On 2022-10-11 18:59:18, user Willow K. Coyote wrote:

General assessment:<br /> D’Costa, Hinds et al. develop a framework to infer protein fitness based on laboratory evolution data using DHFR as a test case. This framework is based on a generalized Pott’s model, which infers parameters from successive rounds of selections using sequencing data. This model works well to predict fitness and epistasis within the experimental data’s bounds but cannot predict beyond it. Overall, we enjoyed the exploration of the model and the data, we had a difficult time understanding what major insights were gleaned from the work, and we found the manuscript was written for perhaps a bit too specialized of an audience. From our perspective, the manuscript could be greatly improved by focusing on the novelty of the models and including additional context for more general audiences.

Strengths:<br /> 1)The manuscript is a rigorous and model driven exploration of how data can be used for training <br /> 2) The authors transparently present the performance of there models even when it does not work as planned.

Weaknesses:<br /> 1) It was difficult to identify what the key takeaways of the paper were. Perhaps working on the narrative would make it easier to identify these takeaways.

2) From our reading we found the manuscript was written at an expert level. We found this made it difficult to interpret some of the results.

High-level Recommendations:

1) The quality of a high throughput assay is largely determined based on the quality of the input library. However, the authors do not show the level of coverage gleaned from the library that could result in some parts of the gene not being well-represented within the model. We would love to see more discussion of the raw sequencing data and its analysis to get a sense of the overall quality/coverage of the DHFR libraries and the signal-to-noise within the selection.

2) From our reading the overall framing of the manuscript appeared to be for a very technical audience. As readers we are very familiar with genotype-phenotype assays and to a limited degree models to represent the data from this assays. However, as we are not deeply familiar with Potts models and their novelty in the context of fitness landscapes, it was difficult for us to understand portions of the manuscript. In addition as using Pott’s model to represent the directed evolution experiments appeared to be a major novelty for the manuscript it would be useful to provide more background for why this specifically is novel.

3) From our reading the interpretation and exploration of the model was focused on describing the smooth or ruggedness of the fitness landscape and testing whether one could extrapolate from this fitness landscape to future predicted generations of sequences. We found it admirable the authors shared that the extrapolation did not work. Fot the fitness landscape exploration more background would likely benefit a non-expert reader. For example, the fitness landscape was described as being ‘Fuji-like’. For those familiar with fitness landscapes, this means a smooth landscape with a single peak. However, for those unfamiliar, this could be somewhat confusing. Furthermore, perhaps it would be worth exploring what we can learn about fitness landscapes with the model, as this could yield further insight.

Specific discussions of results and figures:<br /> Figure 1: Shows the directed evolution experiment using error-prone PCR on DHFR to search the fitness landscape for diverse sequences encoding DHFR activity. A: In the text the authors could explain why they stopped the experiment at round 15, but not earlier or later. B: This panel shows clearly that round 15 is distant from other rounds. Within D the authors show a plot of what fraction of mutations are accumulated throughout the experiment at a given position. A high score here likely implies that there is a beneficial impact of mutating a residue at a position. In the text position, 20 is discussed widely but is not noted on the figure while the figure notes positions 71 and 117 are noted but not discussed until later during the interactions discussion mostly focused on Figure 2. We found this somewhat confusing and it would perhaps be beneficial to note where specifically residue 20 is (presumably within the greyed out ‘Nucleotide Binding’ part of the plot). Perhaps more discussion of why mechanistically N20D is beneficial would be useful for aiding in interpreting the result.

Figure 2:<br /> A: Shows a heatmap of how the model predicts fitness for mutations at each position. Error-prone PCR was used to make libraries for this experiment as in classic directed evolution experiments and is well known to not result in full coverage. Within the model it appears that most positions have a predicted fitness with many being 0 (based on the greyish color). We would like to know the percentage of variants present for the predictions. As illustrated later in the manuscript, the model does not perform well when extending beyond the data. Perhaps if positions with low confidence or without frequent observed mutations at the beginning of the experiment were withheld the performance of the model would improve.Within the text the authors mention how the heatmap highlights the importance of the catalytic residues but as presented it is difficult to see this as we did not know where these are within the heatmap. It would be useful to have annotations for where the catalytic/functional residues are. It appears that the N20D that is enriched with the library is not positive within the model. Why could that be? <br /> B: The authors mapped the average predicted mutational effect onto the structure of DHFR. The discussion of this data is rather limited to negative impacts within the core. Perhaps it would be useful to talk more about what can be learned or compare to previous library studies of DHFR such as recent work done in Kimberly Reynold’s lab (McCormick, J., 2021 eLife).<br /> C-D: The authors then explore interactions within the model as a way to look at epistasis by looking at interactions across residues. Within D the authors show specific examples mapped onto the structure focusing on either side of the NADPH binding site. It would be useful within C to highlight interactions that appear important and highlighted in D. <br /> E-F: The authors compare how well the model can be used to identify contacting residues based on interactions scores. The authors show that including the 15th round improves model performance. For many interactions contacts are incorrectly predicted, however within the text the authors discuss how the model performs well compares to other approaches. For many regions of the contact map especially the end of the gene there are few predictions. Why is that? Overall, we find this exploration of the models to be difficult to interpret as from our view it appears the model is frequently predicting incorrect contacts or no contacts at all. Perhaps this is expected and that should likely be discussed further within the text.

Figure 3: Using their model trained on experimental data, they simulated several evolutionary trajectories of DHFR until a fitness peak was reached. Each simulation converged on the same peak, relatively close to the WT sequence. Following this, simulations were performed starting from the last experimental time point; however, mutants after this point were found to be inactive. In the discussion of these results, the authors note several reasons for this shortcoming of their model, all of which seem reasonable and highlight the need for more data to effectively use such models. Key suggestions include a discussion of how/where the inactivating mutations occur in the enzyme. Are these concentrated in specific hotspots, or are they in areas well sampled by the preceding libraries?

Suggested citation: “Deciphering protein evolution and fitness landscapes with latent space models”, Xinqiang Ding, Zhengting Zou & Charles L. Brooks III, Nature Communications 2019

Format of review: We reviewed this paper as part of our weekly paper discussion. We read a paper in depth weekly and discuss it as a group. In contrast to most journal clubs, we do not make slides, and everyone is an active participant in presenting a paper in which we typically each will describe each figure within the paper. As we are a group with diverse backgrounds, the person with the most experience in that area will bring everyone up to speed if anything is confusing. We focus on going through a paper from figure to figure. For these reasons, most of our comments and suggestions are about the figures. Beyond that, we try to understand the major points a paper makes and whether the data presented in the figures supports it. Different people wrote a description for each figure they wrote.

Authors of Review: <br /> Willow Coyote-Maestas<br /> Matthew Howard<br /> Patrick Rockefeller Grimes<br /> Christian Macdonald<br /> Donovan Trinidad<br /> Arthur Melo

Relevant expertise: Scientifically, we come from many different backgrounds but most people within our group have experience with membrane proteins, high throughput genotype-phenotype assays, and developing assays for measuring how mutations break membrane protein.

On 2022-10-10 09:12:26, user Srijit Seal wrote:

Here's a link to the STRING gene-gene interaction networks for 35 Genes annotations associated with 38 assays in the public dataset (Figure 4)

https://version-11-5.string...

On 2022-10-10 08:53:23, user Pasquale Saldarelli wrote:

the excel files cited in the supplementary material are not present

On 2022-10-09 23:27:46, user jiarong wrote:

I am glad to see to see updated manuscript w/ the contamination in the negative set (microbial + plasmid) removed in the simulated Refseq data, addressing my previous comments (http://disq.us/p/2it17sx) "http://disq.us/p/2it17sx)"). I have a few more comments:<br /> - I am concerned that there are unknown prophages in the microbes in the mock community data that could significantly skew the precision lower. The contaminant screening that is used in the simulated Refseq data might help here too.<br /> - From my experience on VirSorter2, the optimal score cutoff for highest F1 could change a lot with different datasets such as environment types, eg. soil samples generally requires higher cutoff. Thus this SOP (https://www.protocols.io/vi... is the recommended way to run VirSorter2.

On 2022-10-09 16:47:55, user Jordan Willis wrote:

Hello,

Any software that bridges the gap between computationalist and experimentalist is great. However, the primary use case for this will be remote servers. People just don't have the laptops required to run the cellranger software. Even if they did, once you throw docker into the mix, you ask the experimentalist end user to know quite a bit about the command line. If they know enough to use docker, they would probably be able to use a lot of the command line apps available for this. With that in mind, I'd like to see way more documentation on the remote setup. Especially the incorporation of the ShinyProxy. I want to use this for folks on my team but I'm not especially clear on how this app works with the proxy.

My other concern is the closed source code baked into the docker image. I'd encourage you to release it.

Also, the youtube documentation while great, also should be posted as static documentation.

Thanks,<br /> Jordan

On 2022-10-08 20:28:21, user David Bacsik wrote:

All data used in this study is now available in GEO under accession number GSE214938 at https://www.ncbi.nlm.nih.go....

On 2022-10-08 16:37:04, user Michael Ailion wrote:

This paper aims to understand how toxin-antidote (TA)<br /> elements are spread and maintained in species, especially in species where<br /> outcrossing is infrequent and the selfish gene drive of TA elements is limited.<br /> The paper focuses on the possible fitness costs and benefits of the peel-1/zeel-1 element in the nematode C. elegans. A combination of mathematical modeling and experimental tests of<br /> fitness are presented. The authors make a surprising finding: the toxin gene peel-1<br /> provides a fitness advantage to the host. This is a very interesting<br /> finding that challenges how we think about selfish genetic elements,<br /> demonstrating that they may not be wholly “selfish” in order to spread in a<br /> population.

This paper is of interest to evolutionary biologists and<br /> population geneticists. It provides empirical evidence that supports a previous<br /> hypothesis of how selfish toxin-antidote elements spread in non-obligate<br /> outcrossing species. While the experiments and data are appropriate for<br /> addressing this hypothesis, one major conclusion is not supported by the data<br /> and one other major conclusion is supported only weakly.

Strengths

1. The authors support results found with a zeel-1 peel-1 introgressed strain by using<br /> CRISPR/Cas9 genetic engineering to precisely knock-out the genes of interest.<br /> They were careful to ensure the loss-of-function of these generated alleles by<br /> using genetic crosses.

2. Similarly, the authors are careful with<br /> controls, ensuring that genetic markers used in the fitness assays did not<br /> affect the fitness of the strain. This ensures that the genes of interest are causative<br /> for any source of fitness differences between strains, therefore making the<br /> data reliable and easily interpretable.

3. A powerful assay for directly measuring the<br /> relative fitness of two strains is used.

4. The authors support relative fitness data<br /> with direct measurements of fitness proximal traits such as body size (a proxy<br /> for growth rate) and fecundity, providing further support for the conclusion<br /> that peel-1 increases fitness.

Weaknesses

1. One major conclusion is that peel-1 increases<br /> fitness independent of zeel-1, but this claim is not well supported by<br /> the data. The data presented show that the presence of zeel-1 does<br /> not provide a fitness benefit to a peel-1(null) worm. But the experiment<br /> does not test whether zeel-1<br /> is required for the increased<br /> fitness conferred by the presence of peel-1.<br /> Ideally, one would test whether a zeel-1(null);peel-1(+) strain is<br /> as fit as a zeel-1(+);peel-1(+) strain, but this experiment may<br /> be infeasible since a zeel-1(null);peel-1(+) strain is inviable.

2. The CRISPR-generated peel-1<br /> allele in the N2 background only accounts for 32% of the fitness difference<br /> of the introgressed strain. Thus, the effect of peel-1 alone on fitness appears to be rather small. Additionally, this<br /> effect of peel-1 shows only weak<br /> statistical significance (and see point 5 below). Given that this is the key<br /> experiment in the paper, the major conclusion of the paper that the presence of<br /> peel-1 provides a fitness benefit is<br /> supported only weakly. For example, it is possible that other mutations caused<br /> by off-target effects of CRISPR in this strain may contribute to its decreased<br /> fitness. It would be valuable to point out the caveats to this conclusion, or<br /> back it up more strongly with additional experiments such as rescuing the peel-1(null) fitness defect with a<br /> wild-type peel-1 allele or determining<br /> if introduction of wild-type peel-1 into<br /> the introgressed strain is sufficient to confer a fitness benefit.

3. The strain that introgresses the zeel-1 peel-1 region from CB4856 into the N2 background was made by<br /> a different lab. Given that N2 strains from different labs can vary<br /> considerably, it is unclear whether this introgressed strain is indeed isogenic<br /> to the N2 strain it is competed against, or whether other background mutations<br /> outside the introgressed region may contribute to the observed<br /> fitness differences.

4. Though the CRISPR-generated null allele of peel-1 only accounts for 32% of the<br /> fitness difference of the zeel-1 peel-1 introgressed<br /> strain, these two strains have very similar fecundity and growth rates. Thus,<br /> it is unclear why this mutant does not more fully account for the fitness<br /> differences.

5. Improper statistical tests are used. All comparisons use<br /> a t test, but this test is inappropriate when multiple comparisons are made.<br /> Importantly, correction for multiple comparisons may decrease the already weak<br /> statistical significance of the fitness costs of the peel-1 CRISPR allele (Fig 3E), which is the key result in the<br /> paper.

6. N2 fecundity and growth rate measurements<br /> from Fig 2B&C are reused in Fig 3C&D. This should be explicitly stated.<br /> It should also be stated whether all three strains (N2, the zeel-1 peel-1 introgressed strain, and<br /> the peel-1 CRISPR mutant) were<br /> assayed in parallel as they should be. If so, a statistical test that corrects<br /> for multiple comparisons should also be used.

7. It appears that the same data for the<br /> controls for the fitness experiments (i.e. N2 vs. marker & N2 vs.<br /> introgressed npr-1; glb-5) may be<br /> reused in Fig 2A and 3E. If so, this should be stated. It should also be stated<br /> whether all the experiments in these panels were performed in parallel. If so,<br /> this may affect the statistical significance when correcting for multiple<br /> comparisons.

Minor<br /> points

1. Though the mathematical modeling is interesting from a<br /> theoretical point of view, we feel that it oversells the rationale behind the<br /> experiments, setting up a “straw man” argument to knock down. Also, the modeling<br /> relies on rather high assumptions of the possible carrying cost of peel-1/zeel-1. For example, the modeling<br /> of the effect of outcrossing rate on peel-1/zeel-1<br /> frequency assumes a selection coefficient of 0.35, which seems rather<br /> arbitrary and high. Where does this number come from? Is there any precedence<br /> for this high carrying cost? In our opinion, the idea that energy expenditure<br /> or leaky toxicity accounts for such a high carrying cost seems unlikely.

2. The two studies cited for “outcrossing rates typical for<br /> C. elegans” estimated vastly different outcrossing rates (~20% or ~1%).<br /> The model presented in Fig S1 specifically uses the lower estimates (0-2%), so<br /> the Sivasundar & Hey paper is miscited here. It is unclear whether there is<br /> a good rationale to go with the lower rate estimates.

3. The measurement of body-size is unclear in the main<br /> text. Only when reading methods did we realize that body-size is more of a<br /> proxy for growth rate rather than an end-point measurement of worm size.

4. What is the temporal distribution of egg laying of the<br /> N2 and N2peel-1(null) strains? Based on how the<br /> data collection is described in the Methods, the authors should already have<br /> these data. Does egg-laying start at the same time in the two strains? The fact<br /> that strains carrying peel-1 grow<br /> faster but also apparently produce more sperm (which might slow them down)<br /> makes an analysis of this worthwhile, especially since fitness depends on when<br /> eggs are laid, not just how many. Some more characterization of this fitness<br /> trait seems appropriate and useful for beginning to understand how peel-1<br /> may be increasing fitness. Given that the number of sperm limits how many eggs<br /> are laid, the presence of peel-1 apparently results in more sperm. It is<br /> surprising that a gene exclusively expressed in developing sperm can lead to<br /> production of more sperm.

5. Line 65: the statement “similar elements have not been<br /> identified in obligate outcrossing Caenorhabditis nematodes” is somewhat<br /> misleading. TA elements may not have been identified in obligate outcrossing<br /> nematodes because of research bias since genetic experiments are easier to<br /> perform in non-obligate outcrossers and it is unclear that there have been<br /> extensive searches for TA elements in outcrossing nematodes. Furthermore, as<br /> the mathematical models in this study suggest, TA elements will spread quickly<br /> with increasing rate of outcrossing. Since a TA element’s non-fixation within a<br /> species has historically been a prerequisite for its discovery, the rapid TA<br /> element fixation that would generally occur in obligate outcrossers would make<br /> their identification more challenging.

6. Line 209-210: it is stated that this is the “first<br /> measurement of the fitness cost of a TA element to the host” and “first<br /> demonstration that a TA element can benefit the organism.” These claims may be<br /> overstated. It has been previously shown in several cases that TA elements can<br /> provide fitness benefits in bacteria, such as improved antibiotic resistance<br /> (e.g. Bogati et al. 2022, PMID: 34570627).

7. More details about the CRISPR protocol would be helpful.<br /> It is unclear whether Cas9/sgRNAs were introduced as RNPs or plasmids (and at<br /> what concentrations). It is unclear how worms were screened for edits. It is<br /> also unclear how many Dpy or Rol worms were screened and how many peel-1 or<br /> zeel-1 edited worms were found (the efficiency of CRISPR). The meaning<br /> of the shaded portion of the repairing oligo sequences in the table is not<br /> explained. Finally, it is not stated whether CRISPR-generated mutant strains<br /> were outcrossed.

Reviewed (and signed)<br /> by Lews Caro and Michael Ailion

On 2022-10-07 16:32:59, user Alexa Sterling wrote:

The published version of this article is available here: https://doi.org/10.1002/lno... . It is open access. Thank you!

On 2022-10-07 09:11:44, user Iratxe Puebla wrote:

Review coordinated via ASAPbio’s crowd preprint review

This review reflects comments and contributions by Ruchika Bajaj and Sanjay Kumar Sukumar. Review synthesized by Iratxe Puebla.

