTEChi and post-Aire mTECs in RTOC systems, thereby validating ex vivo the previously known in silico-inferred trajectory of cycling mTECs.

Using "revealed" in the title may be misleading, letting think that it's the first time that the cycling mTEC population expressing Aire is identified by single-cell approaches (Wells 2020, Dhalla 2020, Baran-Gale 2020).

The conclusions of the paper are mostly well supported by the data. However, some points need clarification:

1) The authors nicely confirm, by the fine analysis of their scRNA-seq data, that the TAC pop contains Aire+ cells and Ccl21+ cells in a visually mutually exclusive manner. However, they don't formally clarify whether the Ccl21-expressing TACs have differences in their chromatin accessibility pattern compared to the Aire-expressing TACs.<br> It's also worth showing the expression of CD80 in the scRNA-seq UMAP of TACs alone and in the one of all TECs (Fig2). This would notably allow to determine whether CD80 expression is restricted to Aire-positive TACs or encompasses Aire-negative TACs (Ccl21+).

2) Trajectory analyses provide a nice confirmation of published results identifying a trajectory from TACs to mTEChi. However, the authors don't discuss whether their data support a potential trajectory from TACs to mTEClo (already suggested/ reported) which seems to be present in Fig 3-fig sup 2B. Would this mean that some TACs could mature into Ccl21+ mTECs (mTEClo)? and if so, how Aire and Ccl21 are expressed in these TACs? Which TAC sub-cluster do they belong to?

3) The authors focus their study on the CD80high cycling cells (Aire+). Figure 4 show transcription profiles of the isolated cycling CD80+ mTECs. Dots in the "TSA genes" panel (right) don't appear in the "All genes" panel (left). Are those lost dots of TSA genes?<br> In Fig 4E (left) low expressed genes seem to be skewed towards Venus- mTEChi (in comparison to high expressed genes). A statistical assessment for the comparisons of the TSA, Aire-dep TSA and Aire-indep TSA profiles to the general profile (Fig 4E and 4G), considering expression levels, would comfort the visual assessment.

4) In Fig4F, Aire expression is similar between Venus+ and Venus-. Is it compatible with Fig 4A showing less Aire-expressing cells in Venus+ than in Venus-?

Reviewer #1 (Public Review):

A) Main findings:

The authors report that extravesicles (EVs) including exosomes (Exos) provide a new source for membrane expression of CD47 in a process termed as "cross-dressing".

B) Strengths:

The experiments reported provide compelling evidence supporting:<br> - the existence of CD47 cross dressing in vitro.<br> - the functionality of CD47 (in phagocytosis inhibition assay) when transferred after cross dressing.

C) Weaknesses:

In vitro settings. The persistence of cross-dressed CD47 at the surface of recipient cells is not addressed.

The cellular mechanisms underpinning the transfer of CD47-containing membranes is not fully described nor mechanistically explained.

The lack of pro-apoptotic signalling is not explained on a mechanistic basis. For instance, does this implies that cross-dressing does not involve membrane fusion thereby preventing the access of CD47 ICD to intracellular machineries and signal transduction within host cells? Determining whether EVs/Exos tether on membranes of cross dressed cells or provide integral membrane proteins to the target cell membrane (post membrane fusion) is a key point that is not addressed at all within the manuscript.

The role of exosome release in providing membranes for CD47 is not assessed in physiological settings (exosomes released by cells in vitro or in vivo e.g.). Instead preparation of concentrated EVs/exosomes are used and it is questionable if this recapitulates physiologically relevant concentration of exosomes/EVs.

D) Conclusions:

The conclusions raised from the experiments provided are sound.

E) Impact & significance:

Inhibitory pathways activated upon the SIRPa signalling downstream CD47 ligation are of key relevance for the regulation of xenorejection and allorejection at both innate (phagocytosis) and adaptive (antigen presentation by DCs).

The prospect of uncoupling the inhibition of phagocytosis from apoptosis induction at the level of target cells is attractive. However, one might wonder if engineering of CD47 intracellular domain might not achieve the same goal more directly. Are cells expressing a tail-less CD47 would be protected from SIRPa-mediated apoptosis as well?

Also, it is a little bit unclear how this process unveiled in this study can be practically implemented in transplantation and regenerative medicine, especially taking in account the persistence of cross-dressed CD47 on recipient cells which is not assessed in this study.

Reviewer #2 (Public Review):

In cell culture, membrane is known to be shed as vesicles from living and dying cells, and such vesicles have the potential to fuse with other cells and confer novel function. CD47 and SIRPa are integral membrane proteins expressed at low to medium levels on many cell types, and if they are shed and properly fuse with another membrane, then they could be expected to affect phagocytosis of the recipient cell. Such in vitro observations made here are potentially impactful for the CD47 field, and although many mechanistic aspects remain unclear as are in vivo effects, the results do extend past work that has already shown shed membrane particles have functional CD47 that - by design - tends to integrate into recipient membrane (PMID: 28053997). However, a number of concerns temper enthusiasm for the present submission.

1. Recent studies with particles shed from hCD47-overexpressing HEK cells, as a mix of encapsulated virus and simpler vesicles, showed that non-phagocytic cells (HEX cells, A549 lung cells) take up such particles -- with membrane fusion being the common pathway, and they also take up more when SIRPa is present and not blocked (Fig.2b,e,f in PMID: 28053997). This is counter to this submission's claim that "hCD47 cross-dressing is SIRPα-independent".

2. This study does not provide clear mechanism for many key observations, even though demonstrations of mechanism(s) are usually foundational to the reproducibility and rigor of results. For examples: (A) It is unclear what the basis is for the observation that "ligation of the autogenous, but not cross-dressed, CD47 induced cell death." (B) Given the claimed role for hCD47 transfer by extracellular vesicles and/or exosomes, it is unclear whether the process involves any transfer of hCD47 mRNA or even DNA (e.g. plasmid, even episomal). (C) It is unclear why there is a very broad distribution of intensities on cells that supposedly have 'all' received hCD47; indeed, some cells have little to no hCD47 (e.g. Fig.1c). (D) It is unclear how hCD47 transfers from "florescence Celltrace violet" cells to null cells (Fig.1c) without also transferring some of the dye. (E) It is unclear why results for h2 and h4 spliceforms are sometimes the same (e.g. Fig.1a) versus sometimes different (Fig.1b).

3. Microscopy images always provide more rigorous and reproducible evidence of phagocytosis than flow cytometry assays -- especially because of the use of dyes, given the concerns mentioned above. A further concern with Fig.4c,d is that the null or KO cells with hCD47 inhibit phagocytosis to an equal extent as control, even though the transferred levels of hCD47 tend to be lower and also heterogeneous; those cells showing very low transfer levels (e.g. Fig.1,2b,S5c, etc.) are expected to be engulfed more.

