The VarA-CsrA regulatory pathway influences cell shape in Vibrio cholerae

Article activity feed

1. Version published to 10.1371/journal.pgen.1010143

   Mar 28, 2022
2. 

   Review Commons

   Dec 14, 2021

   Note: This rebuttal was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Reply to the reviewers

   Description of the planned revisions

   From Review Commons: Please find below our point-by-point reply that explains what revisions, additional experimentations and analyses are planned to address the points raised by the referees

   **************

   Reviewer #1 (Evidence, reproducibility and clarity (Required)):

   * *

   Comment #1:* In this manuscript, the authors follow up on an interesting finding that varA null mutants of V. cholerae form spherical cells in stationary phase. The authors determine that this cell rounding is due to weakening of the cell wall via less production of tetrapeptide cross links. Mutation of the regulator csrA and the enzyme aspA lead to a …

   Note: This rebuttal was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Reply to the reviewers

   Description of the planned revisions

   From Review Commons: Please find below our point-by-point reply that explains what revisions, additional experimentations and analyses are planned to address the points raised by the referees

   **************

   Reviewer #1 (Evidence, reproducibility and clarity (Required)):

   * *

   Comment #1:* In this manuscript, the authors follow up on an interesting finding that varA null mutants of V. cholerae form spherical cells in stationary phase. The authors determine that this cell rounding is due to weakening of the cell wall via less production of tetrapeptide cross links. Mutation of the regulator csrA and the enzyme aspA lead to a model in which a varA mutant cell lacks aspartate leading to low cross-linked cell wall that is unable to hold the typical curved V. cholerae shape. The data are robust, and the manuscript is clearly written.*

   __Authors’ reply #1: __We very much appreciate the reviewer’s accurate summary and the appraisal of the robust data and the clearly written manuscript.

   Comment #2:* I think the finding is quite interesting, even though it is not clear to me if this observed cell morphology has a biological function or if it is an artifact of completly removing VarA. However, this manuscript builds the foundation to further test this question. *

   __Authors’ reply #2: __We agree with the reviewer. However, it is worth mentioning that this two-component system (TCS) has been first described in 1998 yet the input signal (or repression of the signal under certain conditions) still remains elusive. Maybe, this isn’t too surprising, given that studies on V. cholerae are strongly biased towards its pathogenic lifestyle, while the varA/varS system is highly conserved among Gram-negative bacteria including non-pathogenic environmental V. cholerae strains. These strains can live under very diverse conditions of slow or fast growth, including long starvation periods. Unfortunately, we still lack significant insight into this part of the V. choleraebiology. We therefore believe that the current study is very important, as the elucidation of the molecular mechanism of the observed shape transition within the varA mutant will foster fresh hypotheses on the role that the system plays in V. cholerae and what signals might be sensed.

   We would also like to remind the editor and reviewer(s) that a plethora of studies have been published based on varA and varS (and csrA) deletion mutants of V. cholerae with various readouts ranging from transcriptomics to quorum sensing defects, impairment of virulence, etc. Thus, the argument that the complete removal of varA might cause an artifact seems equally valid for previous work by others and, maybe, even the vast majority of studies in which TCS are investigated for which the sensed signal has yet to be identified.

   In conclusion, we propose to address this valid point of critique in the revised manuscript by clearly stating the caveat of the gene deletion(s). However, as the reviewer correctly stated, “this manuscript builds the foundation to further test this question.”

   Comment #3:* The data all support the conclusions, but I do think the authors could have really confirmed their model by connecting mutations in csrA and aspA to restoration of high cross-linked cell well similar to the WT strain as done in Fig. 2. As it stands, this is still somewhat hypothetical and has not been directly demonstrated, although I do think their model is correct and these experiments will be conformation of it. *

   __Authors’ reply #3: __We thank the reviewer for their comment and the assumption that our model might be correct. It is very unlikely that the csrA suppressor mutant(s) or the ∆varA∆aspA mutant maintain the low level of cross-links and the high level of dipeptides that we observed for the ∆varA mutant. Indeed, it would be unclear how the cells could restore the Vibrio shape that we visualized in the phase contrast image under such conditions. However, as this point seems very important to the reviewer (see also comment #7 below), we will perform the suggested cell wall analysis of these mutants and include the new data in the revised manuscript.