The preprint studied the conformational changes upon binding of the Akt protein kinase to the Akt active site inhibitor A-443654 and the Akt allosteric inhibitor MK-2206, under three states of Akt: inactive monophosphorylated, partially active tris-phosphorylated, and fully activated, tris-phosphorylated bound to PIP3 membranes. The MK-2206 resulted in allosteric conformational changes in all states and restricted membrane binding through sequestration of the PH domain. The A-443654 inhibitor led to allosteric conformational changes in the monophosphorylated and phosphorylated states, with increased protection in the PH domain upon membrane binding. The results can assist the design of Akt-targeted therapeutics.

The reviewers had a few minor comments about the paper:

It could be helpful to include a short explanation early in the text about the use of HDX-MS, how it works and why it is useful for exploring conformational changes.

Figure 2A+B provide a nice representation of the HDX exchange data.

Results ‘3 seconds at 1°C, which is referred to as 0.3 sec in all graphs and the source data)’ - This may be a bit confusing for someone who wants to look at the data in the figures independently. Consider an alternative way of representation or providing some further clarification in the figure legend.

Results ‘Decreases in exchange in the kinase domain were similar to those observed in the absence of membranes, occurring in regions encompassing the αC helix, the ATP binding pocket, as well as changes covering the activation loop and C-lobe:PH interface’ - Please clarify whether the comparison here relates to the data in Fig. 3A/C vs Fig 4A/C.

Results ‘There were multiple regions of significantly decreased deuterium exchange in the kinase and PH domains (Fig. 2B, 2D, 2E).’ - This section mainly focuses on conformational change upon the addition of MK-2206 allosteric inhibitor binding. Figure 2F appears to be the most relevant for the comparison. It is suggested to provide additional combination of data with ATP analogs to understand the coordination of ATP and inhibitor during the inhibition step in the cycle.

Results ‘Both experiments were carried out under saturating concentrations of inhibitor binding, so this difference reflects intrinsic conformational differences.’ - Saturating concentrations implies that most of the population will be in the same conformation. Please comment on the association between saturating concentrations and intrinsic conformational differences.

Discussion - There do not appear to be many structures available for different conformational states of Akt. The study has mapped hdx data on available structures, however, it'd be good to see correlation of conformational changes by HDX with conformational changes in structure.

On 2022-10-07 09:04:58, user Iratxe Puebla wrote:

Review coordinated via ASAPbio’s crowd preprint review

This review reflects comments and contributions by Ruchika Bajaj, Gary McDowell, Sree Rama Chaitanya Sridhara. Review synthesized by Iratxe Puebla.

The preprint studies the process for mitochondrial targeting of mitochondrial precursor proteins. Using a yeast model, experiments show that the cytosol transiently stores matrix-destined precursors in dedicated granules which the authors name MitoStores. The formation of MitoStores is controlled by the heat shock proteins Hsp42 and Hsp104, and suppresses the toxicity arising from non-imported accumulated mitochondrial precursor proteins.

The manuscript is clear and well-written. The reviewers raised a few comments and suggestions as outlined below:

The introduction was extremely clear and provides a good summary of the protein homeostasis dimension of the problem in question. However, there could be a clearer discussion of the processes of import, in particular with respect to the results discussing “clogging”. It is suggested to add a penultimate transitional paragraph in the introduction that facilitates this transition e.g. this could be expansion of the first paragraph in the Results section, moved into the introduction to provide more context about the cloggers, PACE, and the Rpn4-mediated proteasomal regulation.

Figure 2E and Figure S2 - Can some further explanation be provided about what data belongs to delta-rpn otr WT, or whether the associated fold change is reported - delta-rpn/WT.

Results ‘while the levels of most chaperones were unaffected or even reduced in Δrpn4 cells, the disaggregase Hsp104 and the small heat shock protein Hsp42 were considerably upregulated (Fig. 2F, G)’ - Suggest adding some further clarification as to why Hsp104 and Hsp42 are selected despite perturbations in other protein partners. Are there other proteins than proteosomes and chaperones which are significantly up- or down-regulated? STRING or cytoscape tools may help with the interactome analysis.

Figure 3

• Figure 3A - It seems Δrpn4 cells are bigger in size than control cells, suggest commenting on this point.
• Figure 3B ‘Hsp104-GFP was purified on nanotrap sepharose’ - Please clarify on which tag the purification was based.
• ‘grown at the indicated temperatures’ - Please clarify the rationale for using 30 or 40C.
• ‘SN, supernatant representing the non-bound fraction’ - Please report what is total, wash and elute etc.

Results ‘protein accumulated at similar levels as Hsp104-GFP in the yeast cytosol (Fig. S4B)’ - Please clarify whether the image reports qualitative or quantitative data, and how the levels of DHFR-GFP and Hsp104-GFP are compared based on S4B.

‘Owing to the striking acquisition of nuclear encoded mitochondrial proteins in these structures, we termed them MitoStores’ - Suggest providing some discussion about the fraction of Hsp104 that is part of the MitoStores? Does a major portion of Hsp104 in the absence of Rpn4 form MitoStore structures?

Figure S5 C ‘Quantification of the colocalization of Hsp104-GFP with Pdb1-RFP after clogger expression for 4.5 h.’ - Suggest normalizing the intensity with one another.

Results ‘Upon clogger induction, the RFP signal formed defined punctae that colocalized with Hsp104-GFP’ - The Hsp104-GFP pattern seems different between Fig 3A, 5, and S5. In some cases, clear punctae are seen and in others, a diffused pattern. Can some comment be provided on this? This might be important to score the colocalization between Hsp104-GFP and other protein partners tagged with RFP. If different conditions were used in the figures, recommend specifying this in the figure legends.

Discussion ‘We observed that MitoStores are transient in nature and dissolve…’ - Suggest adding some discussion about the half-life of MitoStores, and about what the different stressors that can trigger MitoStores may be.

On 2022-10-07 00:51:41, user Michael Ailion wrote:

This manuscript investigates the cellular basis of reduced insulin secretion in Prader-Willi syndrome (PWS) by generating several independent INS-1-derived cell lines to model PWS. These PWS model cells have reduced insulin secretion and reduced levels of insulin and several other secreted peptides. A possible mechanism is suggested by transcriptomic and proteomic analyses, demonstrating the reduced expression of a number of endoplasmic reticulum (ER) chaperones that may be important for proper folding of insulin; reduced levels of chaperones may lead to reductions in insulin levels, though it should be emphasized that this is just a model and hasn’t been tested. The experiments are performed well and the data are solid and convincing, though the effect sizes are of rather small magnitude and it is unclear how important the small effects seen here are to the pathophysiology of PWS. The extensive and rigorous molecular characterization of the mutations in the PWS model cell lines is a particular strength, and the fact that several independent PWS and control cell lines are generated increases the confidence in the results. The proteomic and transcriptomic datasets generated in this work are important contributions to the field. We have a number of relatively minor critiques, many related to the writing and presentation of the work.

Specific comments:<br /> 1. Though the data appear to be solid, virtually all of the effects are of small magnitude, < 2 fold (e.g. insulin secretion, altered expression of chaperones at both the mRNA and protein levels, ER-stress sensitivity). This is fine, but is not readily apparent from the way the paper is written (e.g. line 389, where insulin secretion is described as “dramatically reduced” or elsewhere where effect sizes can only be gleaned by careful examination of the figures). More transparency and explicit discussion of the effect sizes would be helpful. For instance, it would be helpful to compare the effect sizes seen here to those of PWS mouse models and human patients.

1. The take-home model of the paper is that the effect on insulin in PWS is due to an indirect effect on chaperones. This is a reasonable and interesting model, but given that the magnitude of the downregulation of these chaperones is actually quite small (appears to be less than 2-fold for most or all of them), it seems possible that some other mechanism is at play and this may be a red herring. It is unclear whether such a modest decrease in multiple chaperones would produce the observed effects on insulin content and secretion, though it is an interesting question for future work. However, it would also be nice to see the full lists of proteomic data presented as supplementary tables for interested readers so possible alternative targets can be more easily explored.

2. It is concluded that PWS cells are unable to compensate for decreases in chaperones because many chaperones are simultaneously downregulated. This argument does not make a lot of sense to us, and would seem to depend on the mechanism of compensation, which is not further described here. For example, if chaperone genes are transcriptionally downregulated in PWS mutants, what precludes an independent compensation mechanism from simply turning transcription back up, unless the PWS genes are important for the compensation process itself? It would help to present more about what is known about this compensation mechanism, and whether it occurs transcriptionally or posttranscriptionally. A small decrease in many chaperones does not inherently seem to preclude a possible compensation mechanism.

3. The paper is rather difficult to read with lots of jargon and a poor narrative flow. The reader has to do a lot of work to figure things out on their own without much help. Several examples of this are provided in the next few comments(#5-8).

4. Some genes or proteins come up quite suddenly without mentioning their functions or significance. For example, in lines 85-90, SNRPN, SNORD116 and SNORD107 have not been introduced yet as PWS genes, which makes the subsequent conclusion confusing that PWS genes function in beta cells.

5. In several places in the proteomics and transcriptomics sections, there are long lists of genes or proteins with very little context to orient the reader. It is hard for the reader to make much of these lists, and some guidance as to why they are considered worth pointing out or short take-home messages in these sections would be useful.

6. The description of engineering PWS INS-1 cells is quite hard to follow. Figure 1B is not very intuitive. These sections demanded a lot of work from the reader, much of which required looking at supplementary figures to understand the main Results sections. As many readers may not look at these figures, it would help to make this section more accessible.

7. The rationale for performing RNA-seq of small RNAs is not provided. This section ends up interrupting the main narrative and feeling tangential.

8. Since PWS cells have altered levels of many proteins, it is unclear whether the total protein content is a good parameter to use for normalization of insulin secretion.

9. It would help to see the unnormalized raw data for the insulin secretion experiments. Figure 2 shows pooled data from several cell lines. It would also be helpful to see the data for each line separately in a supplementary figure.

10. It is not stated how many times proteomic and transcriptomic experiments were replicated. It is stated that each was performed on three control and three PWS cell lines, but it is unclear if each line was tested just once. It seems likely that the data depicted in the figures are pooled from the different lines though this is not stated explicitly. More clarity on these points would be useful. Separate figures for unpooled data of each cell line would also be useful in the Supplement so that variability between lines can be seen.

11. The study emphasizes the deficits in secreted peptides and ER chaperones but doesn’t provide an explanation for proteins that are increased. A number of neuronal active zone proteins are reported to have increased expression at the mRNA level, but for most it is unclear whether this effect extends to the protein level (only CHGB is labeled in Figure 3). The possible relevance of these changes is also unclear. It is pointed out that many of these proteins may play a role in insulin secretion, but it is unclear why potentially increased levels would lead to the decreased secretion observed in PWS cells unless these factors are negative regulators of insulin secretion (though that seems unlikely given their neuronal functions). Thus, the relevance of these results is unclear.

12. It would be helpful to explicitly state in the Results how many of the genes with reported changes in RNA levels were validated and were not validated by RT-PCR experiments.

13. It is unclear whether it is standard practice to use an anti-KDEL antibody in Western blots to specifically identify GRP94 and GRP78, given that this antibody would be expected to recognize many proteins. If so, it would be helpful to cite other articles that validate this method or state the same thing.

14. Electron microscopy images in Fig S17 show one picture of each cell line, leading to the conclusion that PWS cells have normal ultrastructure. It is unclear what criteria were used to make this apparently subjective conclusion (no quantitative data are presented). Also, there is no mention of how many cells and sections were examined.

15. Fig S18: confocal cell images are difficult to assess. It would be helpful to zoom in on one cell for better comparison. As with EM data, it is unclear what criteria were used to compare the PWS cells to control and no quantitation is provided, nor is there mention of number of cells examined.

Reviewed (and signed) by Michael Ailion and Chau Vuong

On 2022-10-06 16:52:55, user Christina Nord wrote:

How is the choice of what individual-level variable constitutes a "block" made? In Pearl and Schulman (1983) they quote Schulman and Boorman (1983) and state, "Very roughly, the criterion for comembership of two individuals in the same block is that they should bear similar relationships to the remaining members of the population, evaluating “similarity” simultaneously across all types of networks for which data are available." Could age be considered a block, for primates, if sex is? TIA!

On 2022-10-06 14:59:36, user Sarah wrote:

Please note that the mutation is to amino acid 13 "E13Y" not E12Y. Great work!

On 2022-10-06 13:59:43, user Afshin Beheshti wrote:

This pre-print has been accepted in Cell Reports in October 19, 2021 and can be found here: https://www.cell.com/cell-r...

On 2022-10-06 13:37:37, user Kirk Overmyer wrote:

This work has now been published at Plant Communications:

Fuqiang Cui, Xiaoxue Ye, Xiaoxiao Li, Yifan Yang, Zhubing Hu, Kirk Overmyer, Mikael Brosché, Hong Yu, Jarkko Salojärvi,<br /> Chromosome-level genome assembly of the diploid blueberry Vaccinium darrowii provides insights into its subtropical adaptation and cuticle synthesis,<br /> Plant Communications,<br /> Volume 3, Issue 4,<br /> 2022,<br /> 100307,<br /> ISSN 2590-3462,<br /> https://doi.org/10.1016/j.x....<br /> (https://www.sciencedirect.c...

On 2022-10-06 06:20:06, user Mahendra Gaur wrote:

This protein from monkeypox shows the 25-40% similarity with human Dual specificity protein phosphatase and Protein tyrosine phosphatase. However, for identification of potential therapeutic drug targets, essential and non-host homologous protein were considered. How this protein can be considered as drug-target against MPXV?

On 2022-10-05 17:10:51, user Pierre Joubert wrote:

A few concerns have been raised about this manuscript that we wish to address specifically here as well as in a second version of this preprint.

Concern 1: We report a large number of eccDNA breakpoints in our samples in this manuscript. This large number and diversity of eccDNAs could be the result of contaminating linear DNA. This contaminating linear could call into question our results. In general, our pipeline could be calling other artefacts as eccDNAs, and has not been as thoroughly vetted as other previously published pipelines.

Our response:<br /> • We have performed a thorough degradation of linear DNA in all our samples which follows the standards set by our colleagues in the field. We also verified this degradation using qPCR.<br /> • We provide two PCR experiments that support our claims of little to no linear DNA contamination in our samples.<br /> • We used split reads and discordant reads to identify eccDNA forming regions. Any remaining linear DNA that was directly sequenced would not result in either of these read variants and would therefore not result in calls by our pipeline. We verified this by running our pipeline on whole genome sequencing data from other studies.<br /> • While the phi29 polymerase prefers to amplify circular DNA, contaminating linear DNA may also be amplified, resulting in multimeric sequences that, when sequenced, result in split reads and therefore eccDNA calls by our pipeline. This is an accepted weakness of this protocol in the field. However, this method is still widely used by our colleagues, pointing to the fact that this weakness is not enough to invalidate the analysis of eccDNA sequencing data. Clearly, the preference of the phi29 polymerase for circular DNA is strong enough to allow meaningful analysis. All of the data we re-analyzed in this study used the phi29 amplification protocol and therefore would have been similarly affected by linear DNA contamination, but these samples did not show the abundance of eccDNAs we saw in M. oryzae.<br /> • We sequenced O. sativa samples in addition to our M. oryzae samples to verify that our lab methods were not the source of our observations in M. oryzae. Our sequenced O. sativa samples appeared very similar to the samples produced by a previous study across many characteristics and looked very different from what we saw in M. oryzae. Specifically, the number of eccDNAs identified in those samples were much smaller than in M. oryzae.<br /> • We compared our called eccDNA forming regions to those called using the same sequencing data by a previous study (Møller et al. 2018) and found that our results were largely similar, and our criteria for eccDNA calling was more stringent than those previously published.<br /> • Other previously published pipelines, like ecc_finder, use peaks of sequencing reads in the genome as the primary basis for identifying eccDNAs in the genome. However, given the large diversity of eccDNAs we found in M. oryzae we were unable to rely on a peak calling based approach for our data, and instead wrote our own pipeline that relies entirely on split-mapping reads and opposite facing read pairs which are strong evidence of eccDNAs, and used in conjunction with peak-calling in other pipelines.<br /> • We also compared our Illumina called eccDNAs to eccDNAs called using PacBio data in the same samples. This data is much easier to interpret as long split reads are clear evidence of eccDNA formation. We found substantial overlap between eccDNAs called using our Illumina data and using our PacBio data.

Concern 2: We report very little overlap in eccDNA breakpoints between samples, especially among technical replicates, which calls into question the results, especially when it comes to the relevance of genes being found on eccDNAs. This little overlap, combined with potential linear DNA contamination, could point to this study simply over-analyzing noisy data without biological significance. While we explain this lack of overlap by pointing, in part, to under-sequencing, we also claim that this cannot be the reason a subset of genes are never present on eccDNAs in our data which seems contradictory.

Our response:<br /> • The number of eccDNA forming regions we identified in each of our technical replicates suggests that each sample contained an extremely large number of distinct eccDNA molecules. Our analysis of split read counts per eccDNA showed that the majority of these eccDNA molecules were very likely present at very low copy numbers. Furthermore, replication of individual eccDNA molecules in M. oryzae is likely to be very rare or non-existent.<br /> • We split our technical replicates after DNA extraction. Given the hugely diverse population of low copy number eccDNAs in each of these samples, it is extremely likely that some eccDNAs ended up in one aliquot and not the others. This likely explains the lack of overlap in exact breakpoints between technical replicates.<br /> • We also showed that increasing our sequencing coverage per technical replicate likely would have led to better overlap between technical replicates. However, this likely would not have completely solved the problem given the low copy number of the eccDNAs.<br /> • Given the little overlap in breakpoints between samples, we instead sought to analyze the hotspots and coldspots for eccDNA formation in the genome. We felt that this analysis was meaningful because, while our technical replicates did not share exact breakpoints, they did share many approximate breakpoints (if we allow boundaries to be within 100 bp of each other, for example) pointing to the existence of these hotspots. We compared this overlap in breakpoints to our expected overlap if we had sequenced random segments of the genome in each sample and found no such overlap.<br /> • We chose to focus on these hotspots and coldspots by taking a gene-based perspective and counting how often M. oryzae genes were fully contained within eccDNAs in our data, regardless of the exact breakpoints of the eccDNAs. This helped us identify a group of genes that we never found fully contained within an eccDNA in any of our samples. In this case, we were able to show, through rarefaction and permutation analyses, that we do not expect that increasing our sequencing coverage would have led to the discovery of all of these genes in our samples. We used this analysis as evidence that, while under sequencing may have affected our ability to detect overlap in exact breakpoints between samples, it did not explain this observation.