Reviewer #3 (Public Review):

Here Li et. al. investigates the effects of cross-dressing pig cells with human CD47 in the cocultures of human monocytes and pig cells. Using hCD47 transgenic cell lines, Jurkat cells that constitutively express high levels of CD47, the authors have shown that pig cells can be cross-dressed with hCD47 molecules via a transfer of extracellular vesicles. This transfer can occur in the absence of SIRP-a. While the in vitro evidence presented here may provide a new strategy for minimizing anti-xenograft immune responses by macrophages/ monocytes, the study would immensely benefit from in vivo evidence that addresses whether cross-dressing can indeed be sufficient to protect xenogeneic cells in the host and that the CD47 molecules that land on xenogeneic cells remain on the surface of acceptor cells stably enough. Once the functional relevance of CD47 cross-dressing is carefully addressed and substantiated, the study would provide valuable knowledge to transplantation immunology field.

Reviewer #1 (Public Review):

Kumar et al., combined both RNA structure probing methods and computational approaches to generate the prediction model for splicing. The framework is novel, and the prediction model is interesting. The weakness is the interpretation of DMS reactivity profile which may be affected by the interacting proteins. Overall, the authors achieved their aims but the results require careful interpretations. The framework and the model established in this manuscript is very useful to the field. The capability for predicting splicing will enhance our knowledge on the disease-related mutations.

Reviewer #2 (Public Review):

Kumar et al. developed an intricate framework to predict how a given mutation will affect splicing at exon-intron junctions. The leveraged the power of Mutational Profiling to obtain in vivo secondary and tertiary structure of RNA molecules and the spliceosomal footprints at these junctions. The structural data alone was only able to predict splicing outcome with 72% accuracy. The authors then explored how well the strength of splice sites and splicing regulatory elements predicted splicing outcome. These two features only predicted accuracy of R2 = 50 in splicing events. However, when combining splice site strength, splicing regulatory element strength, and structural data yielded the highest prediction accuracy of R2 = 89. The authors demonstrate the importance in considering both structural and splice site and regulatory element strength when assessing splice-altering mutations. The data adds to the ongoing studies on how to predict, manipulate, and measure splicing.

Strengths:<br> The framework used to create the predictive models are thoughtful and rigorous in the context of MAPT gene. Upon training the model, the authors chose to run 55 variants of unknown significance (VUSs) through the model to predict which of the 3 splicing outcomes (3R, 4R, WT) would occur. They then used a splicing assay to experimentally confirm the computationally predicted outcomes of 6 of these VUSs.<br> The techniques in principle are widely applicable across cell types and any genomic sequences of interest. This framework could be leveraged to determine how known mutations affect splicing, or how gene editing can be used to manipulate splicing events at exon-intron junctions of interest.<br> As a clinically relevant exon-intron junction, the exon10-intron10 junction of MAPT gene serves as an ideal model for developing the framework.

Weaknesses:<br> The authors places emphasis on MAPT gene and its relation to neurodegenerative diseases in the introduction and first figure, but then make broad claims about the applicability of the framework. While known and unknown variants within and surrounding the exon10-intron junction in MAPT were extensively studied and add to the validity of the models, applying the framework to a second exon-intron junction within a different gene would have strengthened the claim that this framework can be used as a general predictive model of splicing.

Reviewer #3 (Public Review):

The manuscript by Kumar et al. describes a modeling study of a well-studied system, namely, the MAPT Exon 10-Intron 10 junction, whereby splicing disruption has been associated with multiple neurodegenerative diseases. It thus addresses an important, fundamental problem, which is of interest to a broad readership. Despite having been studied extensively, splicing is still poorly understood, and in particular, the role of RNA structure is difficult to systematically or quantitatively highlight. The authors developed a comprehensive and highly accurate predictive model of MAPT splicing, which integrates both structural and regulatory sequence motifs. They not only used it to predict measured PSI (percent spliced in index) values, but more interestingly, they used it to dissect the contributions of structure and SRE RBP binding sites to splicing disregulation and to reveal their counteracting effects. This, in turn, sheds light on how disease variants affect splicing outcomes. Additionally, they experimentally validated select model predictions.

Overall, this is an insightful manuscript that introduces several novelties. On the experimental side, the authors performed the first experimental structure determination in vivo of MAPT's two relevant isoforms and its pre-mRNA (as opposed to previous in vitro or computational studies). There is also innovative use of the unique features of the MaP structure probing strategy (as opposed to truncation-based probing) to determine structure at the isoform level while obviating the need to deal with read mapping ambiguities. On the computational side, I liked the novel use of t-SNE and density contour plots to visualize shifts in ensemble composition and how the authors integrated cryo-EM data with computational ensemble-based structure predictions. Both innovations can be useful in a range of other structure-function studies. I also find the final model's accuracy and its success in making robust predictions for all mutation types impressive. Although additional DMS-MaP experiments that could validate some of the predicted mutation-induced structural changes would be ideal, because changes take place at the pre-mRNA level, it is much more complicated to perform such in vivo assays.

Reviewer #1 (Public Review):

The submission "Molecule structure and conformation of stereocilia tip-links elucidated by cryo-electron tomography" represents an attempt to better delineate the structure of the molecular apparatus-the transduction complex-by means of which mammalian hair cells respond to mechanical stimuli. In particular, the authors use cryo-electron microscopic tomography to examine unfixed preparations of murine hair bundles that have been pressed against grid films. By labeling protocadherin 15 (PCHD15) with a novel immunogold construct, they demonstrate the presence of dimeric and possibly monomeric PCDH15 molecules interacting with stereociliary membranes, membranous vesicles, and other filamentous structures perhaps including cadherin 23 (CDH23) molecules.

1) In a broad sense, any filamentous structure that bridges two stereociliary tips could be termed a "tip link." The conventional definition, though, restricts the term as defined by the authors on lines 21-22. Using the term more broadly risks erroneous conclusions: especially during murine development, there are clearly a wide variety of filamentous structures at stereociliary tips, nearly all of which disappear within the first two weeks of postnatal life. The term "tip link" should be stricken from instances in which the nature of the structures is not rigorously documented. This includes the submission's title and sections such as that beginning at line 196, "Tomographic reconstructions of PCDH15-CDH23 heterotetramers." The authors have not confirmed the identities of such structures, and in particular have not demonstrated the presence of CDH23. The same holds for the section commencing on line 217, "An intact tip-link surrounded by non CDH23-bound PCDH15 molecules." The single structure in question is described as "most likely an intact tip-link," but that degree of proof does not inspire confidence in either the section title or indeed the title of the submission itself.

2) The assertion on line 253-254 that "this work is the first look at native tip-links in situ" is unsubstantiated: the authors are staking an unjustified claim of priority. Although such an assertion should be supported by several well-documented structures, the actual evidence amounts to one "putative" tip link. Moreover, the structures are scarcely "native," given the violence of the bundle-blotting procedure and the acknowledged damage as a result; nor are they "in situ," which would imply their presence in a mouse's inner ear, or at least in living, functioning hair cells. The authors should describe what they have nicely demonstrated, the immunologically confirmed occurrence of PCDH15 molecules-many of them dimers-near stereociliary tips, and the association of those molecules with other as yet unidentified filamentous and lipidic structures.