   Comment #4:* I also have a few other suggestions to improve the manuscript, but in sum I think it is a well-done research study that will be interesting to research in V. cholerae and other gamma proteobacteria.*

   __Authors’ reply #4: __Once again, we thank the reviewer for their kind words.

   Major comments:

   Comment #5:* 1. The enrichment for suppressors is very creative and connected the varA impact on cell morphology to misregulation of csrA as 10/10 mutants were ultimately linked to this gene. However, insertion in aspA should also suppress this phenotype, and I am curious why this gene was not identified in the transposon suppressor screen.*

   __Authors’ reply #5: __This is a very relevant and important comment. The reason why we did not isolate ∆varA-aspA::Tn mutants is most likely due to a growth defect that we observed for the double ∆varA∆aspA mutant compared to the ∆varA-csrA suppressor mutant(s). In the figure on the right, respective growth curves are shown [∆varA∆aspA in orange and the ∆varA-csrA suppressor mutant ∆varA-Tn A in gray]. Any ∆varA-aspA::Tn mutant is therefore expected to be outcompeted by the ∆varA-csrA suppressor mutants during the enrichment process. We will include this information and the corresponding data (e.g., final growth curves after 3 biologically independent experiments) in the revised manuscript.

   [figure not shown in online form]

   Comment #6:* 2. The authors should complement at least one of their varA/csrA mutants with csrA. *

   __Authors’ reply #6: __Agreed. We are in the process of performing the suggested experiment and will include the results in the revised manuscript.

   Comment #7:* 3. The changes in cell wall structure are not directly connected to the genetic identification of csrA and aspA. Yes, I agree their model makes sense, but to really nail it down they should analyze the cell wall composition in the varA/csrA and varA/aspA double mutants and show it returns to WT levels of crosslinking. *

   __Authors’ reply #7: __As mentioned above, we will perform those cell wall analyses (see also authors’ reply #3 above), as requested.

   Comment #8:* 4. Does deletion of aspA in the WT or varA mutant impact the growth rate? *

   __Authors’ reply #8: __This is indeed the case in the ∆varA background but not in the WT background (as shown under authors reply #5). These data will be included in the revised manuscript.

   *Minor comments *

   Comment #9:* 5. Are the round cells able to divide? The data in Fig. S2 would suggest they can based on the increase in CFUs from hour 6 to hour 8, but the authors never comment on this point and it might be worth addressing in the discussion.*

   __Authors’ reply #9: __We never observed diving round cells. Indeed, the increase of the optical density and CFUs from 6h to 8h is most likely based on those bacteria that have not yet changed their cellular morphology and therefore keep dividing (see below the imaging/quantification for this timeframe taken from Fig. 2). What we observed though is that upon dilution into fresh medium, the round cells start to elongate and then divide resulting in newborn Vibrio shaped cells. We will include these new data in the revised manuscript.

   [figure not shown in online form]

   Comment #10:* 6. Fig. S2-Why does the OD600 increase from 8 to 24 hours but the CFUs decrease in the varA mutant? *

   __Authors’ reply #10: __This is an interesting observation that might reflect the presence of dead but not yet lysed cells in these cultures. Indeed, while it looks as if the OD600 values are still increasing for the ∆varA mutant at 24h, we cannot exclude at this point that the OD600 values increased during the 16h-time interval and went again down at 24h (e.g., like shifting the WT peak to later time points/the right of the X-axis). Notably, the purpose of this figure was mostly to i) indicate the slower growth of the ∆varA mutant while ii) emphasizing that late during growth (e.g., 24h here) the strain can still reach similar OD600 values as well as CFUs/ml as the WT strain. We will change Fig. S2 to better emphasize these two points in the revised manuscript.