On 2022-10-05 09:40:05, user Iratxe Puebla wrote:

Review coordinated via ASAPbio’s crowd preprint review

This review reflects comments and contributions by Ruchika Bajaj, Pablo Ranea-Robles, Sree Rama Chaitanya Sridhara. Review synthesized by Iratxe Puebla.

In this preprint Munhoz et al. identify adiponectin as the main effector of the protective effects of sera from lean women and calorie-restricted rats on beta-cell integrity and glucose-stimulated insulin secretion. The study reports that sera from obese humans and rats impairs beta-cell integrity and insulin secretion in the absence of nutrient overload. This observation implies that changes in circulating factors between obese and lean individuals would explain the effects on beta-cell function. The levels of circulating adiponectin in rat sera and human plasma were consistent with the metabolic effects observed in beta-cells. Finally, adding adiponectin to islet cultures that were incubated with sera from obese individuals restored beta-cell integrity and glucose-stimulated insulin secretion. The data are reported in a clear way and the manuscript is well written. Data are consistent with a role of adiponectin in the observed protective effects, but some additional experiments are suggested to clarify this role.

Major comments

The paper states that adiponectin is necessary to maintain islet function and integrity. According to the data reported, it is recommended to amend the conclusions to indicate that adiponectin is “sufficient”. A key experiment to demonstrate that adiponectin is necessary would be to deplete the sera of adiponectin and then evaluate the same parameters on beta-cells/islet primary culture. Adiponectin-receptor KO beta-cells would also help to clarify the role of adiponectin in the protective effects of sera. It may also be worth exploring if there were other hormones or other components beyond adiponectin which may have the similar increase in serum samples.

In Figure 3A/3B, a picture of the corresponding Ponceau used for quantification should be shown next to the adiponectin blot. It would also be helpful to provide the full raw blots as supplementary files to allow for further evaluation, e.g. there seems to be a faint band in 3A above the predicted band which might be cropped in 3B, and there seems to be some difference in protein migration in different samples. Please show as a supplementary figure the full blot for adiponectin with all the samples shown in quantification.

In the blot in Figure 3B there does not appear to be a clear difference between adiponectin levels in lean vs obese women, which would argue against adiponectin having a beneficial metabolic effect when treating beta-cells. It would be useful to provide some further comments on this possible discrepancy.

Figure 4E compares different amounts of glucose with either FBS or no serum+adiponectin. Another condition with only no serum + vehicle for adiponectin should be included as a negative control, as shown in Figure 5.

Minor comments

Abstract - Please specify in which model (cell/islet culture) the effects are observed.

Sex-specific differences - The findings in humans are really interesting. However, only male rats are reported in this manuscript. Would there be any difference between male and female CR-rats sera when applied to beta-cells? This experiment would be a great addition to the paper. If the experiment cannot be completed at this time, there should be a mention to this limitation in the discussion.

Results ‘Fig. 1A shows that the animals on the CR diet gained significantly less weight over the course of 15 weeks, but did not lose mass’ - This text refers to mass, the figure legend says weight, the y axis title on the figure states body mass. Please clarify for consistency.

Results ‘They were within the same age range (Table 1), but were clearly distinct in body mass indexes (BMI), which separated them into lean and obese groups: lean women (BMI 22 ± 0.9, Fig. 2B)’ - Please clarify the reference to Figure 2B in this fragment as the figure does not report BMI.

Results ‘these results show a clear modulatory effect of circulating blood factors on metabolic fluxes in β-cells, which are stimulated by factors present in samples from lean and female subjects.’ - It is interesting that this is only observed for females, does this suggest that there may be sex-related factors involved, instead of or in addition to diet status? Could some further comment be added as to why the effect may only be observed in females.

Please mention in the abstract/discussion that the results are obtained from in vitro experiments using beta-cells and islet primary cultures.

Conclusions: suggest specifying “in the blood of lean rats” in the fragment that states “... in the blood of lean animals”.

Methods

Please report the method of euthanasia.

‘experiments were carried out in accordance with the A. C. Camargo Cancer Center Institutional Review Board under registration n°. 3117/21’ - Please clarify whether the study received ethical approval, or was exempt from this requirement at this setting.

Please report what type of fetal bovine serum (FBS) was used (e.g., charcoal-stripped FBS) as well as the FBS catalog number.

‘sera from both groups were collected to be used on cultured INS-1E β-cells, under physiologically relevant conditions’ - Please provide further clarification on the conditions applied.

‘adiponectin supplementation in the plasma from obese donors’ - Please report how this was prepared.

‘Data were expressed as means ± standard error of the mean (SEM)’ - There is a concern about using SEM to illustrate the distribution of data points, please consider using SD.

On 2022-10-03 20:00:55, user Alexander Alleman wrote:

Much more experimentation and scholarship are required to make such an extraordinary claim in the title and such statements as "proved the direct consumption of atmospheric N2 by eukaryotic organisms ". My opinion is that this manuscript offers little evidence of these stated results.

My biggest concern is the contamination of media and agar with residual nitrogen and the lack of multiple replatings to remove residual nitrogen within the cell. Besides the statement on line 23, the authors do not list how many times the yeast strains were replated on nitrogen-free agar to confirm that residual nitrogen is not being used.

Regular agar tends to have some small amount of nitrogen left over from its purification process. We have experienced that nitrogen-fixing bacteria will not derepress nitrogenase unless on pure agar. We use noble agar to make nitrogen-free plates. But for these experiments where absolute proof is required, I suggest using electrophoresis grade agarose as the thickening agent of plates

For the media lacking Mo and Fe, was chelating or acid washing of the growth bottles performed to remove Fe and Mo? Does yeast require some Fe for growth?

Suppl Fig 6 does not seem the yeast grew very well compared to Fig 1. Therefore it seems that Mo or Fe might be required, or there are residual metals or nitrogen in the media.

A positive control in parallel with an acetylene reductase assay is required for the GC-TCD dinitrogen consumption assays. This will show that known biological nitrogen fixation bacteria act similarly to the yeast in your experiment. Azotobacter vinelandii is a model aerobic diazotroph with lots of experimental data to compare to.

I am afraid the dinitrogen consumption assay is a very unusual way of determining nitrogen fixation in a normal atmosphere. Most N2 consumption assays are performed in vitro with nitrogenase under an argon atmosphere as one can measure the change in concentration of N2. I am afraid the authors are measuring the change in partial pressure after the consumption of O2 in the vial.

15N gas enrichment or 15N natural abundance assay must be performed to confirm atmospheric dinitrogen assimilation.

Ammonia or the total N of the cells is never measured. Therefore, it should be easy to compare the biomass before and after and determine if there is more nitrogen in the bioreactor.

Why are the GC dinitrogen consumption assays performed on media with added nitrogen? Where are the assays with nitrogen-free media? Biologically why would yeast be fixing large amounts of nitrogen when it's freely available?

S. cerevisiae is one, if not the most, studied organisms in the history of biology. Is there any evidence in the literature that provides any clues that it is fixing nitrogen under ammonia-supplemented conditions?

Please provide a standard curve of nitrogen and oxygen for the GC data. How do you calculate the initial N2 concentration in the headspace?

Suppl Fig 10. The 16s primers seem to not work well on the positive control of E. coli. If a small population of bacteria was symbiotic with the yeast, this gel does not show they are not there.

The authors have failed to produce any convincing evidence that yeast strains can fix atmospheric nitrogen. While careful experimentation and multiple controls might still prove their hypothesis correct, contamination from nitrogen in the media or a diazotrophic bacteria is most likely allowing the small amount of growth seen on plates. In addition, the GC assay does not have proper controls and is inadequate to show nitrogen fixation.

On 2022-10-03 17:53:16, user Chase Clark wrote:

Not sure if this is a pre-publication thing but one of the links in the paper doesn't resolve:<br /> https://www.user.gwdg.de/~c...

but this one does:

https://wwwuser.gwdg.de/~co...

On 2022-10-03 10:04:15, user Ramesh Vetukuri wrote:

This manuscript is now published in Plosone and the link is below

Antifungal compounds, GC-MS analysis and toxicity assessment of methanolic extracts of Trichoderma species in an animal model

https://doi.org/10.1371/jou...

On 2022-10-03 09:55:42, user Iratxe Puebla wrote:

Review coordinated via ASAPbio’s crowd preprint review

This review reflects comments and contributions by Luciana Gallo, Claudia Molina Pelayo, Sónia Gomes Pereira, Asli Sadli. Review synthesized by Iratxe Puebla.

The preprint examines the meiotic recombination co-factor MND1 and its role in the repair of double-strand breaks (DSBs) in somatic cells. The paper reports that MND1 stimulates DNA repair through homologous recombination (HR) but is not involved in the response to replication-associated DSBs. MND1 localization to DSBs occurs through direct binding to RAD51-coated ssDNA. MND1 loss potentiates the G2 DNA damage checkpoint and the toxicity of IR-induced damage, opening avenues for therapeutic intervention, particularly in HR-proficient tumors.

The reviewers raised some minor comments and suggestions on the work:

Results ‘Therefore, we conclude that MND1-HOP2 are ubiquitously expressed proteins’ - we understand that the study looked at the transcript's expression level and not protein levels, consider revising this sentence.

Figure 1F - Due to the differences in intensity for the loading control, recommend quantifying the normalized level of MND1.

‘we used live-cell imaging of RPE1 cells’ - Are these cells p53 KO? In Suppl. Figure 1K, RPE Delpta-p53 cells are used , but the HALO tag was introduced in the normal (WT) RPE cells. Could some clarification be provided for this difference, and report what's the level of MND1 and the effects of its loss in WT RPE cells?

‘Analysis of 53BP1 foci formation and resolution in asynchronously growing RPE1 cells revealed that MND1 depletion leads to slower repair and retention of DSBs after IR (Figure 2A, Suppl. Figure 2F&G)’ - While the quantification shown in Figure 2A is explicit, the foci in the raw images displayed in Suppl. Figure 2G appears to be more frequent in the siNT, especially in the last 2 time points. It may be worth making the images bigger and maybe clearer?

‘our data show that the role of MND1 in DNA repair is most prominent in G2 phase cells and restricted to repair of two-ended DSBs’ - Can some further context be provided for the last part of this claim. Is this due to the different modes of action of the different drugs used? If so, it would be nice to clarify in the text which drugs induce the two-ended DSBs.

‘These data show that MND1 is recruited to sites of DSBs’ - The data shows that there is an increase in MND1 foci, but whether these are or not the sites of DSBs is not clear. Recommend co-staining with a known DSBs marker.

Methods

• Haploid genetic screen - Please describe how cells were fixed.
• Please detail if/what software was used for the Fisher’s exact test.
• ‘Cells were fixed after 7 days of growth in 80% methanol and stained with 0.2% crystal violet’ - Please report at which temperature and for how long the steps were completed, and provide a reference for the crystal violet reagent.
• ‘Membranes were blocked in 5% BSA’ - Please report the temperature and duration for this step.
• Please describe how the propidium iodide staining was performed.

On 2022-10-03 09:44:44, user Iratxe Puebla wrote:

Review coordinated via ASAPbio’s crowd preprint review

This review reflects comments and contributions by Vasihvani Ananthanarayanan, Sam Lord, Rinalda Proko, Luciana Gallo, Sónia Gomes Pereira, Asli Sadli, Mugda Sathe, Parijat Sil. Review synthesized by Iratxe Puebla.

The preprint studies the molecular function of Arl15, a member of the Arf-like GTPases (Arls) group, which has been linked to magnesium homeostasis. The manuscript reports that Arl15 localizes in the Golgi and plasma membrane, including filopodia. The dissociation of Golgi or the expression of Arf1 dominant-negative mutant leads to a mislocalization of Arl15 to the cytosol. Knocking down Arl15 results in reduced filopodial number, altered focal adhesion kinase organization, and enhanced cargo uptake. Arl15 knockdown decreases cell migration and enhances cell spreading and adhesion strength. The findings point to a functional role for Arl15 in the Golgi.

General comments

Figures 1,2, 3 - The images display one representative example, recommend providing quantification (e.g. PCC/Manders) across several biological replicates, as well as information on the type of images reported, single slice, max Z projection etc.

For the bar plots, the paper reports the number of cells as well as the number of times the experiment was repeated, which is excellent. However, it is unclear whether the SEM error bars and p-values were calculated based on the number of repeats (correct) or based on the number of cells (incorrect). Can clarification be provided for this point. See https://doi.org/10.1083/jcb... and https://doi.org/10.1371/jou....

Throughout the paper there are several references to ‘data not shown’ - please report the data for those items.

Specific comments

Introduction, first paragraph - Suggest shortening the paragraph, particularly regarding the description of the different Arls and their relationship/correlation with all diseases.

‘These results show that similar to HeLa cells, Arl15-GFP localizes to PM along with filopodia and Golgi in all mammalian cell types’ - Suggest revising the fragment to ‘all the mammalian cell types tested in the study’, to avoid generalizing to every mammalian cell type.

‘the localization of Arl15-GFP to PM however remained unchanged as compared to DMSO treated cells (Fig. 2A).’ - Fig 2A only compares mCherry-UtrCH against Arl15-GFP. To support this claim, Arl15-GFP would need to be compared to WGA-AF, as in Figure 1, and their colocalization quantified to confirm that it remained unchanged.

‘We treated mCherry-UtrCH expressing HeLa:Arl15-GFP stable cells with a small molecular inhibitor of Rac1 (CAS 1177865-17-6) or Cdc42 (ML141)’ - Please report the concentration of both inhibitors.

‘Overall, these studies indicate that neither actin depolymerization nor the key regulatory molecules of filopodia/lamellipodia affect the localization of Arl15 to PM/Golgi.’ - The visualization reports Arl15-GFP v mCherry-UtrCH, to support the claim please check against WGA/GM130 as in Figure 1. Also, Figure 2c Arl15 for FAK inhibitor looks different from the DMSO control, recommend confirmation with WGA staining. Can also some explanation be provided for the fact that the Arl-15 in Figure 2A and 2C DMSO looks quite different from 2B and 2D despite the stable cell line with uniform expression?

‘which mislocalized Golgi pool of Arl15 without affecting its PM localization (Fig. 2D)’ - There does not seem to be a marked difference in Arl15-GFP's intracelluar localisation in cells with and without microtubules, and the PM signal appears slightly reduced in the Nocodazole-treated cells. Is it possible to please quantify the localisation?

Figure 2 -The quality of the images from panels B and D looks very different from those of A and C. Can some clarification be provided, were the same microscope, camera, and settings used?

Figure 3 - It would be good to mention the role of brefeldin A as an ATPase inhibitor to provide context for why it is being used.

‘Surprisingly, Arl15-GFP localized to the cytosol as similar to Arf1-GFP in GM130 dispersed cells that are indicative of brefeldin A treatment in HeLa cells (Fig. 3A).’ - It may be worth clarifying the reference to a surprising result, considering the nocadozol results would this result not be expected? It may also be worth providing some comments about the possible PM localisation difference when Golgi is disrupted with nocadozol vs BrefeldinA/golgicide A. It seems that the PM localisation is also affected in the BrefeldinA/golgicideA treatments.

Figure 3A ‘Cells were treated with DMSO (as control), brefeldin A or golgicide A for 24 h followed fixation’ - Please comment on the 24-hour period, BFA would be expected to work in minutes timescale: https://rupress.org/jcb/art...

Supplementary Fig 2A - The blots for Arl15 endogenous are very different between S2A and S2B. Also a 40% knockdown of Arf1 decreases the level of Arl15 by 17%. Can some comments be provided on the significance of this decrease.

Figure 4 - Is the SEM over 3 independent experiments or total number of cells from the three experiments? What was the criteria used to define a structure as filopodia?

‘However, we continued with Arl15V80A,A86L,E122K cytosolic mutant to study the functionality of Arl15 in HeLa cells’ - It may be worth specifying the reason to use the V80A,A86L,E122K form instead of the more simple V80A alone?

‘To test whether the mislocalized Cav-2 and STX6 are targeted to lysosomes in siArl15 cells’ - Please comment on why colocalisation with lysotracker or lamp1 positive structures was not examined instead of treating the cells with bafilomycin A1? Note that bafilomycin A1 also inhibits retrograde membrane traffic at the ER–Golgi boundary: https://www.molbiolcell.org...

Figure 5 - Please clarify whether the quantification of images was done on images taken from the same microscope? Also, suggest arranging the figures in a way that the quantification and images are not so far apart from each other.

Figure 5D - It is unclear how the western blot of EGFR showing total EGFR is indicative of what happened to its trafficking, this appears to be in contrast to the increase in transferrin uptake data. Recommend normalizing the transferrin uptake to surface transferrin levels as one can have higher uptake simply because there is more transferrin receptor instead of actual changes in trafficking rates.

‘Nevertheless, the reason for the partial loss of STX6 and caveolin-2 localization from Golgi in the ASAP1/2 knockdown cells requires investigation’ - Text earlier mentioned "However, we have not observed any significant change in Arl15 and its dependent cargo (caveolin-2 and STX6) localization to Golgi in siASAP1/2 cells " and there does not appear to be any difference in the siASAP1 or siASAP2 on Fig 6. However, in Figure S3 there is a slight reduction in the intensity. Can this be clarified?

Methods ‘Post chase, cells were washed with 1X PBS, fixed with 3% formaldehyde…’ - Please report for how long and at which temperature the fixation step was completed.

On 2022-10-02 18:32:50, user Carrie Partch wrote:

The labs of Seth Rubin and Carrie Partch at UCSC jointly reviewed this preprint. This manuscript examines how the two transactivation domains (TADs) of β-catenin interact with several domains of CBP/p300 to potentially control transcriptional activation. A combination of biochemistry, NMR, and ITC studies narrow down several binding sites for TAZ1 and TAZ2 domains. Overall, the manuscript is well organized and provides new details on these interactions that may play a role in β-catenin function. We have some suggestions that might enhance the clarity of the work below. Thanks for an enjoyable read!<br /> Figure 1: <br /> • The schematics do not depict consistent widths/domain lengths and CBP/p300 is missing some domains, including one implicated in β-catenin binding (Emami et al. 2004, PNAS).<br /> • It would be helpful if your schematic also illustrated all of the constructs used in the study and these names were used consistently.<br /> • If space allows, a simplified diagram of the pathway described in the text could be helpful.