3) The preparative techniques are sound and the tomographic images of generally high quality, as manifested for example by the smooth membrane profiles and nicely ordered actin microfilaments of the stereociliary cytoskeletons. Nevertheless, the intrepretation of the data is shaky. Figure 3C gives striking evidence for a dimer of PCDH15. It is not at all clear, either in the figure or in the associated video, that the structure in panel (B) shows two labels; and that in panel (C) shows a dense blur, perhaps a cluster of gold particles. It is likewise unclear that there is any label bound in panel (D). In view of the ambiguous micrographs, a reader might well be skeptical about the ensuing statistical treatment.

Reviewer #2 (Public Review):

The manuscript presents exquisite images of PCDH15 in its native state, confirming the predictions of earlier studies. While much of what is presented here was predicted from less-direct measurements, it is gratifying to confirm these with direct imaging of single molecules. Specifically, the paper demonstrates that PCDH15 in its native state at the tip of a stereocilium is a dimer; it shows full tip links with PCDH15 bound to strands likely to be CDH23 with N-terminal binding; it shows that PCDH15 can bend in vivo at a location at or near the EC9-EC10 junction; it shows how that bending could allow a tip link to extend upwards from a location below a stereocilium tip; it shows a protein likely to be CDH23 bending near the middle of its extracellular domain; it confirms that the dense cross links of immature bundles do contain PCDH15; and it give some sense of the number of PCDH15 molecules on each stereocilium.

Overall, the claims are supported by the images. The proteins of interest are at the limit of resolution of the cryo tomography, so we do have to trust that the "annotated" tomograms-really the authors' conception of the tomograms-are fair representations. At the same time, studying over 500 tomograms probably trained the authors' eyes to the point that they could see detail that a first-time viewer cannot.

Reviewer #1 (Public Review):

Cell surface proteins are of vital interest in the functions and interactions of cells and their neighbors. In addition, cells manufacture and secrete small membrane vesicles that appear to represent a subset of the cell surface protein composition.

Various techniques have been developed to allow the molecular definition of many cell surface proteins but most rely on the special chemistry of amino acid residues in exposed on the parts of membrane proteins exposed to the cell exterior.

In this report Kirkemo et al. have devised a method that more comprehensively samples the cell surface protein composition by relying on the membrane insertion or protein glycan adhesion of an enzyme that attaches a biotin group to a nearest neighbor cellular protein. The result is a more complex set of proteins and distinctive differences between normal and a myc oncogene tumor cells and their secreted extracellular vesicle counterparts. These results may be applied to the identification of unique cell surface determinants in tumor cells that could be targets for immune or drug therapy. The results may be strengthened by a more though evaluation of the different EV membrane species represented in the broad collection of EVs used in this investigation.

Reviewer #2 (Public Review):

This paper describes two methods for labeling cell-surface proteins. Both methods involve tethering an enzyme to the membrane surface to probe the proteins present on cells and exosomes. Two different enzyme constructs are used: a single strand lipidated DNA inserted into the membrane that enables binding of an enzyme conjugated to a complementary DNA strand (DNA-APEX2) or a glycan-targeting binding group conjugated to horseradish peroxidase (WGA-HRP). Both tethered enzymes label proteins on the cell surface using a biotin substrate via a radical mechanism. The method provides significantly enhanced labeling efficiency and is much faster than traditional chemical labeling methods and methods that employ soluble enzymes. The authors comprehensively analyze the labeled proteins using mass spectrometry and find multiple proteins that were previously undetectable with chemical methods and soluble enzymes. Furthermore, they compare the labeling of both cells and the exosomes that are formed from the cells and characterize both up- and down-regulated proteins related to cancer development that may provide a mechanistic underpinning.

Overall, the method is novel and should enable the discovery of many low-abundance cell-surface proteins through more efficient labeling. The DNA-APEX2 method will only be accessible to more sophisticated laboratories that can carry out the protocols but the WGA-HRP method employs a readily available commercial product and give equivalent, perhaps even better, results. In addition, the method cannot discriminate between proteins that are genuinely expressed on the cell from those that are non-specifically bound to the cell surface.

The authors describe the approach and identify two unique proteins on the surface of prostate cell lines.

Strengths:

Good introduction with appropriate citations of relevant literature<br> Much higher labeling efficiency and faster than chemical methods and soluble enzyme methods.<br> Ability to detect low-abundance proteins, not accessible from previous labeling methods.

Weaknesses: The DNA-APEX2 method requires specialized reagents and protocols that are much more challenging for a typical laboratory to carry out than conventional chemical labeling methods.

The claims and findings are sound. The finding of novel proteins and the quantitative measurement of protein up- and down-regulation are important. The concern about non-specifically bound proteins could be addressed by looking at whether the detected proteins have a transmembrane region that would enable them to localize in the cell membrane.

Authors mention time-sensitive changes but it is unclear how this method would enable one to obtain this kind of data. How would this be accomplished? The statement "Due to the rapid nature of peroxidase enzymes (1-2 min), our approaches enable kinetic experiments to capture rapid changes, such as binding, internalization, and shuttling events." Yes, it is faster, but not sure I can think of an experiment that would enable one to capture such events.

The authors do not have any way to differentiate between proteins expressed by cells and presented on their membranes from proteins that non-specifically bind to the membrane surface. Non-specific binding (NSB) is not addressed. Proteins can non-specifically bind to the cell or EV surface. The results are obtained by comparisons (cells vs exosomes, controls vs cancer cells), which is fine because it means that what is being measured is differentially expressed, so even NSB proteins may be up- and down-regulated. But the proteins identified need to be confirmed. For example, are all the proteins being detected transmembrane proteins that are known to be associated with the membrane?

The term "extracellular vesicles" (EVs) might be more appropriate than "exosomes" to describe the studied preparation.

Reviewer #3 (Public Review):

The article by Kirkemo et al explores approaches to analyse the surface proteome of cells or cell-derived extracellular vesicles (EVs, called here exosomes, but the more generic term "extracellular vesicles" would be more appropriate because the used procedure leads to co-isolation of vesicles of different origin), using tools to tether proximity-biotinylation enzymes to membranes. The authors determine the best conditions for surface labeling of cells, and demonstrate that tethering the enzymes (APEX or HRP) increases the number of proteins detected by mass-spectrometry. They further use one of the two approaches (where HRP binds to glycans), to analyse the biotinylated proteome of two variants of a prostate cancer cell line, and the corresponding EVs. The approaches are interesting, but their benefit for analysis of cells or EVs is not very strongly supported by the data.