   Comment #11:* 7. Lines 309-A little bit more detail here would help the reader. Are the authors examining whole cell lysates or lysates from specific cellular components? I am actually very surprised this worked as there are so many proteins in crude cell lysates. *

   __Authors’ reply #11: __Indeed, these are whole cell lysates, which were prepared as described in the methods section (lines 482 onwards; “SDS-PAGE, Western blotting and Coomassie blue staining”). We fully agree that there are many proteins in the crude cell lysates and realized that we might not have explained well enough that only the gel region containing the overproduced band in the ∆varA strain and the same location in the WT sample were analyzed by mass spectrometry (even though we were referring to the “gel pieces” in line 498 onwards). Please accept our sincerest apologies for this neglect. During the revision, we will ensure that this information is explicitly stated and that these details are included in the main text and the methods section.

   * *

   Comment #12:* 8. Lines 320-321-I don't think there is evidence that CsrA enhances aspA RNA translation, merely that CsrA enhances AspA protein production. It is likely through increasing translation, but this cannot be concluded without direct evidence. *

   __Authors’ reply #12: __We fully agree and thank the reviewer for this important comment. Indeed, we meanwhile know that the aspA mRNA levels also increase in the ∆varA mutant strain (which might or might not be linked to enhanced translation). We will add these transcript level data to the revised manuscript and discuss all possibility that could explain the AspA overproduction.

   Comment #13:* 9. Line 348-350-I do not understand the logic of this sentence stating that the "..until now, the signal that abrogates VarA phosphorylation..." as this manuscript does not contribute to our understanding of the VarS signal. *

   • *__Authors’ reply #13: __We apologize that this sentence or the logic behind it wasn’t clear. As this is a combined result and discussion section, the aim of the sentence was to put the observed shape transition of the bacteria into a broader context, which required us to mentioned that the input signal is still unknown. We will make sure that this becomes more obvious in the revised manuscript by rephrasing this sentence.

   Comment #14:* 10. I am curious if the total volume of the round versus curved cells is constant at 20 hours. This should be easy to determine using ImageJ and worth reporting.*

   __Authors’ reply #14: __We are not entirely sure how this question is relevant for the study (e.g., for this report on the observed shape transition phenotype it doesn’t matter if the cells maintain the same volume or not). However, given the importance for the reviewer, we will perform these volume measurements on our images and add a sentence to the revised manuscript on the analysis’ outcome (plus include the data as a supplementary panel).

   Reviewer #1 (Significance (Required)):

   *Comment #15: **Understanding changes to cell morphology and their biological implications is a growing area of microbiology. This study makes a new contribution to this area by demonstrating a round, spherical form of V. cholerae that is driven by alterations to the cell that decrease cell-wall cross linking. *

   __Authors’ reply #15: __Once again, we thank the reviewer for this summary and for placing our study into context. We agree that cell morphological changes and the underlying molecular mechanism(s) are an exciting and growing area of microbiology.

   Reviewer #2 (Evidence, reproducibility and clarity (Required)):

   __*Comment #16: *__In this manuscript, Rocha et al studied the effect of the VarA response regulator on cell shape of Vibrio cholerae. VarA is part of a two-component system that also includes the histidine kinase VarS. It has previously been shown that VarA activates the expression of three redundantly acting regulatory RNAs called CsrB, CsrC, and CsrD. All three Csr RNAs share the same regulatory principle, which is to sequester the activity of the RNA-binding protein CsrA. CsrA in turn can bind hundreds of mRNA species in the cell, which in the majority of cases results in reduced translation of these mRNAs (in addition to various other modes of action that have been reported). Here, the authors discovered that deletion of the varA gene results in an abnormal, spherical cell shape in stationary phase grown V. cholerae cells. Biochemical analysis revealed an unusual peptidoglycan composition in varA-deficient cells, showing increased levels of dipeptides, a reduction of tetrapeptides, and an overall decrease in peptide-cross linkage. Interestingly, the varA phenotype was complemented by the addition of conditioned medium from wild-type cells, which are likely to provide peptidoglycan building blocks in trans. The authors further discover that varA-deficiency results in AspA over-production, which could be linked to the activity of CsrA. The authors speculate that high AspA levels deplete the cell of aspartate, which is required to produce peptidoglycan precursors. The manuscript is interesting, well-written, and the rationale of the experiments is easy to follow.