Figure 2:<br /> • It would be helpful to add dashed line for predicted secondary structure cut-off at 0.3 in panel B.<br /> • It would enhance the rigor of the work to show expression levels for the constructs used in panel C.

Figure 3:<br /> • Labeling the MW markers, light and heavy IgG chains, and proteins (does the 666-781 fragment overlap with the light chain?) in panel A, along with the input, would make this figure easier to read.<br /> • The results of the pulldown seem pretty straightforward so quantification from n = 2 experiments seems unnecessary. If you do so, please define what ‘relative’ means in the quantification and make sure that your statistical methodologies are appropriate for this low n.

Figure 4:<br /> • There is some concern that the ITC data might be overfit to a two-site binding model without more information on the fits. Additional rationale or evidence that justifies use of the two-site model would also be welcome.

Figures 5 & 7:<br /> • It might be easier to interpret the binding interface if you used surface representation for the TAZ domains instead of ribbon.

Figure 8:<br /> • It was a bit confusing to show an analysis of the NMR data from the construct used in Fig. S3 in panel A, but then use data in panels B – E from a larger construct. We struggled a bit throughout the manuscript to match domain names and fragments (see Fig. 1 comment above) with the data.<br /> • It could be helpful to conclude with a cartoon or schematic that illustrates what was learned here.

Other:<br /> • Discussion text mentions a possible role for phosphorylation of serines; if citations for this exist, please add them or perhaps broaden this to a possible role for PTMs in general.<br /> • Consistent labeling of ITC data throughout the paper would help clarify which constructs were used in each experiment.

On 2022-10-02 17:56:40, user Carrie Partch wrote:

The labs of Seth Rubin and Carrie Partch at UCSC jointly reviewed this preprint. This manuscript examines how the two transactivation domains (TADs) of β-catenin interact with several domains of CBP/p300 to potentially control transcriptional activation. A combination of biochemistry, NMR, and ITC studies narrow down several binding sites for TAZ1 and TAZ2 domains. Overall, the manuscript is well organized and provides new details on these interactions that may play a role in β-catenin function. We have some suggestions that might enhance the clarity of the work below. Thanks for an enjoyable read!<br /> Figure 1: <br /> • The schematics do not depict consistent widths/domain lengths and CBP/p300 is missing some domains, including one implicated in β-catenin binding (Emami et al. 2004, PNAS).<br /> • It would be helpful if your schematic also illustrated all of the constructs used in the study and these names were used consistently.<br /> • If space allows, a simplified diagram of the pathway described in the text could be helpful.

Figure 2:<br /> • It would be helpful to add dashed line for predicted secondary structure cut-off at 0.3 in panel B.<br /> • It would enhance the rigor of the work to show expression levels for the constructs used in panel C.

Figure 3:<br /> • Labeling the MW markers, light and heavy IgG chains, and proteins (does the 666-781 fragment overlap with the light chain?) in panel A, along with the input, would make this figure easier to read.<br /> • The results of the pulldown seem pretty straightforward so quantification from n = 2 experiments seems unnecessary. If you do so, please define what ‘relative’ means in the quantification and make sure that your statistical methodologies are appropriate for this low n.

Figure 4:<br /> • There is some concern that the ITC data might be overfit to a two-site binding model without more information on the fits. Additional rationale or evidence that justifies use of the two-site model would also be welcome.

Figures 5 & 7:<br /> • It might be easier to interpret the binding interface if you used surface representation for the TAZ domains instead of ribbon.

Figure 8:<br /> • It was a bit confusing to show an analysis of the NMR data from the construct used in Fig. S3 in panel A, but then use data in panels B – E from a larger construct. We struggled a bit throughout the manuscript to match domain names and fragments (see Fig. 1 comment above) with the data.<br /> • It could be helpful to conclude with a cartoon or schematic that illustrates what was learned here.

Other:<br /> • Discussion text mentions a possible role for phosphorylation of serines; if citations for this exist, please add them or perhaps broaden this to a possible role for PTMs in general.<br /> • Consistent labeling of ITC data throughout the paper would help clarify which fragments of b-catenin were used in each experiment.

On 2022-10-01 15:44:36, user Phillip Gordon-Weeks wrote:

This study is an elegant confirmation of the well-established fact that growth cones need dynamic microtubules to execute a turn while providing new optogenetic tools to interrogate +TIP functions. The demonstration that EB3 cannot substitute for EB1 in maintaining microtubule growth perhaps should not come as a great surprise since EB3 occupies a more proximal position than EB1 at the microtubule plus end as first shown in cell lines (Dart et al; 2017, Oncogene 36, 4111-4123 https://doi.org/10.1038/onc... Roth et al., 2019, J. Cell Sci., 132, 1–18. https://doi.org/10.1242/jcs... and, more recently, in cortical neurons (Poobalsingam et al., 2021 https://doi.org/10.1111/jnc....

On 2022-09-30 22:29:46, user MIT Microbiome Club wrote:

Small things: S6, S7,... There's a typo in multiple figures, "Addative". The sign of Figure 4D and E might be reversed. In Methods "rfc" is supposed to be "rcf". The author's point could be strengthened by mentioning the RMSE of the trio models in the text (rather than just in the Figure), as was done for the pair models. Figure 3A is a little confusing regarding the relationship between all the bar graphs-- could be useful to add an "=" before the 4th bar graph in each series so show what each model predicts the outcome to be.

On 2022-09-30 22:29:26, user MIT Microbiome Club wrote:

Figure 2D displays nearly bimodal distribution of effect on focal by pairs of affecting species. Could be nice to explore if this difference is consistently (across RP, BI, CF) due to the same groups of affecting species.

On 2022-09-30 22:29:17, user MIT Microbiome Club wrote:

The authors should clarify if growth rate is included in "metabolic profile" in the list of things that correlated with effect size across focal species.

On 2022-09-30 22:29:03, user MIT Microbiome Club wrote:

It is know that the effect of a drug, and therefore perhaps a bacteria, can be non-linear with concentration, and that dose-additivity (Bliss) predicts drug combinations better than effect-additivity (DOI: 10.1038/s41564-018-0252-1). How might this impact the results? Of course, it can hard to half the concentration of the bacteria in the setup to create the Bliss model (although maybe something with lower glucose might help?) Might the fastest grower in each pair/trio reach the highest concentration, limit the growth of other species, and therefore provide an effect quite similar to itself in isolation? How would a "fastest grower" model compare to a "strongest" model? The authors note that growth rate did not correlate with effect size for "some" focal species, but might such a model work for the remaining focal species?

On 2022-09-30 22:28:44, user MIT Microbiome Club wrote:

The discussion should mention a limitation of this method is its inability to detect interactions that depend on spatial structure (it is known that higher order interactions can depend on spatial structure DOI: 10.1038/nature14485) to fast growers, and to only single carbon source (DOI: 10.1038/s41559-020-1099-4). It should also be noted that any interactions that affects flourescence rather than growth would be misinterpreted.

On 2022-09-30 22:28:34, user MIT Microbiome Club wrote:

The selection of 30 strains chosen for the trio experiment seems biased towards negative interactions, for which the strongest model is particularly good in the doublet model. The authors might consider repeating this with a subset biased towards mixed and/or positive interaction to get a more complete picture.

On 2022-09-30 22:28:26, user MIT Microbiome Club wrote:

Please report what method is used for adjusting optical density and what is the potential range in values. Were all cells in stationary phase before renormalizing?

On 2022-09-30 22:28:15, user MIT Microbiome Club wrote:

The additive interaction model is particularly bad for interactions of negative-negative interaction pairs. Is there a "floor" below which negative interactions cannot be measured or realized? This seems to be the case in Figure 3B: all observed effects are -4 or higher. While the authors explored adding a ceiling (carrying capacity) to the model, a floor should be explored as well and might improve the accuracy of the additive and mean models.

On 2022-09-30 22:28:04, user MIT Microbiome Club wrote:

How good were the different models at predicting the *sign* of the interactions? (Minor)

On 2022-09-30 22:27:27, user MIT Microbiome Club wrote:

Focal species were transformed to constitutively express a fluorescent protein - is this on a plasmid or integrated in the chromosome?

On 2022-09-29 04:07:22, user ygc_smu wrote:

now published on https://onlinelibrary.wiley...

On 2022-09-29 00:11:28, user John Hulleman wrote:

This work is now published in iScience:<br /> https://www.cell.com/iscien...

On 2022-09-29 00:09:40, user John Hulleman wrote:

This work is now published in Current Eye Research:<br /> https://www.tandfonline.com...

On 2022-09-29 00:08:22, user John Hulleman wrote:

This work is now published in ACS Chemical Biology:<br /> https://pubs.acs.org/doi/10...

On 2022-09-28 19:26:59, user Lori O'Brien wrote:

Interesting findings, great to see this translated to mammalian TFs. This was originally shown for the Drosophila TF bicoid in 2000, that the ARM was important for RNA-binding and it was similar to HIV proteins (https://pubmed.ncbi.nlm.nih.... The authors do not cite this though, it would be great to see that study recognized.

On 2022-09-28 12:46:20, user Gary J. He wrote:

The snATAC-seq in this preprint seems to be sci-ATAC-seq?(https://www.science.org/doi...

On 2022-09-27 23:04:37, user anonymous wrote:

Throughout: it is confusing to shift between common and scientific names. Both Marine Iguana and Amblyrhynchus are fine but switching between them in the discussion is confusing.

On 2022-09-27 23:04:28, user anonymous wrote:

line 75: I wonder if it is worth also mentioning valved nostrils/nares here - in fact there is no mention of "valves" throughout - why not? This seems like a central secondary adaptation for a transition from terrestrial to aquatic ecosystems.

On 2022-09-27 23:04:22, user anonymous wrote:

line 67: What is meant by "over-coming buoyancy"?

On 2022-09-27 23:04:14, user anonymous wrote:

lines 58-59: It might be helpful for some readers if you identify the fossil reptiles you're referring to. Several of them are "house-hold" names (like Mosasaurs) so will be useful to some readers not as knowledgeable about such groups.

On 2022-09-27 23:03:56, user anonymous wrote:

Lines 54-56: It's not clear to me what you mean here: does "reptile" refer to crown-group Reptilia or a more "evolutionary" definition of reptiles that includes stem amniotes and synapsids as well? It is not clear to me how a modern clade could have played a role in the origin of other tetrapod clades.

On 2022-09-27 22:37:19, user hunterk wrote:

Line 142-144-I had a quick thought about the use of brain tissue for signs of a stress response due to elevational changes. I might expect to see those changes more in the fat body because that's often cited as the main hub of the immune and stress response. If it feels useful, it might be good to describe how the brain response could directly affect behavior more than a gene expression response in the fat body. (Apologies if you do describe this, I might have missed it.) In general, I'm also curious how these gene expression patterns might relate to aging in these bees-do they show similar signatures of aging (perhaps as defined by DEGs related to the Ti-J-LiFE pathway https://royalsocietypublish...

On 2022-09-27 21:32:48, user Charles Warden wrote:

Hi,

Thank you for posting this preprint.

I receive an error message when I try to access the following link:

https://chenlabgccri.shinya...

Is there something in the path that should be changed and/or something on the server that is needed to activate the link?

Thank you again!

Sincerely,<br /> Charles

On 2022-09-27 10:22:51, user Matthew Herron wrote:

Very cool stuff. If I may make one minor suggestion, I'd have liked a short description of the selection protocol earlier in the manuscript. Fig. 2 kind of shows it, but a sentence or two in the Intro would add to the context for the Results (this is assuming the current order with the Methods at the end).

On 2022-09-27 09:53:01, user Emmanuel Gaquerel wrote:

Hi, thanks for the feedback. The corresponding zenodo link should work now.

On 2022-09-20 13:48:52, user Galano Jean-Marie wrote:

Such an impressive work. How long did it take...? so good.

anyhow still reading but cannot get to the zenodo link for QcxMSin the supporting<br /> https://doi.org/10.5281/zen...<br /> congrats

On 2022-09-27 00:29:34, user Joshua Mylne wrote:

This work is now published (22 Sep 2022) at Nature Communications under a variant title "A fungal tolerance trait and selective inhibitors proffer HMG-CoA reductase as a herbicide mode-of-action" PMID: 36137996 PMCID: PMC9500038 DOI: 10.1038/s41467-022-33185-0<br /> https://doi.org/10.1038/s41...

On 2022-09-26 15:10:46, user anonym wrote:

There are some mistakes in the influent column in the tables. The values for S_h2 and S_ch4 should be 1.0E-... instead of 10E-..., and the influent value of S_H+ should be 0.0 instead of 1.0

On 2022-09-26 14:59:23, user Xiaoyu Weng wrote:

This manuscript has been accepted by Molecular Biology and Evolution: https://doi.org/10.1093/mol...

On 2022-09-26 13:27:22, user Gauthier wrote:

Where can I find the supplementary data including the species list? They are not supposed to be submitted to biorxiv along with the main text ?

On 2022-09-26 12:05:30, user Miha Moškon wrote:

The final version of this paper has been published at https://doi.org/10.1016/j.biosystems.2022.104778

On 2022-09-25 19:03:27, user smd555 smd555 wrote:

Good day, dear authors!<br /> I have a question: do these newly described hunter-gatherers from the Middle Don show any genetical similarity to Sredniy Stog samples from Ukraine (I4110 and I5882)?

Best regards

On 2022-09-25 03:03:28, user Peter Uetz wrote:

I would say explicitly what the yellow cavity is in Fig. 7 (I guess it's the foregut, as shown in other figures), but it's a good idea to make this explicit for non-experts. I was wondering already when I looked at the Liem paper before I found your paper.

On 2022-09-23 12:58:38, user Laura Rossini wrote:

Pleased to announce that the final updated and peer-reviewed version of this manuscript was published in Frontiers in Plant Science. Laura Rossini

Bretani G, Shaaf S, Tondelli A, Cattivelli L, Delbono S, Waugh R, Thomas W, Russell J, Bull H, Igartua E, Casas AM, Gracia P, Rossi R, Schulman AH and Rossini L (2022) Multi-environment genome-wide association mapping of culm morphology traits in barley. <br /> Front. Plant Sci. 13:926277. doi: 10.3389/fpls.2022.926277

https://doi.org/10.3389/fpl...

On 2022-09-23 05:03:03, user Bela Toth wrote:

Hi Michi,<br /> impressive work, although I slightly disagree with you on the basic assumption of evolution. Nevertheless, I have a technical issue with this manuscript: The number of mRNA transcripts contigs that you show for the different genes in Figure 5 are not well defined. You forget that genes can produce several transcripts by the process of alternative splicing. Therefore labeling the x-axis with the gene names is a bit misleading.

Take care!

On 2022-09-22 18:44:51, user john wallingford wrote:

The FAK-based ciliary adhesion complex is an enigmatic structure in motile ciliated cells, and this paper is a welcome contribution for both the technical advance (new fixation methods) and new cellular insight (links to the apical microtubule network).

On 2022-09-22 18:25:45, user john wallingford wrote:

The paper also has me thinking about patterns of cytoplasmic mechanics. Microheology shows that cytoplasmic stiffness differs in different regions of migrating cells. How do such patterns relate to the propagation of forces at the cortex/membrane?

On 2022-09-22 18:23:46, user john wallingford wrote:

Using an elegant new technique, this paper reveals new insights in the role of the plasma membrane and the actin cortex in the propagation of forces across single cells. For this developmental biologist, the paper provides an exciting new paradigm to explore further in multi-cellular tissues, in particular as we seek to understand recent findings of mechanical heterogeneities in individual cell-cell junctions during morphogenesis (e.g. Huebner, 2021: https://pubmed.ncbi.nlm.nih... Cavanaugh, 2022: https://pubmed.ncbi.nlm.nih...

On 2022-09-22 18:20:07, user Robert J. Huebner wrote:

Belly et al investigate membrane tension transmission across individual cells. They find that membrane tension is strongly propagated in response to cellular protrusions or pulling on membranes and the actin cortex. However, pulling on the membrane alone does not stimulate tension propagation. One exciting conclusion is that the cell cortex opposes tension propagation when force is applied to the membrane alone. It would be interesting if the authors proposed a mechanism for how the cortex resists tension propagation when force is only applied to the membrane.

On 2022-09-22 18:12:57, user Shinuo Weng wrote:

Beautiful work! I'm curious why the actin flow upon actin pulling is in the opposite direction to the membrane tension propagation. Thank you!

On 2022-09-22 18:07:02, user Austin T. Baldwin wrote:

De Belly et al describe different dynamics of membrane tension propagation dependent on whether they perturb the cell membrane or cortical actin. Adhesive linkers between the membrane and the cortex are essential in their model of how this tension is propagated, but what these linkers are or could be is poorly explained. More discussion on the possible identities of these linkers and subsequent perturbation of these linkers (if possible) would enhance an already-fascinating set of experiments.

On 2022-09-22 16:24:14, user PSauer wrote:

This bioRxiv manuscript, combined with its companion manuscript "Structures of the Cyanobacterial Phycobilisome", has now been puplished in Nature doi: 10.1038/s41586-022-05156-4

On 2022-09-21 03:31:40, user Daniel E. Weeks wrote:

What a fun and interesting paper! It even applies Student's t test to Student's data!

I am interested in your variable selection (Table S2) with the disparate results between the different methods.

One minor suggestion would be to merge Table S3 into Table S2, as it would be nice to be able to see these metrics at the same time as we're seeing which variables were selected.

Regarding variable selection, I really like this discussion of variable selection issues:

Heinze G, Wallisch C, Dunkler D. Variable selection – A review and recommendations for the practicing statistician. Biometrical Journal. 2018;60(3):431–449. DOI: https://doi.org/10.1002/bim...

I like their recommendation to "assess selection stability and model uncertainty", which is what we ended up doing in this recent paper:

Heinsberg LW, Carlson JC, Pomer A, Cade BE, Naseri T, Reupena MS, Weeks DE, McGarvey ST, Redline S, Hawley NL. Correlates of daytime sleepiness and insomnia among adults in Samoa. Sleep Epidemiology. 2022 Dec;2:100042. DOI: https://doi.org/10.1016/j.s...