First, the authors honestly show (fig2-suppl figures) that only 35% of the proteins identified after biotinylation with their preferred tool actually correspond to annotated surface proteins. This is only slightly better than results obtained with a non-tethered sulfo-NHS-approach (30%). Indeed the list of identified proteins in figures 4 and 5 include several proteins whose expected subcellular location is internal, not surface exposed, and whose location in EVs should also be inside (non-exhaustively: SDCBP = syntenin, PDCD6IP = Alix, ARRDC1, VPS37B, NUP35 = nucleopore protein)... The membrane proteins identified as different between the control and Myc-overexpressing cells or their EVs, would have been identified as well by a regular proteomic analysis.

Second, the title highlights the benefit of the technique for small-scale samples: this is demonstrated for cells (figures 1-2), but not for EVs: no clear quantitative indication of amount of material used is provided for EV samples. Furthermore, no comparison with other biotinylation technics such as sulfo-NHS is provided for EVs/exosomes. Therefore, it is difficult to infer the benefit of this technic applied to the analysis of EVs/exosomes.

In addition, the WGA-based tethering approach, which is the only one used for the comparative analysis of figures 4 and 5, possibly induces a bias towards identification of proteins with a particular glycan signature: a novelty would possibly have come from a comparison of this approach with the other initially evaluated, the DNA-APEX one, where tethering is induced by lipid moieties, thus should not depend on glycans. The authors may have then identified by LC-MS/MS specific glycan-associated versus non-glycan-associated proteins in the cells or EVs membranes. Also ideally, the authors should have compared the 4 combinations of the 2 enzymes (APEX and HRP) and 2 tethers (lipid-bound DNA and WGA) to identify the bias introduced by each one.

As presented the article is thus an interesting technical description, which does not convince the reader of its benefit to use for further proteomic analyses of EVs or cells. Such info is of course interesting to share with other scientists as a sort of "negative" or "neutral" result.<br> Maybe a novelty of the presented work is the differential proteome analysis of surface enriched EV/cell proteins in control versus myc-expressing cells. Such analyses of EVs from different derivatives of a tumor cell line have been performed before, for instance comparing cells with different K-Ras mutations (Demory-Beckler, Mol Cell proteomics 2013 # 23161513). However, here the authors compare also cells and EVs, and find possibly interesting discrepancies in the upregulated proteins. These results could probably be exploited more extensively. For instance, authors could give clearer info (lists) on the proteins differentially regulated in the different comparisons: in EVs from both cells, in EVs vs cells, in both cells.

Reviewer #2 (Public Review):

In this work, Lechtreck et al showed that the IFT-dependent transport of radial spokes in Chlamydomonas requires ARMC2/PF27. They first identified the pf27 mutant strain as an armc2 mutant strain and showed that radial spoke localization in this mutants is restricted to the proximal part oft he flagella. The authors showed by using tagged ARMC2 and the radial spoke protein RSP3 that ARMC2/PF27 is highly enriched in growing flagella and that ARMC2/PF27 and RSP3 comigrated on anterograde trains. Furthermore, after unloading at the flagellar tip, only ARMC2/PF27 diffused back to the cell body and RSP3 was attached to the flagellar axoneme, supporting the role of ARMC2/PF267 as an IFT cargo adapter specific for radial spoke proteins. The authors also show that ODA16 (an outer dynein arm adapter protein) and IDA3 (an inner dynein arm cargo protein) move independently of ARMC2/PF17 indicating that unrelated cargoes are distributed stochastically onto IFT trains. The conclusions of this paper are well supported by the results.

Reviewer #1 (Public Review):

Building cilia requires bidirectional intraflagellar transport (IFT) that moves components in and out of the growing organelles. This process involves multi-component IFT trains being driven by kinesin and dynein motors. Cargos are known to attach to these trains allowing them access to the distal cilium where offloading generally occurs. This process provides a mechanism by which axonemal components reach the site of assembly. Defects in this process have profound impact on mammalian physiology and development. Chlamydomonas is an extremely powerful model system in which to dissect this complex process as numerous genetic, biochemical and microscopic tools can be readily applied. Although numerous proteins have been shown to require IFT for their transport into cilia, how many different components interact with the IFT trains to achieve this and whether those associations are inter-dependent or even exclusionary remains poorly understood. In this manuscript, the authors take advantage of a Chlamydomonas mutant, pf27, which assembles normal looking radial spokes (RS) but only in the very proximal part of the organelle even though reconstitution experiments indicate that the assembly sites in the remainder of the axoneme are fully functional. Using a candidate gene approach, they identify the gene product PF27 as ARMC2 a protein containing multiple armadillo repeats and whose dysfunction in mammals is known to disrupt lung function and yield male infertility. They find that ARMC2 and RS are trafficked together by IFT, while in the absence of RS, ARMC2 is moved by itself. The conclusion is that ARMC2 is an adaptor linking RS to IFT trains. Assembly of RS in the proximal region of pf27 cilia is proposed to occur by diffusion of RS from the cell body. This would seem to imply a certain leakiness to the ciliary gate even for large mega-dalton sized complexes. The authors suggest that IFT-adaptor interactions are critical for control of cargo entry into cilia. While that certainly seems quite reasonable, control of the adaptor-cargo association is likely equally important for productive transport, off-loading and axonemal incorporation.

Using fluorescent tags and TIRF microscopy, the authors observe that ARMC2 is moved by IFT to the ciliary tip but is there off-loaded and diffuses back to the cell body - this has been seen with several other ciliary proteins previously. They carefully characterize the co-transport and release of ARMC2 and a RS marker protein RSP3, and clearly demonstrate that RSP3 transport fails in the absence of ARMC2.

One important issue is whether individual IFT trains are dedicated to a particular cargo or set of cargoes (i.e. are active), or if all trains have similar cargo binding capacity and associations of individual complexes with any given train is stochastic. The experimental results reported for ARMC2 and outer arm dynein closely fit values derived from calculated probabilities of chance interactions with any given train. This is an important result and is well supported by the numeric data.

Reviewer #3 (Public Review):

In this manuscript the authors make several conclusions, according to the abstract:

1 - LTG activity is essential by contributing to a process independent of PG recycling.<br> 2 - LTGs are important because of their catalytic activity rather than because of a protein-protein interaction.<br> 3 - LTG mutants are hypersusceptible to production of periplasmic polymers.<br> 4 - LTGs prevent toxic periplasmic crowding and their function is temporally separate from PG synthesis.

The authors perform a series of genetic experiments that lead to their conclusions. Their first conclusion is well supported by data showing that a PG recycling mutant does not have the same defects as their LTG mutant.