We had originally wanted to use a lasso where we forced in a few variables that we thought had to be in all models, but couldn't get existing software to work in our hands to enable proper post-selection inference. When no variables are forced in, proper post-selection inference after variable selection via lasso can be done using these approaches:

1. Taylor J, Tibshirani R. Post-Selection Inference for ℓ1-Penalized Likelihood Models. Can J Stat. 2018 Mar;46(1):41–61. PMID: 30127543 PMCID: PMC6097808 DOI: https://doi.org/10.1002/cjs...

2. Lee JD, Sun DL, Sun Y, Taylor JE. Exact post-selection inference, with application to the lasso. The Annals of Statistics. Institute of Mathematical Statistics; 2016 Jun;44(3):907–927. DOI: https://doi.org/10.1214/15-...

On 2022-09-20 11:36:44, user Prof. T. K. Wood wrote:

YgiW (referred to here several times) has been studied previously and linked to hydrogen peroxide stress resistance:

https://sfamjournals.online...

So perhaps this should be cited.

On 2022-09-19 14:09:03, user Gregory Way wrote:

We reviewed this preprint as a part of Arcadia's preprint review initiative: https://twitter.com/Arcadia...

Peidli et al. present a data resource (for single-cell perturbations) and apply energy distance (e-distance) to quantify differences in perturbations. For the data resource, the authors focus on curating single-cell RNAseq and ATACseq measurements perturbed with CRISPR, drug treatments, and a few other perturbation types. The authors curate a total of 44 datasets. Overall, the paper is very well written with a sound logical flow. However, many elements of the paper seem incomplete. We provide several specific comments regarding our views on how the paper could improve. We thank the authors for posting their preprint and code publicly.

Our two primary comments are:

1. The data are not harmonized from reads. Instead, the authors process (in most cases) already processed read count by gene matrices. The authors also use different versions of scanpy to process different datasets. This is definitely still valuable, but the authors should state these facts earlier and probably decrease the use of “harmonization”. Additionally, there is no evaluation to determine the effect or benefit of this read count harmonization. Calculating e-distance before and after harmonization across datasets might be helpful.

2. E-distance is not sufficiently benchmarked. The math and intuition are described marvelously, but how does E-distance behave across datasets and common perturbations? How does subsampling read depth impact E-distance calculations? How does drug dose impact e-distance? How does sequencing technology impact e-distance? How does modifying the distance metric within the E-distance calculation impact calculations?

We also have several general comments on different aspects of the paper and github repository. We hope that the authors can benefit from our deep dive on the paper. Thanks again!

Introduction

• Definition of single-cell perturbation data (SCPD)

Overall, this subsection is more of a “methods/techniques overview” of how to collect SCPD rather than defining what SCPD actually is. What is output from these techniques?<br /> - The authors should define these data in more detail.<br /> - The authors should also further define the techniques as it is helpful to have a general idea of why the data collected from the techniques are “good” and not just “more data are better”.

Motivation for distance measure of high-dimensional profiles:

• The authors claim that E-distance can identify strong or weak perturbations. It’s unclear what a strong or weak perturbation is. I was unable to find this information from a quick google search so I think they should define that here (not found in methods either).

Motivation for unifying datasets

• Their motivation only seems to be “it doesn’t exist yet because it’s difficult to do” so therefore we should do it. What will/could come of the integrated and standardized datasets? What would we hope to find?

Web Interface

• The authors claim, “a web interface for data access, analysis and visualization is available at scperturb.org.” There is data access on that site, but analysis and visualization appear to be absent using Brave and Safari browsers.
• It seems that one would require a computer with enough memory (500G) to run scPerturb to reproduce the analysis. The authors present solutions for how to overcome these requirements, but it did not seem that they attempted to solve them.
• The authors state that there are Quality Control plots for each dataset on the website but we could not find.

Results<br /> - The authors should briefly describe the methods underlying the statement “dense low-dimensional embeddings of the original data (see Methods for details)” in a bit more detail upon introduction.<br /> - It is surprising to me that there are so many cells with 2 perturbations (proportionally to a single perturbation) (sup fig 1). Is this because of an overweighting of a specific study?<br /> - It might be helpful to add targeted sequencing depth to table 1 per study, also helpful to add the sequencing platforms used.<br /> - Data source trust: Zenodo sources appear to be auxiliary data downloads as opposed to direct sources. How might other researchers assume trust in the sources? Are the included metadata implied or entrusted to the authors?<br /> - Are the UMAPs in Figure 3E the same UMAP space or are the spaces fit independently in both panels?<br /> - Need to provide a bit more rationale for why the authors chose E-distance over the other options.<br /> - Did they calculate E-distance for all perturbations? Sup Fig 3 shows this, so maybe? It was not obvious where to find the measurements.<br /> - There are only 11 drug perturbations in common. This is a very interesting observation! How many genes are perturbed in common datasets?

Methods<br /> - For the scATAC-Seq data, it’s not clear to me if they perform LSI jointly across all samples or not. This would cause non-interoperability across datasets if not done jointly since each LSI dimension may mean something different in each dataset. In addition, they provide peaks x counts matrix -- which is dataset specific. I would suggest aligning jointly using a uniform set of peaks -- Running MACS2 on all datasets would be a huge benefit to the community.<br /> - How do the different versions of scanpy impact data processing? Typically, harmonized data are generated with a single pipeline.<br /> - When performing subsampling to fit PCA, did the authors transform the full data subsequently? In other words, does the PCA fitting step impact cell count for e-distance calculation?<br /> - What distance measure is used in the E-distance calculation for ||x_i - x_J||? L2? For perturbations, comparing L2 to other metrics would help benchmark the method.

Code/Github<br /> - It seems to us a good idea to spend time improving the existing model / code at https://github.com/theislab.... The authors should justify why they are not contributing to existing open source code.<br /> - I can’t find the script “fragments2outputs.R” in their github. From their paper: “All features described in the overview above were computed with ArchR functions. For details inspect the “fragments2outputs.R” script in our code repository (see Data Availability).”

Data Repo comments:<br /> - Manual data testing for reproducibility within https://github.com/sanderla... (one must perform the steps, the repo doesn’t provide or outline within the code itself)<br /> - Suggests using “mamba” but does not provide instructions on how to install mamba <br /> - Would suggest a small description for each folder in the directory (README) explaining its contents <br /> - There’s no usage example on how to download the data or use the program<br /> - Would be best to have a notebook (or bash script) that describes the entire workflow. <br /> - The notebooks are not sequentially executed and there are no execution instructions<br /> - What environment (OS/hardware/configuration/etc) is required to run the code?<br /> - Is notebook (.ipynb) output expected within committed code? (should these have been scrubbed with nbconvert/jupytext?)

Data Availability<br /> - Based on this section, their website only contains the first three bullet points (e.g scRNA-seq data, scATAC-seq data, and details about the datasets). We could not easily find the last three bullet points (Quality control plots for each dataset, Filtering, e.g., by readout or type of perturbation, Commands for direct file download using the Unix command curl)

This review was produced jointly at The University of Colorado by:

Gregory P. Way, PhD<br /> Natalie Davidson, PhD<br /> Erik Serrano<br /> Parker Hicks<br /> Jenna Tomkinson<br /> Dave Bunten

On 2022-09-19 08:47:03, user ricky wrote:

I suppose, In fig.2 the result of SMEM should be "ATCTCAGTC" rather than "ATATCAGTC"

On 2022-09-19 07:58:28, user zhljude wrote:

Hi Thomas Burger:

This article is a nice work. However, the low resolution of the Figures makes it confused to understand the content of the article. Could you provide clearer Figures ?

Best regards<br /> Jude

On 2022-09-18 23:37:46, user Sebastian wrote:

Hi, cool paper!<br /> I had a bit of trouble understanding one thing though: how do you prevent or even correct for false positives in this procedure? Could you measure FPR?

It seems to me that the narrow selection of literature (enriched in associations with Depression/AD), the relatively long reasoning chains, and the paths through hub nodes such as "Nervous System Process" can very easily result in hits that "biologically make sense"; what prevents this process from just returning each and every entity that is connected to one of the parent nodes of Depression and AD? You even encourage hierarchical chains, so the chance for parent hub nodes is almost 1 I would assume.

Generally, what is the reasoning for allowing these chains to cross through nodes with such low specificity?

On 2022-09-17 01:01:54, user Chengxin Zhang wrote:

A standard approach for protein structure compression is MMTF, which is a lossless compression format supported by RCSB PDB. After including metadata, does PIC outperforms MMTF in compression rate, with and without gzip compression?

On 2022-09-16 21:43:47, user Simon Zhongyuan Tian wrote:

Simon Zhongyuan Tian, Guoliang Li, Duo Ning, Kai Jing, Yewen Xu, Yang Yang, Melissa J Fullwood, Pengfei Yin, Guangyu Huang, Dariusz Plewczynski, Jixian Zhai, Ziwei Dai, Wei Chen, Meizhen Zheng, MCIBox: a toolkit for single-molecule multi-way chromatin interaction visualization and micro-domains identification, Briefings in Bioinformatics, 2022;, bbac380, https://doi.org/10.1093/bib...

On 2022-09-16 12:07:25, user EM wrote:

This is a very important paper. It indicates that classifying and understanding the crystal polymorphisms that occur during protein/enzyme reactions with its ligand in crystals can lead to a detailed understanding of protein reaction mechanisms. This paper will become increasingly important with the development of the 4th generation synchrotron radiation facilities.

On 2022-09-15 15:36:04, user Foster Birnbaum wrote:

The difference between Figures 4 and 5 is striking. In Figure 4, UniRep/BO is clearly superior to the other versions, whereas in Figure 5, even a random sequence performs well until higher iterations and there is never a clear difference between UniRep/BO and the other methods. You highlight that this task is very specific and is convex as explanations for why UniRep/BO is not clearly better, but I am still wondering why the performance for any method is not much better than random. Also, you also state that the sequence length is fixed at thirteen residues. In part A of the results, you mention that an advantage of BO is that the sequence length can change during optimization. Have you experimented with letting BO run with a variable length sequence on the unknown target matching problem? In addition, can you perform the AlphaFold task using the two ablated methods? I think including the ablation results for all three tasks would be helpful.

I think Figure 1 could be made clearer if the sequences and labels proposed from the logits went directly to the train step, instead of being directed first to a separate shape. This would make Figure 1 have a more triangular structure, with the logits at the top, the UniRep vector in the bottom left, and the prediction plot in the bottom right. I think this could help make the flow of information clearer.

On 2022-09-15 01:35:35, user Sebastian Swanson wrote:

When designing peptide binders, how does bayesian optimization compare to alphafold hallucination as far as runtime? It seems like the main benefit of BO is that it could require less forward passes through the network to find a binder with high pLDDT, but since extra time is required to embed the sequence with uniref, train the MLPs and find the gradient, it’s not clear how significant the speed up will be. Regarding the alphafold predictions of your design and the patgiri peptide, it seems like loop 1 is collapsed into the protein domain, as opposed to wrapping around the peptide, as one might expect it to given the Ras-SOS complex (PDB ID: 1NVW). Do you think it’s possible that alphafold is struggling to model the binding of the patgiri peptide due to the flexibility of the loops, and that is responsible for the abnormally low pLDDT? Have you considered using alternative methods, like openfold (which has different weights) or Rosetta PIPER-FlexPepDock to test whether this peptide is predicted to bind?

On 2022-09-15 07:03:40, user Iratxe Puebla wrote:

Review coordinated via ASAPbio’s crowd preprint review

This review reflects comments and contributions by Joe Biggane, Luciana Gallo, Rachel Lau, Sam Lord, Dipika Mishra, Claudia Molina. The comments were synthesized by Iratxe Puebla.

The study reports two two Bcl-2 family proteins, BNIP5 and Bcl-G, which inhibit Bak-dependent apoptosis through engagement of MODE 2 inhibition. The BH3 domains of these proteins act as selective Bak activators, while not inhibiting anti-apoptotic proteins, leading to increased binding of activated Bak to Mcl-1, which prevents apoptosis.

The reviewers raised a couple of questions about the methodology and several other suggestions for the paper, outlined below:

Methodology

Throughout the study various BH3 mimetics are used, but the combinations in which they are used and/or the doses employed could be more clearly reported. For example, in Figure 1E and 1F ABT-737 and S63845 are used at 1 μM. Then, in Figure 1H, A-331852 is substituted for ABT-737 in combination with S63845 and the concentration is not reported. In Figure 1H, ABT-737 and S63845 are used again, but this time at a concentration of 2 μM each. Other concentrations are used in Figures 2, 3, and 5. There seems to be a dose-response assay in Figure 3B, but it is used for a specific use case. It would be beneficial to report all combinations and doses employed, and the rationale for them in the main text, to allow readers to fully interpret the data presented.

In various figures, there is a concern about the statistical approach to calculate p-values based on multiple measurements or cells within each sample. The t-test and ANOVA assume that each measurement is independent, and multiple nuclei within the same sample are not independent. Recommend either not reporting p-values or averaging together the values from each biological replicate to calculate the p-value using those sample-level means. For more information, see https://doi.org/10.1371/jou... and https://doi.org/10.1083/jcb...

Specific comments

Introduction ‘The two MOMP effectors Bcl-2 associated x (Bax) and Bcl-2 antagonist killer (Bak) are inactive in resting cells as these cells exhibit low levels of proapoptotic BH3-only proteins (e.g. BIM)....and some are additionally able to activate Bax and Bak (sensitizers and direct activators, e.g. BIM)’ - Recommend revising the fragment for clarity, adding references to support the statements and possibly an introductory figure to help visualize the proteins involved.

Introduction, last paragraph ‘We found that two Bcl-2 proteins, Bcl-2 interacting protein 5 (BNIP5) and Bcl-G, act as selective inhibitors of Bak-dependent but not Bax-dependent apoptosis…’ - The fragment is unclear, BNIP5 and Bcl-G are first reported as Bak-inhibitors, then activators and back to inhibitors. Does this mean to describe protein-protein interaction and changes in conformation?

Figure 1

• Recommend using a different color scheme for Figure 1E to assist visual interpretation of the results, in particular consider using a color-blind friendly color palette.
• ‘colony formation (F)’ - The text later on refers to ‘clonogenic survival’, would it be possible to clarify in the legend or text what is being assessed, i.e. recovery assay, clonogenic survival or colony formation?
• Figure 1G - Please clarify whether 2 uM of each are used in this experiment.
• ‘We transduced PC9 lung cancer or A375 melanoma cells…’ - It is nice to see that different cell lines were assessed to address any cell line-specific effects. Would be interesting to see if this effect occurs in normal cell lines and not just cancer cell lines.

Figure 2 - The inline color-coded legends are useful when bars are displayed but in the figure several bars are close to zero, consider an alternative method to label the bars.

Results ‘...suggesting posttranscriptional regulation of Bak levels by BNIP5’ - Maybe large proteome databases of multiple cell lines (e.g CCLE) can be datamined to determine the correlation between BNIP and Bak expression?

Figure 3B, 3D and 3E - Please clarify the concentrations used for each treatment in the figure legend.

Methods, Cell viability and cell death measurements - The study assessed cell death or cell viability with either live cell imaging, or in fixed cells, can the methodology for this be elaborated upon? Also, propidium iodide staining is used in several sections of the results, recommend adding information about this under the Methods section.

Methods - There are several missing references in the Methods section.

Suggest adding Supplemental Figure 6 as a graphical abstract.

On 2022-09-15 06:44:56, user Iratxe Puebla wrote:

Review coordinated via ASAPbio’s crowd preprint review

This review reflects comments and contributions by Ashley Albright, Luciana Gallo, Sam Lord, Dipika Mishra, Claudia Molina, Arthur Molines, Sónia Gomes Pereira, Parijat Sil, Rinalda Proko. The comments were synthesized by Richa Arya.

The reviewers like the motivation behind the study as a lot is still unknown about the impact of fluorescent tags on various mechanisms in biology. The work is impactful. However we outline below some major questions and several minor points:

1. Related to the data analysis:

The findings are valuable however the analyses may not be sufficiently sensitive to pick up morphological changes. Maybe other more sensitive approaches for measuring interference in the biology of these neurons could also be tested, like bulk growth rates, a stimulus added to the culture medium or other?

Some of the phenotypes (see Figure 1D and Figure 3D) are relatively subtle and the manuscript relies heavily on statistics to support the claims. Independent of the statistics, the differences are not striking by eye examination. Perhaps more data is necessary to bolster some of the reported claims.

As continuation to the expression analysis, it is important to estimate the expression levels of the actin binding probes used, in order to rule out the fact that some of the observed differences between LifeACT-GFP and AC-GFP may be due to discrepancies in the extent of overexpression of these probes. It would greatly add to the study to include, at least for some of the phenotypes, whether the measured parameters respond to the low versus high expression levels of the same probe.

1. Related to the transient expression:

Figure 1: The transient expression method used in the manuscript shows a lot of variability in expression levels, between cells, and between replicates.

Expression levels could confound the interpretation. One of the constructs could be expressed more or less than the other, resulting in stronger or weaker phenotypes, not because it is more or less toxic than the other per se but because its expression level is different. It would be relevant to "normalize" for the expression level of each construct. Another way to circumvent this, at least partially, would be to substantially increase the number of cells analyzed, which would allow for a range of expression values to be represented in the data.

1. Related to FRAP analysis Last paragraph result 1: ‘and depends only on their affinity to F-actin, that is similar in AC-GFP and LifeAct-GFP (Figure 1A, Figure 1B, Supplementary Video 1). In addition, our results suggest that even without photomanipulation, minor changes in actin reorganization cannot be revealed when using actin-binding probes…’ - Based on the images reported, it is not possible to establish how much of the signal is due to the population of probes being bound to actin versus the population that is free floating in the cytoplasm. The recovery could be due to the diffusion of free-floating probes and therefore give no information about affinity for actin. EGFP alone was used as a baseline for cytoplasmic diffusion, the slower recovery from the EGFP-actin implies that some portion of the EGFP-actin is incorporated in filaments. Recommend replacing "Affinity" with "relative ability to incorporate into filaments." A possibility to address the issue of size-based diffusion in cytoplasm is to complete FRAP measurements in latrunculin-treated cells that depolymerize most of the actin filaments. This will enable to set a baseline for each of the probes here (which will now probably be either free or G-actin bound) and provide a complement to the Jasplakinolide treatment.

‘In addition, our results suggest that even without photomanipulation, minor changes in actin reorganization cannot be revealed when using actin-binding probes.’ - There has been only two actin binding probes tested, both with similar turnover as measured in FRAP in their own assay. It might be worth making a comparison in this experiment with a very strong actin binding probe as control, such as Utrophin.

1. Figure 2: Theat measurement shown is not a very good proxy for filopodia motility.

The study used an intensity-weighted center of mass. This means that the center of mass moves, not only because the shape of the filopodia changes but also because the signal intensity changes. In other words, the shape of a filopodia could be constant (no motility) and yet have a center of mass that moves because the mCherry signal fluctuates inside it. This could be avoided if the center of mass of the shape is used, not weighed by intensity. This is especially a concern because the signal from the cytoplasmic mCherry is used for the analysis. If a folipodia locally thickens in the Z-direction, the cytoplasmic signal will increase locally, displacing the intensity weighted center of mass even if the 2D contour has not changed. Using a membrane signal would provide a better alternative. It would also be possible to make use of the resource Filotracker, that tracks the length of the filopodia as a measure of filopodia dynamics. Find the paper and the resource here: https://www.molbiolcell.org...

1. Result 2 last para, ‘We found no significant difference in center of mass displacement between actin probe expressing cells and EGFP expressing control filopodia (Figure 2B)…’. This section needs more clarity and evidence to conclude that the probes do not alter filopodia dynamics. Maybe filopodia growth rate or some additional measurements? Failing to find significance does not equate to finding evidence of absence. It may be that this one parameter is not sufficiently sensitive. Maybe this possible uncertainty should be discussed in the last sentence of the paragraph, to note that the data highlights the possibility that the tested actin labeling proteins do not interfere.

Minor Comments

Introduction: ‘Actin is a key cytoskeletal element in mammalian cells involved in many cellular mechanisms’. mammalian cells can be replaced with eukaryotic cells. It would also be nice to mention some of the cellular mechanisms involved such as cell division, and migration, among others.

Introduction: it would be good to describe the various phenotypes observed in previous studies when actin was labeled or when actin-binding proteins were used. It would give readers context about the level of toxicity and what phenotypes to expect.

Introduction last paragraph: ‘…and to exclude certain actin structures from labeling (Munsie et al., 2009)’: one more reference could be added for this statement: Sanders et al., 2013 https://www.ncbi.nlm.nih.go...

Figure1: It would be nice to have grayscale images of the actin channel in addition to the overlay.

Figure 1 B, C: For all of the FRAP recovery curves, recommend providing insets, zooming in on the first 30 to 60 sec of the recovery, as that's when most of the recovery happens. The last 120 sec of the plots show a "flat" plateau.

Figure 1D, in the fluorescence recovery plateau %."In addition, our results suggest that even without photo manipulation, minor changes in actin reorganization cannot be revealed when using actin-binding probes." This claim relies on the p-values. In looking at Figure 1D, left panel, EGFP-actin (orange) dots there appears to be an outlier. Independent of the outlier, the collection of dots does not appear that different by eye, recommend providing additional data to support this claim.

Result 1: ‘In Jasplakinolide-treated neurons, as expected, we observed an almost immediate recovery of fluorescence in the EGFP expressing group, whereas the EGFP-actin signal did not recover…’. The fact that the EGFP-actin signal did not recover is surprising. Normally not all of the actin present in the protrusion is incorporated into filaments, some of it is floating around freely. Hence, some of the signal should be recovering, even after stabilization of the actin filament, simply due to diffusion. For example, the EGFP signal recovers in presence of Jasp. due to diffusion of the free-floating probe. Recommend some discussion about the absence of recovery for the EGFP-actin.

Figure 2A: ‘Red lines show the movement track of intensity weighted center of mass..’. The red dots for the center of mass cluster and overlap, recommend color coding the dots so that it is clear visually what the displacement of the center of mass was and showing an overlay of the contours used for the analysis. Additionally, in the Supplementary Video 2 it looks like EGFP and EGFP-actin centers of mass are more displaced than AC-GFP or LifeAct-GFP. It would be good to clarify if this is exactly the same example as shown in the figure.

Figure 2B: ‘Average intensity-weighted center of mass displacement over 60s time periods…’ Why was only a 60 sec interval considered when there are images up to 120 sec and the video goes until 180 sec? Additionally, please specify if these are the first 60 sec of imaging.

Result 3: It is known that expression levels of actin binding probes can alter actin structures and their dynamics. It would have been great to do the following: (a) estimate the levels of expressed lifeact-GFP/AC-GFP and see how they compare with each other, (b) note or look for phenotypic differences as a function of the expression levels of these probes. It might be worth plotting the spine morphometric data in categories of low, medium and high expression levels of the two actin binding probes as well as EGFP-Actin (since this can affect nucleation/treadmilling etc at very high expression levels). Just as the identity of the actin binding probe being used is an important consideration in studies of actin dynamics, so is the expression levels of these probes.

Result 3: use p-values to compare different cell lines, the n used in the statistics should be the number of samples, not the number of spines.

Result 3: ‘This is like due to the known high background fluorescence level of LifeAct, originating from its affinity to G-actin (Melak, Plessner and Grosse, 2017)…’. Actin chromobody is also known to bind G actin. Is there a significant difference in G Actin binding affinity for LifeACT versus AC that can account for this explanation?

Figure 3C ‘Expression of EGFP-actin or LifeAct-GFP for 24 hours did not influence total protrusion density’ - Please indicate whether these morphological analyses were done blinded as to what the cells were expressing, or any steps taken to reduce bias.

Figure 3D: There is a similar concern here as for Figure 1D. Here the number of cells is higher, but the density of the points is not shown. By eye the box plots do not look very different, violin plots may be better for these data so that the distribution of data points is more apparent.

Figure 3F: it would be useful to have a representative image of each (stubby, thin, and mushroom) class, to help non-experts better visualize what's being analyzed .

Result 4, paragraph 1, ‘..whether dendritic arborization is altered within 24 h after the transfection of the tested probes…’ All the experiments were performed 24h after transfection, would it be worth testing different time intervals (e.g. 12-16h and/or 48h)?

Result 4C,D E: suggest adding quantification to enhance the data.

Materials and Methods section, ‘Live cell imaging and FRAP experiments, post-bleach in every case (Supplementary Video 2)…’. Should this read video 1.

Materials and Methods section, ‘Live cell imaging and FRAP experiments,Then, cumulative displacement curves were calculated, and the 60 sec points were compared and statistically analysed (Supplementary Video 1)…’. Should this read video 2.

Materials and Methods section: There are several custom-made plugins used in this work. It is good practice to make these available to the community by depositing them in a repository (e.g. GitHub, zenodo).

On 2022-09-14 23:03:56, user Alex wrote:

The new localization predictor built based largely on this dataset, PB-Chlamy, will be an invaluable tool for the community and may inspire hypothesis-based questions. We were curious if the authors could leverage the combination of protein localization and sequence information to (1) predict sub-organellar localizations within the chloroplast (e.g. stroma vs. pyrenoid etc.) and (2) correlate protein physicochemical properties (based on primary sequence) with localization.

On 2022-09-14 23:03:47, user Alex wrote:

Using immunofluorescence to directly detect PRM1, PHB2, and SNE1 validated localizations from the author’s over expression reporters to the ER/nucleus, cytosol, and nucleoplasm. For proteins localized to novel punctate structures in the chloroplast, localizing a subset of these proteins at endogenous expression levels via epitope tagging and immunofluorescence would rule out the formation of these punctate structures as a consequence of overexpression, especially for puncta that exhibit a robust mobile fraction by FRAP.

On 2022-09-14 23:03:37, user Alex wrote:

It was really impressive to see how visualizing localization though microscopy revealed new structures within the chloroplast and the manual effort it took to classify localization patterns. More information on how localization patterns were classified in the methods and results section would be helpful. Currently, we understand that an individual z-stack was independently analyzed with two people with disagreements resulting in ambiguous/no assignment. But more information on how the number of classes was determined is needed, especially since many novel punctate structures in the chloroplast were discovered. We also wondered whether an automated method of clustering image profiles would yield similar localization assignments.

On 2022-09-14 23:03:28, user Alex wrote:

The ~3,000 overexpression vectors cloned and ~1,000 strains generated will be a valuable resource and the accompanying website will also facilitate access and use of these tools. It was unclear whether localization studies were attempted with all ~3,000 vectors or just the ~1,000 studied. If only 3,000 were attempted, but only 1,000 successfully localized it would be useful for the authors to comment on why the ~2,000 could not be localized (fitness defects as a consequence of tagging, post-transcriptional regulation, whether a subset were localized with other methods/approaches, etc…) and suggest potential alternative approaches.

On 2022-09-14 22:40:22, user Sky wrote:

Currently, the collision rate is based on some assumptions about the efficiency of transfection. I was wondering if a closer approximation of the collision rate could be inferred from the rate of non-self collisions available in the single cell seq data.

On 2022-09-14 20:55:58, user Hannah wrote:

Really interesting integration of single-cell and MPRA techniques. More information about how these CRS's were chosen would be helpful.

I would be curious to see how these promoter activity changes you observe in the HEK293 cells and K562 cells also carry over to changes in the expression of the genes that these promoters control.

For the cell cycle analyses, have you looked at what transcription factors might be driving these changes in promoter activity?

On 2022-09-14 22:22:49, user Ohainle Lab wrote:

These were all empty particles, what would happen if these capsid structures contained HIV genomes? Is there any effect of reverse transcription on “switching”?

On 2022-09-14 22:22:40, user Ohainle Lab wrote:

Is there a way to explicitly measure the stability of pentamers and hexamers?

On 2022-09-14 22:22:30, user Ohainle Lab wrote:

How do we explain that although IP6 binds both hexamers and pentamers and seems to play a more important role stabilizing pentamer form IP6 induces hexamers in vitro?

On 2022-09-14 22:22:18, user Ohainle Lab wrote:

How is hexameric capsid binding to host factors facilitating viral infection? What is the model and at what stage of viral replication would this be important? Assembly? Post-assembly? Would there be a way to test this?

On 2022-09-14 22:22:05, user Ohainle Lab wrote:

Very nice data in Figure 3: we like how you made predictions from the structures and can introduce mutations to toggle capsid protein conformations to go in both (hexamer or pentamer) directions

On 2022-09-14 21:12:15, user Luc Roberts wrote:

Now published in NAR with some serious edits: http://dx.doi.org/10.1093/n...

On 2022-09-14 20:58:14, user CK wrote:

It's mentioned that Replacing Cs with Ch results in a 133-fold difference in the ratio of favorable (hippurate and cinnamoylglycine) to unfavorable (phenylacetylglycine) Phe metabolites. Given that it is known Phenylacetylglycine plays a causative role in cardiovascular disease, did you see any phenotypic differences in ΔCh vs ΔCs mice in terms of cardiovascular disease/mouse growth?

On 2022-09-14 20:56:20, user Chris wrote:

Great paper!

Would it be possible to feed mice with ∆Ch∆Cs communtities with 7alpha-dehydroxylation to see if the metabolite alone is sufficient to prevent the bacterial diversity changes?

Would it also be possible to transform 7a-dehydroxylation synthesis genes into a different bacterium other than Ch or Cs to see if that would rescue ∆Ch∆Cs?

Additionally, were any obvious phenotypes noticed in ∆Ch∆Cs colonized mice?

On 2022-09-14 19:27:49, user David Amici wrote:

Please note that this work has undergone substantial improvement and is now published as two separate articles; one focusing on C16orf72/HAPSTR1 (https://www.pnas.org/doi/10... and one on the coessentiality method, FIREWORKS (https://www.life-science-al...

On 2022-09-14 14:47:30, user Grimm wrote:

It's not the topic of the paper but I'd love to see the STRUCTURE results for k < 6. Your reference data are another mosaic stone in revising the species concept of western Eurasian beeches.

Also, I wonder, given the fit of your reference data (Fig. 1) with studies that focussed on western Eurasian beeches and the underestimating current species concept (cited Denk 1999a,b; Gömöry & Paule 2010; Cardoni et al. 2022), whether one should still use "Oriental beech" in the singular. I'd write "Oriental beeches" to stress the fact it's more than one biological entity. Especially given that you can identify introduced hybrids between Greater Caucasus Oriental beeches (i.e. F. orientalis s.str.) and European beech (F. sylvatica s.str.) and not just between Oriental (F. sylvatica subsp. orientalis) and European beech (F. sylvatica subsp. sylvatica); and in the light of of this being indeed a showcase for the genetic diversity in the Oriental beeches, which, indeed, surpass that in the European beech, in all data assembled so far and highlighted in this study.

Regarding the aspect of level of diversity (especially with regard to F. sylvatica and detection-capacity for [introduced] hybrids), adding the STRUCTURE profiles with k < 6 (it says, you run with k = [1,10]) as supplementary information could be an additional selling point for highlighting the different Oriental beeches as not a single but several, possibly also climatically non-identical, and differentially related to the European beech, genetic resources for the European beech forests and with respect to climate change and AGF potential (which may differ between the Oriental beech spp.)

On 2022-09-13 22:06:06, user Hae Kyung Im wrote:

In this paper, the authors attribute the wrong null hypothesis to the standard TWAS approach. The issue seems to stem from a confusion between the true parameter (a number) with its estimator (a continuous random variable). They state that the null hypothesis is that the estimator = 0, which is an event of probability 0.

The better way to think about the error in the genetic predictors of gene expression is not to change the null hypothesis but in terms of an error in variables problem. Under reasonable assumptions of independence between reference and target sets, error in variables leads to attenuation and not inflation. Many papers have addressed this problem.

More details

On 2022-09-13 21:46:19, user Damien F. Meyer wrote:

Acknowledgements <br /> The authors gratefully acknowledge Géraldine Bossard and Valérie Rodrigues for technical assistance in the development of ELISA assays.

Funding information<br /> The authors acknowledge the financial support from Franco-Slovak bilateral project PHC Stephanik 2014 n°31798XB and from European Union in the framework of the European Regional Development Fund (ERDF), n° 2015-FED-186, MALIN project “Surveillance, diagnosis, control and impact of infectious diseases of humans, animals and plants in tropical islands”.

Conflict of interest disclosure<br /> The authors declare they have no conflict of interest relating to the content of this article.

Author contributions<br /> VP and DFM conceived the original idea. NV and MB acquired the funding. VP, EB, IM, OG, CP and DFM performed. VP, EB, IM, CP and DFM analysed data. VP and DFM drafted the manuscript. All authors read and approved the final manuscript.

Data availability<br /> Data are available on Zenodo public repository at the following address https://zenodo.org/record/5...

On 2022-09-13 18:32:44, user ABHINAV JAIN wrote:

Thanks for providing Ingres tool for GRN.<br /> I am wondering did you guys also compared it with the other existing tools? If yes which one and how was your experience?

Thanks

On 2022-09-13 16:39:13, user Joachim Goedhart wrote:

Great work, I enjoyed reading it.<br /> One suggestion is to split the graph with the 'Bleaching profile' (figure 1D) according to the laser line that was used for excitation. I don't think it is fair to compare FPs that are excited with different laser lines as the illumination intensity of the lines may differ (as far as I can tell the power of the lines was not measured). It would also simplify the comparison between different spectral classes.

On 2022-09-13 14:09:31, user Stephan Emmrich wrote:

Dear Reader,<br /> Please note that the two bioRxiv manuscripts with doi:<br /> https://doi.org/10.1101/859454<br /> https://doi.org/10.1101/202...

Have been published as one paper in EMBO J 2022, please check out<br /> https://www.embopress.org/d...

Thanks,<br /> Stephan

On 2022-09-13 13:34:43, user Miles Markus wrote:

In this interesting paper, earlier studies are referred to in which it was concluded that the first Plasmodium vivax malarial relapses early in life are genetically homologous; and that parasites which gave rise to sequential recurrences in a particular patient with P. vivax malaria were hypnozoite-associated meiotic siblings. The previous authors’ conclusions might or might not be correct. An alternative possibility is that some or all of these recurrences are/were, in fact, hypnozoite-unrelated in that they were recrudescences (not relapses), homologous recurrences being highly suggestive of a clonal merozoite (perhaps non-circulating) origin. SEE (try clicking on the link below): Markus, M.B. 2022. Theoretical origin of genetically homologous Plasmodium vivax malarial recurrences. Southern African Journal of Infectious Diseases 37 (1): 369. https://doi.org/10.4102/saj...

On 2022-09-08 22:06:06, user Walter S Leal wrote:

I read this preprint with vivid interest and took the opportunity here to comment on a couple of issues.

I like the approach of performing repellency tests with a higher throughput assay but having an exit hands-on-cage (real-world) assay. A few relevant issues occurred to me, which the authors may want to consider:

1) It might be more appropriate to report repellency in terms of protection (per EPA & WHO guidelines). Also, it seems that 75% repellency is too low. A good repellent should provide approximately 100% protection for a couple of hours.

2) As you know, the core principle of scientific publication is to provide enough information that a qualified person can perform the work. Several qualified researchers could test whether the newly discovered repellents are indeed more effective than DEET. However, the names of the new repellents, their sources, and chemical characterization (if newly synthesized) were not disclosed. {Is there a SI file I missed?]

3) Suppose there is an issue of intellectual property. In that case, this issue should be addressed first, and then the work should be considered for publication with a complete list of compounds tested, particularly those claimed to be more effective than DEET.

Other minor issues for your consideration:<br /> 1) In the introduction, the sentence referring to the discovery of picaridin needs to be re-phrased. Picaridin was developed before insect ORs were discovered, let alone the co-receptor Orco.

2) Insect ORs are not G-protein coupled receptors, as implied in the introduction.

Thank you for sharing a preprint in BioRxiv. It is a remarkable development.