Their second conclusion needs more justification/explanation. They show a catalytic mutant of RlpA is unable to sustain growth as the only LTG in the cell. However, I am confused by their wording around RlpA in general. In the text they note that their delta_7 mutant, which encodes RlpA, 'has no highly active LTGs' (lines 130-131). Does that imply that RlpA is not an LTG? In the discussion they note that E.coli RlpA has no LTG activity. Is this enzyme known to have LTG activity in V.cholerae? One important control would be to show that the catalytically inactive protein is stable (i.e. that the defect is not due to protein misfolding). This could be supported by looking at protein stability via Western or even quantifying the fluorescence data in Figure S3b.

Their third conclusion also needs more support. The authors do a series of experiments showing that delta7 is more susceptible to SacB. What are the data that show sacB produces large polysaccharides molecules in the periplasm rather than (or in addition to) the cytoplasm? This would be important to show as these data are the main test of the authors model.

The authors have other data that all argue for their model that LTG deficient strains have an excess of periplasmic crowding. The suppressor of delta_opgH is intriguing, but does not restore the morphological defects in delta_7, suggesting that the increase in length during prolonged growth may not be caused by periplasmic crowding, or at least is not alleviated by deletion of OpgH. What then does the deletion of OpgH suppress? Here, I was confused by the experiments in low salt. The authors write that the cells lyse (line 222) but this is not shown anywhere. Growing the cells continually in low salt may not be the hypoosmotic challenge the authors presume. A challenge typically implies an acute change in osmolarity, rather than a prolonged exposure, which may allow cells to adapt.

I was also highly confused by the antibiotic + BADA staining experiments. Do the authors stain the cells, treat, and then visualize? Are they then studying the fate of old PG? How does BADA get incorporated into PG in V.cholerae? Is it through LDT activity or some other way? Without more explanation, it is hard to interpret the results.

The last conclusion is not supported by data. There are no data showing that LTG activity is temporally separate from PG synthesis.

Reviewer #1 (Public Review):

The authors used a combination of multiple gene knock-outs of lytic transglycosylsases with extensive phenotyping of the mutant strains to conclude that the essential role of Ltgs for bacteria is to reduce periplasmic crowding with soluble macromolecular fragments of peptidoglycan. The extensive phenotyping is based on complementation assays, growth phenotypes, morphological phenotypes, analysis of suppressor phenotypes, sensitivity to accumulation of periplasmic polysaccharides and to antibiotics promoting the peptidoglycan futile cycle, and finally, analysis of soluble peptidoglycan accumulated in the periplasm.

The work is well written, easy to follow and technically sound. Collectively, it provides circumstantial evidence for the authors final model.

This work supports previous seminal work on the futile cycle of peptidoglycan synthesis in the presence of beta-lactams but by extending it to homeostasis and not only during cell envelop stress.

Reviewer #2 (Public Review):

The paper "Lytic transglycosylases mitigate periplasmic crowding by degrading soluble cell wall turnover products" by Weaver et al. reveals multiple findings, some of which are transformative for the cell wall field. 1) First and foremost, they finally nail down the role of lytic transglycosylases in the evolution of the cell wall, revealing that they cleave and process uncrosslinked glycans arising from endopeptidase activity. 2) They also document the periplasmic stress that occurs when these strands are not degraded, raising interesting future questions about the role of glycans in the resistance to osmotic stress and perhaps the balance of turgor pressure in the cytoplasm. 3) This work also suggests that the lytic transglycosylases are separate from the synthetic enzymes, further adding to the "disconnected" model of the enzymes that synthesize and process the cell wall.

I very much appease a large amount of work done here and that the authors were able to infer some sort of "cumulative damage" from their initial experiments, and then figure out the non-obvious cause behind these phenomena.

The experiments are all well documented, with a large amount of data supporting their claims, all of which appear well supported, and the more speculative claims are phrased as such.

Reviewer #3 (Public Review):

In this study, Roy et al., have focussed on investigating what happens in infections associated with impaired healing in diabetic wounds. Specifically, they have identified that a certain type of white blood cell (i.e. a neutrophil) is dysfunctional, leading to a delay in its ability to help fight off the infections.<br> The findings of the study are interesting, and the therapeutic possibility of treating diabetic wounds with CCL3 is novel and is likely to be of interest to the field. However, the very low n numbers used in the study questions the validity and robustness of the data.

Reviewer #1 (Public Review):

Roy et al. investigated glucose-induced changes of selected neutrophil functions using neutrophils from diabetic mice and glucose-exposed murine and human neutrophils. They reconfirm earlier findings that glucose renders neutrophils less responsive to fMLF-mediated chemotaxis and show that expression and surface presentation of the corresponding receptor FPR1, a chemotactic receptor that is high in the signaling hierarchy, is downregulated within the first hour of glucose treatment. Similarly, other elements of neutrophil chemotactic responses including the phospholipase PLC and the cytokine MIP-1/CCL3 are also affected, while the expression of the chemokine receptor CCR1 remains unaltered. Interestingly, supplementing the CCFR1-targeting cytokine CCL3 could restore neutrophil chemotactic fitness and wound healing and thus, might be beneficial for diabetic wound management.

Conclusions are supported by the data but the study, in its current stage, needs further analysis. The findings suggest a more general effect on the neutrophil expression pattern induced by glucose and unfortunately, this is not addressed and mechanistic insights to explain the observed effects are entirely missing. The finding that CCL3 levels are reduced and that external addition brings neutrophil chemotactic response back to normal is of high translational potential.

Reviewer #2 (Public Review):

Diabetic wound closure remains a major clinical problem, in the sense that in diabetes, skin wounds do not repair on time and may get infected, installing a feed-forward cycle of inflammatory and tissue damage outcomes. This study chiefly demonstrates that this vicious cycle can be broken by ensuring an exuberant and effective neutrophilic response takes place.

As such the data presented herein are exemplar in demonstrating: i) the substantial non-redundant role of physiological inflammation (that is, a good inflammatory response when needed is a good thing); ii) the fact that tissue inflammation resolves after a proper and effective neutrophil response and iii) a specific receptor target (FPR1) is affected by diabetes (high glucose) and is central to a defective inflammatory response to infective agents.

All in all I found the experimental model used very helpful in demonstrating this important physio-patholoigical link between: inflammation onset -> inflammation resolution.

Reviewer #1 (Public Review):

This study aims at better characterizing the synaptic and cellular integration mechanisms orchestrating the interplay between the cortical feedback arriving in the shell of inferior colliculus (IC) and the feedforward input. Combining optogenetic tools with in vivo and in vitro patch clamp the authors describe the strength and kinetics at different time scales of corticofugal inputs, and theoretically deduce based on this that the top-down and bottom-up inputs are ideally timed to collide in the shell of the IC. They then show that coincidence of top-down and bottom-up inputs leads to supra-linear summation in an NMDA-dependent manner. The data provided in the first descriptive part of the paper is sound and useful for the community to better model corticofugal projections. The finding that coincident arrival of corticofugal and bottom-up inputs leads to non-linear boosting of IC neurons is sound and represents a very important result for further understanding and modeling of corticofugal inputs to IC, even if the authors did not illustrate it with a concrete example during realistic auditory processing where this phenomenon would be at play (some hypotheses are given). The NMDA dependence is even more striking in that it depends on receptors located on bottom-up synapses, a surprising and mind-blowing result.

Reviewer #2 (Public Review):

Oberle et al. provide a detailed analysis of how descending projections from the auditory cortex interact with ascending auditory projections on neurons in the shell region of the inferior colliculus on a cellular basis. Using optogenetic activation of auditory cortical neurons or projections and electrical stimulation of fibres in combination with whole-cell patch clamp recordings in vivo and in vitro, they show that most neurons in the shell region of the inferior colliculus receive several monosynaptic cortical inputs. In vitro, these descending synapses show sublinear summation with a major tonic component for prolonged stimuli. Both in vivo and in vivo experiments support the idea that descending cortical inputs and ascending inputs from the central inferior colliculus temporally overlap and both activate NMDA and non-NMDA receptors. This cooperativity of inputs leads to supra-linear summation and boosting of the response.

Strengths:<br> • The manuscript provides a first detailed analysis of a loop between the cortex and midbrain. It elegantly combines in vivo and in vitro electrophysiological techniques to study this network on a cellular/synaptic level.<br> • These experiments thoroughly characterize the nature of cortical and midbrain excitatory inputs onto shell IC neurons and elucidate how they integrate the ascending and descending inputs on a cellular level.

Weaknesses:<br> • A major weakness of this study is that they do not directly show that ascending and descending inputs to the IC shell neurons actually coincide, but only imply that this should be the case, considering different latency measurements. Latencies that are measured in the anesthetized preparation may change in the awake behaving animals which may change the timing of the respective inputs. In addition, the authors do not show to what extent coincidence of ascending and descending inputs to shell IC neurons is maintained for longer and more complex sounds as compared to click stimuli.<br> • The manuscript does not address the question of whether the different neuron types that they encounter in the shell region based on the firing pattern to current injections, vary in their input latencies, their number and distribution of NMDA receptors or their integrative properties. This may have some additional effect on how these neurons process ascending and descending information.<br> • The authors have not demonstrated that silencing of descending inputs from the AC affects IC shell activity.

Reviewer #3 (Public Review):

Overall, this manuscript is generally nicely written and well-illustrated. I don´t really have any major issues. I like the manuscript but I have a few comments and some issues that need to be addressed.

My main concern is that the authors claim several times that the projections to the central nucleus of IC are weak and they neglect their potential functional role. I think this is a little bit unfortunate. It is true that the large AC projection primarily targets the cortical regions or shell of IC, but it is beyond doubt that it also targets the central nucleus (e.g. Saldaña's studies) . We cannot know whether it is a weak projection or not without central nucleus recordings. Admittedly, these experiments would be challenging, so I would ask the authors to tone down a bit these comments throughout the ms. Also, the reason for the 'weak' projection to the central nucleus may be due to the size and location of the injections made in the auditory cortex. Thus, I would like to see the injections site of Chronos if possible. Likewise, fig 1B is too small and of low quality (at least in my pdf file for review) to appreciate details of labeling. I would suggest that the authors make a separate figure showing the injection site in the AC and larger and clearer labeling in the IC.<br> Also I wonder if the title of the manuscript should refer to the non-lemniscal IC as most of the data is related to this area.<br> While the dogma is that the descending projections are glutamatergic, the authors may care to consider a recently published paper<br> https://www.frontiersin.org/articles/10.3389/fncir.2021.714780/full, which challenges this view by showing that inhibitory long-range VIP-GABAergic neurons target the IC. It would be interesting if the authors could comment on how this projection may have influenced the results of the present study.

Reviewer #1 (Public Review): 

Lee et al. identify miR-20b as a molecular regulator of hepatic lipid metabolism through the post-transcriptional regulation of the nuclear receptor PPAR alpha. Through mechanistic studies the authors identified the 3'UTR of PPARa as a direct target for miR-20b regulation of expression. The experiments are well controlled and the study provides deep mechanistic insight into the miR-20b/PPARa circuit in modulating hepatic lipid metabolism. Furthermore, the authors provide evidence that targeting the miR-20b pathway to enhance PPARa activation via synthetic ligand fenofibrate. The studies provide much needed mechanistic insight into molecular regulators of hepatic lipid metabolism in response to nutrient stress such as high fat diet. While this is a detailed and thorough assessment of this pathway, there are several issues that were identified in the review of this article outlined below: 

1) The authors state there is no off target expression of miR-20b in adipose tissue in their over expression experiments. However, per figure 4 supplement 1, EpiWAT has increased expression over controls in HFD fed conditions. Furthermore, figure 4 supplement 2 shows a functional difference in EpiWAT weight in HFD where miR-20b treated mice have higher fat weight. The authors need at the least to discuss the potential role of adipose tissue in promoting their observed phenotype. 

2) Figure 5 shows anti-miR-20b essentially restores PPARa expression. However, the rescue effects in terms of body weight, liver triglycerides and liver damage are only modestly improved. The authors need to discuss this modest effect and potentially offer alternative mechanisms aside from PPARa as the physiological target. 

3) The authors performed experiments with mutated 3'UTR of PPARa and show mutated PPARa is refractory to regulation by miR-20b. However, the authors provide no functional evidence that mutating the 3'UTR of PPARa elicits changes in hepatic lipid metabolism. Discussion of this point is needed at the minimal.

Reviewer #2 (Public Review): 

1) In the experiments depicted in Figures 1D and E, did OA treatment of HepG2 and/or Huh-7 cells produce a reduction in the levels of mRNA encoding PPARalpha (or PPARalpha protein levels) in concordance with the shown rise in mRNA for miR-20b? 

2) Moreover, Figure 1 shows a fuller landscape of the transcriptional impact of microRNAs in context of obese livers in mice and human. Given this, what made miR20-b more interesting than, for example, miR106a, miR-17, or others that also appear to be robustly regulated? Why focus on miR20b? 

3) What does the rank and p-value exactly represent in tabular part of Figure 1A? This is very unclear as shown, including the figure legend. 

4) Figure 1, supplement 1 shows characteristics of patients involved in data for Figure 1, etc. This shows that the normal patients are younger than the other two groups, the M-F ratio is not identical (more female in the normal group), and the total cholesterol levels are not well matched either. What other parameters are available? Hemoglobin A1c? Fasting glucose? In the end, we need to know that the groups, apart from the severity of NAFLD and NASH, were well matched. Given the small size of each group (n = only 4-5, this matching is critical to avoid confounding of the relationship between miR-20b, PPARalpha, and NAFLD/NASH progression. 

5) The title of Figure 2 relates to PPARalpha. However, in Figure 2G, it is clear that several NRs are downregulated by miR20b overexpression in cells. Although the paper focuses on PPARalpha, should the authors not explore at least some of the other hits to ensure that the impact of PPARalpha is of particular importance vs. others? 