Walter Leal

On 2022-09-08 13:36:55, user Giovanni Dalmasso wrote:

This preprint has been now published in Developmental Cell:<br /> Dalmasso et al., 4D reconstruction of murine developmental trajectories using spherical harmonics, DevelopmentalCell (2022)<br /> https://doi.org/10.1016/j.devcel.2022.08.005

On 2022-09-07 14:26:19, user Feng Yang wrote:

I am the corresponding author of the original study. [Journal name redacted to follow bioRxiv's policy] rejected this Preprint based on our Concerns on their concern. Unfortunately, I do not know how to publish the PDF file of our response (it does not fit BioRxiv since our PDF file does not contain additional experimental data). I am pasting it below. We welcome open discussion based on solid experimental data and are looking forward to more independent studies in this area.<br /> Re: On the therapeutic potential of MAPK4 in triple-negative breast cancer <br /> Feng Yang<br /> Department of Molecular and Cellular Biology, Baylor College of Medicine, Houston, Texas<br /> * Corresponding Author: Feng Yang, Baylor College of Medicine, One Baylor Plaza, Houston, TX, 77030. Phone: 713-798-8022; Fax: 713-790-1275; E-mail: fyang@bcm.edu<br /> Boudghene-Stambouli et al. recently published “On the therapeutic potential of MAPK4 in triple-negative breast cancer” in BioRxiv concerning our Nature Communications publication, “MAPK4 promotes triple negative breast cancer growth and reduces tumor sensitivity to PI3K blockade.”, published 11 January 2022 (1). We want to reply to their comments as follows.<br /> Boudghene-Stambouli et al. essentially detected a similar MAPK4 protein expression pattern (Our report (1) vs. Boudghene-Stambouli et al., Fig. 1c) in the human TNBC cells, when using the same commercially available antibody AP7298b. However, they claimed, “We failed to detect a specific ERK4 band in any of the cell lines, including Hs578T cells transfected with human ERK4 cDNA.” They then used their own “validated custom polyclonal ERK4 antibody that we use routinely in our laboratories” to produce a different MAPK4 expression pattern (Boudghene-Stambouli et al., Fig. 1c). They provided a siRNA knockdown for the “validation” of their antibody. In this case, Boudghene-Stambouli et al. largely ignored our previous publications using the commercially available AP7298b to successfully confirm the overexpression, knockdown (up to five independent shRNAs), and knockout of MAPK4 in many human cancer cell lines and in “normal” cells (1-4). AP7298b can also detect a purified GST-MAPK4 fusion protein in the GST pulldown assays and the purified Flag/His-tagged wild-type and mutated MAPK4 proteins in the in vitro kinase assays (2). It should be noted that instead of our extensive validation of AP7298b using many MAPK4-overexpressing, knockdown (up to five independent shRNAs), and knockout cells as well as purified MAPK4 proteins (overexpressed/purified from both prokaryotic and eukaryotic cells), Boudghene-Stambouli et al. only used a single siRNA to “validate” their un-named custom antibody. Besides, they did not confirm HA-MAPK4/Erk4 overexpression in their Hs578T cells (Boudghene-Stambouli et al., Fig. 1c). Please note, due to the sensitivity of different antibodies, even if an HA-positive western blot is provided, it may not confirm significantly increased ectopically overexpressed MAPK4 expression over the endogenous MAPK4. Finally, their custom antibody detected many non-specific bands compared to AP7298b (Boudghene-Stambouli et al., Suppl. Fig. 1c, which was included in their submission recently rejected by [Journal name redacted to follow bioRxiv's policy] after peer-review). Therefore, we have concerns over Boudghene-Stambouli et al.’s concern on MAPK4 protein expression levels in the MAPK4-high TNBC cell lines that we used in our study (1).<br /> It is well-known that mRNA and protein abundances may not correlate well in biological systems. Therefore, Boudghene-Stambouli et al.’s concern about the variation of MAPK4 mRNA expression across the cell lines will not carry that much weight. We also noticed that Boudghene-Stambouli et al. used our reported 5’ primer but a modified 3’ primer for their qPCR data in Fig. 1a. We wonder whether they have performed qPCR using our reported 5’ and 3’ primers to detect MAPK4 expression (3), and what were the results? Besides, although we have not systematically examined MAPK4 mRNA expression in human TNBC cell lines as we did for human prostate cancer cell lines (3), we did qPCR confirmed MAPK4 expression in MDA-MB-231, SUM159, as well as the non-small cell lung cancer H1299 cells. Besides, Zheng et al. independently showed MAPK4 mRNA and protein expression in HCC1937 and MDA-MB-231 cells (5), two of the TNBC cell lines concerned by Boudghene-Stambouli et al. Without knowing the quality of Boudghene-Stambouli et al.’s RNA-seq data, we could not comment on their Fig. 1b data.<br /> Another concern of Boudghene-Stambouli et al. is their failure to verify our reported MAPK4-AKT signaling axis, a conclusion drawn from their Fig. 2 data. Without providing their data, the corresponding author Dr. Meloche has communicated with me about this issue. At that time, I provided the following answer. “I am not sure if you did a transient transfection in the 293 cells. Unlike MK5, phosphorylation of AKT is subjected to many more direct and indirect regulations in the cells. It is hard to imagine that you can easily detect MAPK4 phosphorylation of cell endogenous AKT in the transiently transfected 293 cells. It can be a hit and miss, especially if you do not carefully monitor cell confluency. I think that we only reported data from the stable 293T cells overexpressing MAPK4 or MAPK4 phosphorylating a co-transfected AKT in 293T cells. In the latter case, we suspect that these ectopically overexpressed AKT are less susceptible to endogenous cellular posttranslational modifications and more susceptible to the regulation of overexpressed MAPK4. Again, unless you can’t repeat our data, such as MAPK4 phosphorylating a co-transfected AKT in 293T cells, I do not see a common ground for our debate here either.” Now I see the experimental data, and Boudghene-Stambouli et al. did perform a transient transfection and tried to detect phosphorylation change of endogenous AKT, which we have already expressed concern about in our previous personal communications. Interestingly, as a positive control for their Fig. 2 data, Boudghene-Stambouli et al. showed MAPK4 enhanced the phosphorylation of an ectopically overexpressed but not endogenous MK5, raising concern about this so-called positive control per se. We are also unsure how much MAPK4 was overexpressed compared to endogenous MAPK4 (Western blots on GFP could not provide that information) nor the nature of the seemingly increased AKT T308 phosphorylation in the MAPK4 transfected 293 cells (Boudghene-Stambouli et al., Fig. 2).<br /> I want to finish this discussion using what I wrote to Dr. Meloche in another email. “Without detailed information from your side, it is hard for me to guess what happened. I want to emphasize several technical details that may help. 1. Please collect cells at about 50%-70% confluency. If your lab collected cells at very high confluency, please try this. 2. We have been using Dox-inducible knockdown and overexpression approaches. We typically maintain the cell culture without Dox induction and do a couple of days (such as three days) induction just before the experiments. 3. If you use a non-induction system as we did in some of our studies, please ensure that you only use the engineered cell lines at early passages. You can do this by freezing down many vials from a very early passage and only using the thawed-out cells for minimal additional passage(s). The cancer cells in culture may adapt to the cellular “stress” from long-term MAPK4 overexpression or knockdown.”<br /> We welcome open discussions based on solid experimental data. We will do our best to help if any group meets technical difficulty in repeating our data under the reported experimental conditions. We have validated our MAPK4-AKT signaling in more than 20 human cancer cell lines (Ref. (1-3), and unpublished data), and additional independent reports also confirmed MAPK4 phosphorylates/activates AKT in human cancer cells (5, 6). We welcome and are looking forward to more independent studies in this area.<br /> References <br /> 1. Wang W, et al. MAPK4 promotes triple negative breast cancer growth and reduces tumor sensitivity to PI3K blockade. Nat Commun. 2022;13(1):245.<br /> 2. Wang W, et al. MAPK4 overexpression promotes tumor progression via noncanonical activation of AKT/mTOR signaling. The Journal of clinical investigation. 2019;129(3):1015-1029.<br /> 3. Shen T, et al. MAPK4 promotes prostate cancer by concerted activation of androgen receptor and AKT. The Journal of clinical investigation. 2021;131(4).<br /> 4. Cai Q, et al. MAPK6-AKT signaling promotes tumor growth and resistance to mTOR kinase blockade. Sci Adv. 2021;7(46):eabi6439.<br /> 5. Zeng X, et al. MAPK4 silencing together with a PARP1 inhibitor as a combination therapy in triplenegative breast cancer cells. Molecular medicine reports. 2021;24(2).<br /> 6. Tian S, et al. MAPK4 deletion enhances radiation effects and triggers synergistic lethality with simultaneous PARP1 inhibition in cervical cancer. J Exp Clin Cancer Res. 2020;39(1):143.

On 2022-09-07 14:03:33, user Tobias Lee wrote:

Is this article's code open-sourced?

On 2022-09-07 10:19:53, user Scott Hayes wrote:

Firstly I would like to congratulate the authors on their study. I really love the inventiveness of the experimental set up! The results are really unexpected. The finding that phosphate starvation proceeds ABA mediated drought responses is interesting on a mechanistic basis and will likely have direct implications for crop management practices. The field-to-lab experimental pipeline looks really effective and I look forward to more people taking up this approach.

I do have a couple of points that I believe the authors could discuss in more depth. In their ridge trials, the geometry of the soil may play a role. Phosphate starvation induces a lot of lateral roots close to the soil surface, to maximise phosphate capture. In the ridge set-up, lateral roots are restricted to only a single plane. Is it possible that this contributes to the PSR in ridge-grown plants? It is also possible that increased rooting depth under mild-drought treatment also reduces phosphate uptake. If the authors see a reduction in PSR gene expression after re-watering ridge plants, this may help to rule out geometric explanations. It should be noted that the pot experiments do already indicate this to some extent.

I am also not so sure about the authors’ conclusion on why PSR is induced at mild drought but suppressed under severe drought “Under severe drought conditions, given the circumstantial evidence, our observations would support the notion that PSR induction is suppressed by a relative increase in Pi concentration due to a decrease in leaf water content (Fig. 4).” I would argue an alternative, more straight forward hypothesis, that plants simply prioritise drought stress over this level of phosphate starvation when drought is severe. Would the authors expect to see a recovery of the PSR if Pi levels dropped even lower?

Once again, thanks to the authors for their very through-provoking study!

I look forward to hearing from you,

Scott

On 2022-09-06 15:02:12, user Fernando Barroso wrote:

Final version available at JPhysChem B: https://pubs.acs.org/doi/pd...

On 2022-09-06 09:43:18, user Prof. T. K. Wood wrote:

Also, filamentous phage and biofilm formation in P. a. has been shown to be controlled by nutrient levels via substrate-binding protein DppA1 of ABC transporter DppBCDF (doi: 10.3389/fmicb.2018.00030).

On 2022-09-06 09:30:00, user Prof. T. K. Wood wrote:

The seminal discovery of phage inhibition by the toxin/antitoxin Hok/Sok system should be mentioned herein given Hok/Sok was discovered 15 years earlier compared to those cited here and provided the mechanism that was confirmed by the Laub group 26 years later (ref 6). See doi: 10.1128/jb.178.7.2044-2050.1996 and https://journals.asm.org/do....

On 2022-09-05 08:53:44, user Kristian Unger wrote:

Congratulations to this great work! Have you made the software already available somewhere?

On 2022-09-04 13:55:42, user Shrishti Singh wrote:

Cheers to the team on this work.! This is a highly-reproducible way to make contrast agents.

On 2022-09-03 02:22:59, user Babu Sudhamalla wrote:

Now published in ACS Biochemistry doi: 10.1021/acs.biochem.2c00297.

On 2022-09-02 03:31:28, user Milind Watve wrote:

The manuscript received an unusual response from a reputed journal to which it was communicated on 11th Feb 2022.<br /> Our correspondence with the editor as under. Name of the journal and editor is not revealed following the policy of BioRxiv.

Fri, Aug 12, 6:17 PM

to Milind

Dear Dr. Watve,

I am writing with the difficult news that we have not been able to secure an Academic Editor to handle your manuscript "Hyperglycemia in type 2 diabetes: physiological and clinical implications of a brain centered model" (MS number ----------). Additionally, we have been unable to secure feedback from peer reviewers. We have therefore reluctantly decided that we must return your manuscript to you without review.

I recognize that this decision will be frustrating -- it is our desire to provide every suitable manuscript the opportunity for review and evaluation by experts in the research community -- and I sincerely apologize that we have not been able to do so in this case. We have exhausted the pool of potential (journal name) Academic Editors qualified to handle your manuscript but have not been able to secure a commitment to handle the submission. We have also invited a number of peer reviewers with relevant expertise, but we have not been able to secure the reviews required to support an editorial decision. We are withdrawing your manuscript from consideration to prevent further delays in the assessment of your submission, and so that you can move forward immediately if you choose to submit your work elsewhere.

Again, I am very sorry not to have more positive news for you. I wish you the best in finding an alternative venue for this work.

Best regards,

Editor-in-Chief

Milind Watve milindwatve@gmail.com<br /> Sun, Aug 14, 10:23 AM

to --------- bcc: Akanksha

Dear ---------,<br /> I understand the agonies of editors. No issues. But I have one request. <br /> I would like to have your consent to post this letter in the public domain. It is very likely to be a remarkable event in the history of science and students of the history and philosophy of science need to have access to this information. How people in a field react to a paper challenging an existing dogma is a very important question in the history and philosophy of science and making this letter public is extremely essential. Therefore I want to append it to the preprint, as well as write an article about it on my blog on which I have often written about problems in science and science publishing. Link here if you want to view it (https://milindwatve.in/)<br /> Awaiting your response.

milind<br /> (Dr. Milind Watve)<br /> https://milindwatve.in/

Journal name ><br /> Sun, Aug 14, 10:24 AM

to me

Dear Milind Watve

Thank you for contacting --------. We will reply to your query as soon as we are able.

In the meantime, please take a look at the following links for more information about our processes:

We appreciate you reaching out and will be back in touch shortly.

All the best,

Milind Watve milindwatve@gmail.com<br /> Mon, Aug 29, 9:20 PM

to --------

Dear Editor,<br /> This is to inform you that since I did not get any reply from you for over two weeks, I am assuming that you have no objection if I publish your letter in any appropriate context, in a respectful manner.

milind<br /> (Dr. Milind Watve)

On 2022-09-02 01:38:59, user Matthew Templeton wrote:

What similarities are there to the Myrtle rust (Austropuccinia psidii) genome, given that both organisms have broad host ranges and very large Tn-rich genomes?

On 2022-09-01 18:11:32, user Matt Wersebe wrote:

Hi Miguel,

We haven't yet quantified this but we have done some preliminary life table experiments with some of these clones. Anecdotally, high tolerance clones may have a smaller "r" or intrinsic rate of increase. This may reflect a trade off in salt tolerance versus fecundity.

On 2022-08-11 06:49:50, user Miguel Cañedo wrote:

Really nice work, congratulations! I wonder if adaptation has any "cost" for the population. Are there any trade-offs? Maybe the adapted populations would be outcompeted by non-adapted populations under certain conditions?

On 2022-09-01 16:43:00, user Rath R. Weird wrote:

Updated info on the first author:<br /> Vadim Molodtsov Rospotrebnadzor Russian Federation<br /> https://sysbiomed.ru/team/m...

On 2022-09-01 04:23:39, user Titos wrote:

Published: https://www.nature.com/arti...

On 2022-08-31 16:46:02, user Yichi Xu wrote:

Mistake in line 71: SHF is not derived from neural crest. We will correct it in revision. Many thanks to Dr. Etchevers!

On 2022-08-31 13:08:23, user Iratxe Puebla wrote:

Review coordinated via ASAPbio’s crowd preprint review

This review reflects comments and contributions by Amrita Anand, Richa Arya, Aurora Cianciarullo, Luciana Gallo, Dipika Mishra, Sanjeev Sharma, Ryman Shoko and Rajan Thakur. The comments were synthesized by Ehssan Moglad.

The study conducted by Doyle et al. aimed to test the lipid phosphatidylinositol 4-phosphate (PI4P) transfer activity of the human ORP5 protein via orthogonal targeting of different sites of membrane contact, namely, between the plasma membrane (PM) and the mitochondrial outer membrane.

Major comments

Figure 1: The idea behind the experiment is great, however, there are some questions about the data presented:

• Recommend using better representative images, the morphology of the cells in panels B and C appears distorted, are these cells undergoing some death? This would make interpretation of the data difficult.
• In Figure 1B, the control experiment should be done with a known mitochondrial marker to test if the construct works as expected.
• P14P measurement using the P4M probe is unclear from the images and quantification provided. Please provide a multi-plane image of the probe showing its distribution on the PM and at the ER-PM or ER-mito contact site.
• Please provide an additional graph to show the relative change of PI4P at the PM compared to the rest of the cell or respective contact site.
• For a better comparison, recommend showing the normal (control) distribution of PI4P in the images.

Results: ‘As expected, we did not see the accumulation of PI4P at these contact sites (see graph in Fig. 1C), presumably due to SAC1 activity in the ER. Instead, the fluorescence of PM PI4P seemed to decline’: Please indicate whether this result is statistically significant.

Figure 2: The experiment with the FKBP-PI4KC1001 construct is not discussed in the text. Also, further clarification would be helpful for the results presented in panel B. In +SAC1mito, it is showing accumulation after Rapa treatment, please discuss why PI4P is not showing accumulation.

'The rationale was that without inhibition of PI4P synthesis, observing reductions in PM PI4P catalyzed by transport of PI4P out of the PM would require a rate that exceeded synthesis, which may not be possible through reduced flux at the much smaller surface area of induced PM-mitochondria contact sites, compared to ER-PM contact sites (compare Figs. 1C and D). We also imaged by Total Internal Reflection Fluorescence Microscopy (TIRFM) to more sensitively detect changes in PM PI4P with the high-affinity PI4P biosensor, P4Mx2.': Recommend revising the fragment for clarity.

Figure 3: The shape of the cells across figure 3 varies substantially, can some text be added to discuss why this is the case?

Figure 3: The authors have already shown in a previous paper that SAC1 predominantly acts only in the 'cis' configuration. However, induced coupling of overexpressed ORP between ER-PM and mito-PM using rapamycin might bring these membranes closer than usual or cause the formation of more membrane contact sites. Thus, there may be some possibility for SAC1 to act in 'trans'. Alternatively, there could be indirect changes in PM PI4P due to increased activity of endogenous ORP5 at these induced contact sites. To address this:

• Would it be possible to confirm if there was an increase in the number/size of contact sites by checking for mapper expression and localization when ORP5 constructs are expressed and coupled with Rapamycin? Lipid binding mutants of ORP5 could also be used to show that those lipid binding mutants do not cause a depletion upon coupling with Rapamycin.
• For experiments where SAC1 (Fig 3) was overexpressed along with FRB::FKBP-ORP5-ΔTMD, please show control conditions where the SAC1 alone was expressed without the ORP5. Also, in a control condition where lipid binding mutants of ORP5 are expressed along with SAC1, there should be minimal effects on PM PI4P depletion compared to WT ORP5. Adding these controls will further confirm prior observations and substantiate the effects of FRB:: FKBP-ORP5-ΔTMD expression, and rule out any potential artifacts from overexpression of just SAC1.