6) In Figure 3, the data show, presumably, that OA induces miR20b, which then represses PPARalpha and, in turn, CD36 downstream of PPARalpha. If this is the case, then how does OA continue to get into the cells? Once CD36 expression falls dramatically, doesn't the key OA uptake mechanism fall with it? Then, does the induction of miR20b abate? Or, does FATP6 or another uptake mechanism account for OA entry into these cells? 

7) Similarly, what happens to AGPAT, GPAT, and DGAT expression in context of OA treatment and modulation of miR20b? Does the capacity of the cell to store OA in the form of triglyceride inside of lipid droplets change, so that the amount of free OA or oleyl-CoA inside the cell rises? Could this impact the transcriptional phenotype? 

8) In Figure 3P, would the impact of anti-miR on the effect of OA on FASN be lost in PPARalpha KO cells? This would really test the functional relevance of the purported transcriptional hierarchy. 

9) The authors should really at least perform a bulk RNAseq analysis to confirm the similarity of the effect of miR20b or anti-miR seen in cells, at the mouse or human liver tissue level. As it is, they only look at 3 FAOX genes, 2 FA uptake associated genes, and 2 FA synthesis genes. This is not very comprehensive as a validation of the in vitro data, although it is intriguing. Or, at the very least, look at a large validated set of PPARalpha target genes in vivo. 

10) Notably, the figures in general do NOT show individual data points. This is the standard for visual display, rather than bar graphs with simple SEM bars. 

11) The in vivo data (e.g. Figure 4) are very low n values. Augmenting this would add confidence to the data. As an example, of inconsistencies potentially stemming from very low n, the liver weights (Figure 4F) are not very different across groups, although the triglyceride levels in the livers (Figure 4H) are more than twice as high. The images of liver specimens shown as examples (Figure 4F) are also more dramatic than the weights would indicate. Note also that the body weights of the mice (Figure 4C) are different as well, and this alone could explain the livers being modestly heavier. Indeed, the extent of body weight excess mirrors the extent of liver weight excess, suggesting that the entire animal may be larger across multiple metabolic tissues including adipose. This is proven in Figure 4D, where the fat mass looks to be larger as well. To this end, Figure 4 supplement 2 shows multiple tissue weights to be increased in this model, suggesting that specificity for hepatic steatosis may be low. 

12) In figure 5 S1, the anti-miR20b substantially reduces the weights of multiple tissues in mice fed a HFD, given this, why does overall body weight (figure 5c) show such a modest difference. Figure 5 E and F also suggest that the overall weights would have been lower than shown in Figure 5C. In the end, instead of bar graphs of the final weights, the entire weight curve for the mice fed the HFD should have been shown. 

13) How well were the NAFLD vs. normal GSE individuals matched? This is very important, since PPARalpha emerges from comparing these data sets. Matching is very important to make sure that the differences in NR expression does not stem from a confound that went along win parallel with the NAFLD cohort vs. the normal GSE cohort.

Reviewer #3 (Public Review): 

In this manuscript, Le et al. use an elegant combination of cultured cells, patient samples, and mouse models to show that miR-20b promotes non-alcoholic fatty liver disease (NAFLD) by suppressing PPAR-alpha. The authors show that miR-20b inhibits PPAR-gamma expression, resulting in reduced fatty acid oxidation, decreased mitochondrial biogenesis, and increased hepatocyte lipid accumulation both in vitro and in vivo. Inhibition of miR-20b in mouse NAFLD models leads to increased PPAR-gamma, reduced hepatic lipid accumulation, decreased inflammation, and improved glucose tolerance. Overall, the data are well-controlled and support the authors' conclusions. 

Strengths:

1) In Figure 1, the authors show miR-20b is increased in NAFLD patients, mouse obesity/NAFLD models, and cultured liver cancer cells treated with oleic acid (OA). The use of multiple complementary approaches is very powerful, although more information regarding the diagnoses in the 13 patient samples would be helpful (see below). 

2) In Figure 2, the authors show that PPAR-alpha is a direct target of miR-20b. These data include a luciferase reporter assay regulated by the 3'UTR of PPAR-alpha. Importantly, when the 3'UTR is mutated, suppression of luciferase expression by miR-20b is no longer observed. The authors use multiple different algorithms to predict miR-20b targets, look for overlap, and then confirm PPAR-alpha as the most important "hit" in vitro. 

3) Figure 3 highlights changes in fatty acid metabolism in HepG2 cells transfected with miR-20b, miR-NC, or anti-miR-20b and treated with oleic acid. Figure 3, supplement 4 shows that anti-miR-20b can alleviate OA-induced hepatic steatosis in both HepG2 cells and primary hepatocytes. The use of another (primary) cell line here is important, because HepG2 is a liver cancer cell line, and metabolic changes in HepG2 cells might not be representative of non-neoplastic hepatocytes. 

4) In Figure 4, the authors show that miR-20b promotes hepatic steatosis, increases liver weight, increases liver injury markers, and impairs glucose tolerance and insulin sensitivity in HFD-fed mice. Conversely, anti-miR-20b inhibits hepatic steatosis, decreases liver weight and liver injury markers, and improves glucose tolerance and insulin sensitivity in HFD-fed mice (Figure 5). Anti-miR-20b also inhibits hepatic steatosis and fibrosis and decreases liver injury markers in MCD-fed mice (Figure 8). These in vivo studies provide excellent support for the authors' hypothesis regarding the role of miR-20b in promoting fatty liver disease. The liver readily takes up small nucleic acids, including miRs and anti-miRs. Thus, the possibility of using anti-miR-20b as a therapeutic for fatty liver disease is intriguing, and supported by these experiments. 

5) In Figure 6, in HepG2 cells, the authors demonstrate that PPAR-alpha overexpression (or to a lesser extent fenofibrate treatment) is able to rescue the transcriptional effects of miR-20b overexpression. Conversely, siPPAR-alpha can rescue the transcriptional effects of anti-miR-20b. Similar results are shown in Figure 7-fenofibrate is able to at least partially suppress some of the metabolic phenotypes that are exacerbated by miR-20b overexpression in HFD-fed mice (the decreased lean/BW ratio, elevated fasting glucose, some transcriptional changes). Again, it is nice to see that the in vitro data is supported by in vivo results. 

Weaknesses:

1) In Figure 3, figure supplement 2, it seems the effects of miR-20b overexpression in primary hepatocytes may be a bit overstated. While it does seem that miR-20b enhances the accumulation of fat in primary hepatocytes upon OA treatment, miR-20b overexpression alone does not seem to have significant effects on steatosis (A), cholesterol (B), or triglycerides (C). 

2) Histologic analysis of mouse liver samples by a pathologist is lacking. In Figure 4, is there increased inflammation and/or fibrosis with miR-20b overexpression, or just increased steatosis? In Figure 4 and Figure 8, it would be helpful if steatosis, fibrosis, and inflammation were quantified/scored histologically. 

3) The effects of anti-miR-20b on hepatic triglycerides and inflammatory markers in vivo are modest (Figures 5 and 8). Perhaps an enhancement could be seen by combining anti-miR-20b with fenofibrate. While the authors show that fenofibrate's effects are suppressed with miR-20b overexpression, they don't examine what happens when fenofibrate is combined with anti-miR-20b. To me, this experiment is critical to determine if PPAR-alpha activity could be further maximized to combat NAFLD (beyond what is seen with fenofibrate alone).

Reviewer #1 (Public Review): 

Regression models are a widespread statistical technique used in epidemiological studies. Most commonly used regression models do not explicitly parameterize the relationship between the independent variables and the variance or skewness of the dependent variable. Generalized Additive Models for Location, Scale and Shape (GAMLSS) is a regression technique that provides the flexibility to estimate parameters of the dependent variable distribution (mean, median, variance, e.t.c) as a function of independent variables. This manuscript uses data from the 1970 British birth cohort study to showcase the use of GAMLSS in epidemiological studies and further compares the results to quantile regressions. 

The primary concern with this manuscript is its overall goal. In its current form, it is hard to assess whether the manuscript is meant to be a tutorial on how to fit GAMLSS and interpret its output from an epidemiological context, or it is meant to be a research report investigating the association between three risk factors (sex, social class, physical activity) with two outcomes (BMI and mental wellbeing). 

The modelling choices in the manuscript are only suited if it is aimed to be a tutorial. For example, the rationale for the choice of the outcomes (BMI and mental wellbeing) is reported to be the fact that they are often measured on a continuous scale. Similarly, authors only interpret the unadjusted estimates because they were similar to those from an adjusted model. Although these are acceptable choices for a tutorial, if the manuscript's goal was to estimate the true association between these variables, it has several shortcomings. Such as i) the disadvantages of dichotomizing a continuous independent variable are well known(1); ii) it is recommended to choose potential confounders based on a Directed Acyclic Graph (DAG) to ensure the estimates are unbiased(2); iii) a clear rationale for estimating this effect and what is already known in the literature about the association should be mentioned in the introduction. 

Yet, interpretations provided in the results section and parts of the discussion imply that they are to be taken as estimates of a true association. For example, i) estimates for variable sex is contrasted with that of social class (Page 7 Line 204), ii) argument comparing results of previous studies on BMI and use of a national representative sample (Page 9 line 252 to 258), iii) using GAMLSS and British Birth Cohort data are reported as strengths of the manuscript (Page 10 Line 309), iv) arguments about limitations to make causal claims for the estimates and other data complexities (Page 11 Line 319 to 334). 

The data generating process in epidemiological studies, especially in observational designs, is complex and needs to be taken into consideration when conducting statistical analysis. Statistical models are often oversimplified mathematical representations of this real-world data generating process. Often in practice, this simplification (e.g. mean the only model) and strong assumptions (e.g. homoscedasticity) are chosen to aid in estimating quantities that are easy to interpret and that have high clinical or public health utility. The popularity of logistic regression in epidemiology, compared to other fields, is a clear example of this practice. On the same lines, complex models should not be adopted at the expense of interpretability or utility of the model outputs. Users of such complex models should clearly explain the interpretation of the parameter estimates and should provide clinical and/or public health utility of the same to avoid misinterpretation of the outputs by potential future users of the model and by policymakers. For example, GAMLSS provides considerable flexibility in modelling choices compared to more standard techniques (e.g. GLMs, and GAMs). Tutorials clearly explaining how to fit GAMLSS, interpretation of its output along its utility from an epidemiological context are needed. There are some shortcomings to the current manuscript if assessed from a perspective of a tutorial: i) it would be pertinent to provide a comparison to linear regression, which only models the mean of the outcome, and elaborate how and why the more complex model help interprets the observed relationship or lack of it; ii) no clear lay interpretation of the effect measures on SD, Coefficient of Variance, and Skewness; iii) guidelines of choosing outcome distribution type (e.g., Normal distribution vs Box-Cox Cole and Green family) from an epidemiological context.

Reviewer #2 (Public Review): 

The authors demonstrated the utility of GAMLSS to investigate the association of risk factors and outcomes in variability as well as central tendency. They used BMI, mental wellbeing as outcomes and three risk factors (sex, socioeconomic circumstances, and physical inactivity) in this study. The strength of this study is that they successfully demonstrated the utility of the approach using a large empirical data set. The limitation of the study is that the data is from observational study, thus causal inference was not feasible. 

Most of clinical studies have been focused on the difference in mean rather than other characteristics of distributions such as variance. However, recent studies have demonstrated that intervention effect is heterogeneous (some are benefitted from the intervention, but others are not). From the perspective of developing personalized medicine, it is important to know whether interventions have response heterogeneity as the first step. If such heterogeneity is identified, the next step will be to identify the factors associated with the heterogeneity (or those who will be benefitted from the intervention). Therefore, this study contributes to the first step by investigating the possibility of response heterogeneity.

Reviewer #3 (Public Review): 

The manuscript by Bann and Cole assesses how risk factors such as sex, socioeconomic status, and physical inactivity affect outcome variability and skewness of outcomes such as BMI and mental well-being. The authors also explored how these risk factors affect the skewness of the distributions of these outcomes. The GAMLSS distribution was used to determine how these risk factors influence both the mean and variability of the outcomes. Additionally, the authors also investigated how these risk factors influence quantile function of the outcomes.

The authors provide an important contribution to epidemiologic studies by examining how risk factors influence the variability and skewness of outcomes. Their discussions provide topics for other researchers to consider when examining intervention effects. While most current approaches often focus on discussions of how interventions influence the mean functions, the authors provide adequate discussions on the importance of also considering how interventions also influence the variances. In addition to providing empirical evidence for why more authors should consider how risk factors influence the variability and not just the mean of outcomes, the authors provide strong justifications for such considerations with the use of important large cohort data.

Reviewer #1 (Public Review): 

How the brain is organized to represent various concepts has long been a central cognitive neuroscience research topic. Most notably, several areas in occipito-temporal cortex are known to be specialized in representing faces. However, as a type of complex stimuli, faces contain multi-dimension features. By using high-field 7T fMRI, the present paper aimed to examine how face-selective brain areas might be organized to represent face features/face parts in fine-scaled spatial patterns. 

This is an overall well-executed study with a careful experimental design that includes helpful control conditions, although the external validity might be somewhat limited due to the small sample size of the present study. Nonetheless, the within-subject results appear to be reliable and the results support their conclusions. The findings complement non-human primate studies on face-selective patches. In addition, there are some agreements but also discrepancies between the present findings and a previously proposed "faciotopy" hypothesis, i.e., the spatial organization of patches selective to different face parts would reflect the actual distances between these face parts.