A major confound across all the experiments is the activity of endogenous ORP5 that is not measured. Is it possible to perform experiments (such as in Figure 3) where the endogenous ORP5 is downregulated using siRNA or shRNA and a siRNA/shRNA-resistant version of ORP5 is overexpressed in this background? There could be potential compensatory effects from other ORP5, and this would require simultaneous knockdown of multiple ORPs.

Minor comments

Figure 1: It could be helpful to start Figure 1 using a scheme of the two hypotheses on how ORP5 regulates PI4P levels at the plasma membrane. This will help easily assess the data presented in the Figures for and against the hypotheses.

Figure 1A: It is difficult to see the co-localization in the current color scheme. It would be helpful to use a different color combination, provide zoomed-in images, or use pointers to highlight.

Figures 1C and D: Please specify in the legend the timepoint when rapamycin was added and the subcellular membrane the measurements were made from.

In discussion: 'In principle, this observation does not demonstrate lipid transfer (though it is compatible with it). Although tethering at a site of membrane contact seems to facilitate access of PI4P to SAC1, ORP5 could simply be presenting the lipid to the phosphatase, as opposed to depositing PI4P into the membrane for subsequent hydrolysis by SAC1. If ORP5 works in such a presentation mode, it is not clear to which membrane the resulting PI lipid is released: either back into the PM, or into the tethered membrane. In other words, lipid transfer is not necessarily part of the reaction'. Are there ways in which this can be tested? Suggest proposing some future experiments in the text.

On 2022-08-31 11:17:49, user Nándor Lipták wrote:

Dear Authors,

In our previous study, we found mosaicism in founder (F0) rabbits, generated by CRISPR/Cas9 gene editing:<br /> doi: 10.3390/app10238508

It is also a common phenomenon in CRISPR/Cas9 gene edited mice.

Have you also detected mosaicism in your founder rabbits?

On 2022-08-31 07:56:29, user Dr. Jaimini Sarkar wrote:

This study @biorxivpreprint has isolated bioactive phytochemical from mangrove - Sonneratia apetala, effective against human pathogenic bacteria.The study also shows that the potency of the extract varies with geographical location of the plant.

On 2022-08-30 11:26:14, user Nándor Lipták wrote:

Dear Authors,

Unfortunately, knockout lethal phenotypes are quite common in mice. We collected the most promising rescuing methods in a review last year, you may find it useful for your present preprint or future experiments:

10.33549/physiolres.934543<br /> https://www.ncbi.nlm.nih.go...

On 2022-08-30 10:08:52, user Prof. T. K. Wood wrote:

There is a mature literature not cited on AMPs killing persister (i.e., dormant) cells that includes AMPs, mimics, and AMPs + antibiotic combinations.

On 2022-08-29 14:32:38, user Victoria Hewitt wrote:

Thanks to the hard work of Alex Whitworth and the co-authors listed here and contributions from Madeleine J Twyning finally officially published at https://www.life-science-al...

On 2022-08-29 12:05:44, user Manuel Ruedi wrote:

This is a very fine new piece of evidence that the Myotis radiation is both quick... and complexe. I have a single comment regarding the place of M.brandtii within the Old World clade, rather than within the New World (as evidenced elsewhere, incl. in large phylogenies using 1610 UCE to recover that topology): The branch linking brandtii to the few other Old W taxa is very short, so that the root of the whole Myotis tree is very fragile. The authors used distant Vespertilionids to place this root (instead, they could have used Kerivoulinae or Muriniae representatives, i.e. the sister-group of Myotinae, which would have been more effective in placing this root of Myotis). Also because they used only few Old World species, they gave little chance for that group to represent its full diversity.<br /> But what is clear from this brilliant study is that the brandtii lineage appears more basal to the New World radiation than previously reported.

On 2022-08-29 10:14:15, user David Curtis wrote:

I just wanted to point out that I performed a similar study, though on a smaller scale with fewer gene-trait combinations:<br /> Exploration of weighting schemes based on allele frequency and annotation for weighted burden association analysis of complex phenotypes.Curtis D. 2022. Gene. 809:146039. https://doi.org/10.1016/j.g...

I didn't test VARITY but in my analyses REVEL was not an obvious winner and some other predictors performed better on some genes.

On 2022-08-28 23:09:38, user Peter Hickey wrote:

Where can this software be downloaded? I could not find a link in the paper or via Google.

On 2022-08-28 14:12:53, user John Hulleman wrote:

This work is now published in Human Mutation: https://onlinelibrary.wiley...

On 2022-08-28 11:13:55, user David Curtis wrote:

I recently investigated the performance of different predictors of pathogenicity and what I found was that some approaches worked well for some gene/disease combinations but less so for others - there was no universal best method which consistently out-performed the others. Also, the problem with using ClinVar/HGMD for validation is that you may end up only dealing with the kinds of variants that people judge to be pathogenic.

My paper is here:<br /> Exploration of weighting schemes based on allele frequency and annotation for weighted burden association analysis of complex phenotypes. David Curtis. Gene 2022 30;809:146039. doi: 10.1016/j.gene.2021.146039. Epub 2021 Oct 22.<br /> https://pubmed.ncbi.nlm.nih...

On 2022-08-28 09:00:20, user Iratxe Puebla wrote:

Review coordinated via ASAPbio’s crowd preprint review

This review reflects comments and contributions by Ruchika Bajaj and Gary McDowell. Review synthesized by Bianca Melo Trovò.

This study demonstrates the utility of an L-Methionine analog - ProSeMet - to tag and enrich proteins which have residues that are methylated in vivo, ex vivo and in vitro. Furthermore, the study demonstrates that this can be used in combination with mass spectrometry to identify these sites. Overall this is a useful, well-verified and well-described approach that will be helpful for future identification and investigation of methylation sites.

Major comments

It would be helpful if the manuscript could additionally discuss the reversibility of methylation generally, and the reversibility of the modification of protein residues by the alkyne group specifically, in the discussion, and whether that has any implications for their results. It may be that the dynamics of methylation and demethylation vary between the two; or it may be that they are the same - either way, that may affect how they suggest others use this method and interpret its results.

Perhaps related to the question of reversibility, it would be helpful if the manuscript would comment on whether these are “true” methylation sites or not; i.e. whether they consider all these methylation sites to be functional. Trying to determine this would be an interesting direction for future work, but for this study a reflection on whether these novel functional methylation sites are simply capable of being methylated, or are likely to be methylation sites that are meaningful biologically, would be helpful.

Results, ProSeMet competes with L-Met to pseudo methylate protein in the cytoplasm and nucleus: the manuscript claims that ProSeMet is not incorporated into newly synthesized proteins but rather converted to ProSeAM and used by native methyltransferases. There does appear to be some reduction in the labeling with ProSeMet on cycloheximide treatment in Figure 2D - could this suggest that it is incorporated into newly synthesized proteins as well as being converted to ProSeAM? If not, could the manuscript explain why not? This experiment clearly shows that in contrast to AHA labeling, there is still use of ProSeMet as a substrate when translation is inhibited; however, it is not clear how this demonstrates that it is not incorporated at all into newly synthesized proteins. If methyl has been incorporated in previously present proteins, perhaps this can be clarified in the text.

Results, ProSeMet competes with L-Met to pseudomethylate protein in the cytoplasm and nucleus: the conclusion that “Cell fractionation of the cytosolic and nuclear compartments followed by SDS-PAGE fluorescent analysis revealed no fluorescent labeling of the L-Met control” is correct but may be overstated as there appears to be some background in the cytosolic fraction.

Minor comments

Introduction: Recommend including a mention to ProSeMet's permeability.

Introduction, Figure 1: the last step with CuAAC and N3 labeling in the description of the Chemoenzymatic approach for metabolic MTase labeling is not clear. Please, add the description in the legend.

Results, Figure 2D: the image suggests an overloaded gel, consider using an alternative gel image.

Supplementary Material, Fig. S1: the data with L-met is only shown with T47D stacks.

Supplementary Material, Fig. S3: please add the control for the no treatment condition.

Results, Fig. 2A ‘ incubating for 30 m in L-Met free media’: Please confirm that the length of incubation was 30 minutes.

Results, Enrichment of pseudo methylated proteins used to determine breadth of methyl proteome: Please provide some description for the SMARB1-deficient G401 cell line. Why smarb1 deficient?

Results, Figure 3: Please define BP, MF, HP, NES, and label the x and y axes in panel D.

Results, ProSeMet-directed pseudo methylation is detectable in vivo: Please, clarify if the administration was oral.

Comments on reporting

Results, ProSeMet competes with L-Met to pseudo methylate protein in the cytoplasm and nucleus: Please verify the quantity reported: 5µg on SDS-PAGE gel seems low.

Results, ProSeMet-directed pseudo methylation is detectable in vivo: the manuscript reports that “mice starved prior to ProSeMet injection had increased ProSeMet labeling in the heart, whereas mice fed prior to ProSeMet administration had increased labeling in the brain and lungs”. The error bars are large, it would be helpful to show the individual real data points for the graphs in Figure 4.

Results, Figure 4C: please report the mathematical expression used to calculate the relative fluorescence.

Supplementary Material, Fig. S7: please provide more details on the antibody employed.

Suggestions for future studies

Future studies could investigate the biological functionality of the novel methylation sites - but this is a great proof of principle.

On 2022-08-28 06:38:09, user Omri Wurtzel wrote:

The peer-reviewed version of the manuscript is found, open access, in the following link:<br /> https://doi.org/10.15252/em...

On 2022-08-27 15:52:20, user Mark A. Hanson wrote:

The first version of this article was accidentally missing its Acknowledgements section. This has been rectified in v2. To ensure this information is present regardless of manuscript version, we would like to additionally post this information here:

We would like to thank Samuel Rommelaere, Jean-Philippe Boquete, Emi Nagoshi, Lukas Neukomm, Kausik Si, and Anzer Khan for helpful discussion. We would also like to thank Brian McCabe, Mariann Bienz, Barry Ganetzky, Steven Wasserman and Lianne Cohen, the Vienna Drosophila Resource centre, and the Bloomington Drosophila Stock Centre for fly stocks requested over the course of this research. This research was supported by Sinergia grant CRSII5_186397 and Novartis Foundation 532114 awarded to Bruno Lemaitre.

On 2022-08-27 06:56:43, user Iratxe Puebla wrote:

Review coordinated via ASAPbio’s crowd preprint review

This review reflects comments and contributions by Claudia Molina Pelayo, Demetris Arvanitis, Pablo Raneo-Robles, Sónia Gomes Pereira. The comments were synthesized by Vasanthanarayan Murugesan.

In this preprint, Hughes et al. describe the interaction between the ER protein PERK and the mitochondrial protein ATAD3A. During ER stress, PERK phosphorylates elF2a leading to reduced global protein synthesis. The authors show that increased interaction between PERK and ATAD3A during such stress attenuates elF2a phosphorylation locally around mitochondria, resulting in continued translation of mitochondrial protein despite a reduction in global protein translation. The authors present multiple lines of evidence to support this claim and the experiments were well performed. The findings may have important implications for the understanding of mitochondrial protein synthesis and the interactions between mitochondria and the ER.

The following suggestions were raised:

Experiments

The manuscript would benefit greatly by measuring protein translation explicitly showing that mitochondrial protein translation is retained despite a reduction in global protein synthesis under certain conditions. That would help determine whether mitochondrial protein translation is protected under certain conditions driven by ATAD3 expression.

The specificity of ATAD3A towards PERK activation requires further experimental validation. Some specific suggestions are:

• Changes in activation of other pEIF2a kinases, such as GCN2 or PKR, could be measured to discard their involvement.
• In Figure S2, protein levels of ATF6 should accompany changes in spliced XBP1.
• ATF4 levels, a downstream marker of the signaling pathway, could be measured.

Manuscript

Recommend providing more details about the experimental protocol when treating cells with ER stressors. Different treatment durations are found throughout the manuscript (30min, 1h, 8h…). More information would be helpful in understanding the election of those time points for different experiments.

In Figure 2, recommend including the blots for the downstream targets ATF4, GADD34 and CHOP at the 30 minutes time point, where the upstream activation starts.

In Figure 2, the differences shown in the representative images for p-eIF2a and ATF4 appear milder than what is shown in the graph. In particular when compared with the interpretation of blots in Fig. S2. It is suggested to include all the blots used for quantification in Figure 2 in a supplemental figure so it can be clear how overexpressing/downregulating ATAD3A has a meaningful effect on this signaling pathway.

Figure 2B shows 5 different (phospho)proteins using the same loading control blot. This approach would require stripping of the membrane after each blotting, can this be specified in figure legends and in the Materials & Methods. Was the membrane stripped after each blot or were different membranes used? If different membranes were used, please indicate so and present the individual beta-actin blots corresponding to each protein as a supplemental figure.

In Figure 3A, arrows indicating the contact sites between ER and the mitochondria would be helpful in highlighting the colocalization of the two proteins. Please also provide scale bars for the images.

In Figure 3D, the #contacts per mitochondria, it is important to specify the area of images analyzed. It is unclear that n=45 images from 3 separate experiments refers to 45 images per experiment or a total of 45 images pooled from 3 experiments. Please clarify.

Recommend discussing the limitation of experiments using a single siRNA for loss-of-functions studies and experiments using cell culture.

On 2022-08-26 18:08:12, user Guido Lenz wrote:

Nice work - how does this relate to the reduction in variance of colonies with the growth of these colonies? https://doi.org/10.1158/000...

On 2022-08-26 14:15:53, user Anthony Gitter wrote:

The manuscript refers to a Methods section, which is not included in the full text. Could the authors please update their manuscript to include the Methods?

On 2022-08-26 13:40:17, user Matt Higgins wrote:

We were extremely interested to see these impressive structures of a chimeric Sec translocon, held in a state ready for post-translational translocation through interaction with Sec62/63, bound to eight different inhibitors. Particularly noteworthy to us is that the inhibitor mycolactone binds in a different location in this study when compared with our previous structure of the same inhibitor bound to a ribosome-bound translocon, primed for co-translational translocation (Gerard et al Molecular Cell 79 406-15).

The authors speculate that “Our data suggest that the density feature previously assigned as mycolactone is unlikely to be mycolactone.” While it is true that our previous structure has a resolution of ~5A in the region of the map attributed to mycolactone, and therefore also true that we cannot unambiguously place mycolactone in this density, we remain confident that this density is mycolactone for the following reasons:

(i) Our procedure involved incubation of microsomes with mycolactone at a concentration of ~0.3µM (compared with 100µM used by Itskanov) before detergent/digitonin treatment and purification of ribosome-associated Sec complexes. A similar sample was prepared without mycolactone. When these two protein complexes were studied by cryo-electron microscopy, the Sec translocon adopted a substantially different conformation when mycolactone-bound compared with free. The only difference between these two samples was the presence of mycolactone, indicating that this structural difference is due to mycolactone binding.

(ii) We confirmed the presence of mycolactone in our mycolactone-bound purified using mass spectrometry of a sample taken immediately before addition to grids for structural analysis.

(iii) Analysis of the electron density for the mycolactone-bound translocon did not reveal any density feature in the mycolactone-bound sample in the location of the binding site observed by Itskanov. Therefore ribosome-bound, mycolactone-bound translocon is different from Sec62/63-bound, mycolactone-bound translocon.

(iv) The only additional density feature observed in the ribosome-bound, mycolactone-bound translocon is that which we have attributed to mycolactone and molecular dynamics simulations confirm that mycolactone is stable in this binding site.

It is therefore our view that we did not misattribute the electron density into which we have placed mycolactone. Instead, it is our view that the difference between these two structures is likely to be genuine and mechanistically interesting.

There are possible technical differences which could account for the different binding sites observed when comparing our structure with that of Itskanov:

• While we added mycolactone to the Sec translocon while still in the native membrane environment of microsomes, and then extracted the mycolactone-bound complex, Itskanov added to mycolactone to translocon after its purification and integration into a non-lipid peptidisc. Our model for how mycolactone reaches its binding site in our system relies on translocon “breathing” within the physiological situation of a lipid bilayer, and mycolactone itself being present in this bilayer. It is not known if the translocon within a peptidisc is able to undertake similar “breathing”, nor how highly hydrophobic mycolactone may interact with this material.

• While we used native canine microsomes, Itzkanov et al used a hybrid translocon, comprised of human transmembrane regions and yeast extracellular regions. It is not known if this hybrid translocon is functional for translocation, or whether the translocation of model substrates by it is inhibited by mycolactone.

• There is also a large difference in mycolactone concentration used in the different studies. Mycolactone is effective at sub-nM concentrations on live cells. To provide sufficient molar ratios of mycolactone in concentrated microsomes, we used ~300nM in our studies, while Itskanov used the much higher concentration of 100µM mycolactone. It would be interesting to know whether this was the minimal concentration required for them to see binding, indicating a lower affinity binding site, or was simply the concentration selected.

While there are technical differences which might account for the different binding sites observed, there is also the far more interesting possibility that both studies have correctly identified binding sites for mycolactone and that this inhibitor acts differently in post-translational and co-translational translocation.

The Sec translocon can act through either a post-translational (involving Sec62/63) or a co-translational (involving ribosomes) mechanism. McKenna, Simmonds and High have previously shown (PMID 26869228) that mycolactone-mediated blockade is different in these two systems. While mycolactone shows a broad effect, preventing co-translational translocation of a wide range of substrates, it has a more restricted effect during post-translational translocation, only affecting translocation of a subset of substrates. Together with the differences in mycolactone binding between these two structures, this suggests the intriguing possibility mycolactone might have two different binding sites; perhaps one site which occurs during co-translational translocation where mycolactone is stably wedged into the cytosolic side of the lateral gate (Gerard et al), and one site which operates in post-translational translocation and is more easily overcome by signal peptide binding (Itskanov et al). Future studies will be required to test this intriguing possibility.

Sam Gerard, Matt Higgins and Rachel Simmonds