   __Authors’ reply #16: __We thank the reviewer for this excellent summary and the kind words on the quality of the manuscript.

   __*Comment #17: *__However, I have two major points of criticism, which reduce my enthusiasm for this work. First, the molecular pathways that links varA-deficiency to increased AspA levels is incomplete: please clarify how CsrA activates AspA levels and if this phenotype is linked to direct binding of CsrA to the aspA mRNA and if so how is activation is achieved at the molecular level.

   __Authors’ reply #17: __We thank the reviewer for the comment. However, we never had the intention to decipher the entire pathway and it is indeed possible that intermediate regulators might be involved. Notably, the first part of the signaling pathway (VarA -> CsrB,C,D -> CsrA) seemed well established in the literature and we truly believe that our work supports this part of the pathway (given the numerous csrA suppressor mutants that we obtained in the varA-minus background). For the link between CsrA and AspA, we indeed do not provide direct evidence. Nonetheless, we discuss recent work in Salmonella by mentioning “Interestingly, previous studies identified the aspA mRNA amongst hundreds of direct CsrA targets in Salmonella using the CLIP-seq technique to identify protein-RNA interactions (32)”. Notably, this finding by Holmqvist et al. (2016, EMBO J.) has been reproduced for E. coli by Potts et al. (2017, *Nat. Commun.; see Supplementary Data file 1, CsrA CLIP-seq data; with three highly significant peaks corresponding to aspA mRNA binding), *an information that we will add to the revised manuscript. Of course, neither Salmonella nor E. coli belongs to the same genus as V. cholerae (though, of course, all are gamma-Proteobacteria). Thus, to accommodate the reviewer’s comment, we will revise the manuscript to include the caveat that direct CsrA binding of the aspA mRNA has been shown in both *Salmonella and E. coli *but that it is still feasible that intermediate regulatory proteins might be involved in the case of V. cholerae. We will also revise the model to show such potential intermediate steps between CsrA and AspA.

   __*Comment #18: *__Second, I am not convinced about the biological relevance of the findings. The authors speculate in the discussion section that the VarA-pathway could modulate cell shape under physiological conditions, however, I am not sure such conditions exist given that VarA activity is not only controlled by VarS, but rather integrates information from multiple histidine kinases. I have a several additional comments, which I listed below.

   __Authors’ reply #18: __We regret the referee’s personal opinion that our findings might not be of biological relevance.

   However, we respectfully disagree with the notion that such physiological conditions would never occur just because several histidine kinases can feed into VarA signaling. Indeed, as discussed above under authors’ reply #2, the (VarS/)VarA-CsrA pathway is highly conserved in Vibrio species and other proteobacteria. Yet, for V. choleraemost studies have focused on virulence-inducing conditions, while the species’ environmental lifestyle has been vastly neglected in the past. Indeed, even our own work on several V. cholerae’s phenotypes (natural competence for transformation [Meibom*, Blokesch* et al., 2005, Science]; T6SS production in pandemic strains [Borgeaud et al., 2015, Science]; pilus-mediated aggregation [Adams et al., 2019, Nat. Microbiol]; etc.) has remained unknown for decades, given the chitin dependency for their induction – a substrate not commonly studied in lab settings. Interestingly, several of these findings have also initially been considered biologically irrelevant, “artifacts”, or even non-reproducible by reviewers in the past, while nowadays all these phenotypes have been extensively reproduced by many different research groups and are well accepted in the field as biologically highly relevant. Thus, we truly believe that one should be open to new phenotypes and, as reviewer #1 rightfully acknowledged, consider that “this manuscript builds the foundation to further test this question.”

   It should also be noted that the (VarS/)VarA-CsrA system has been studied for >15 years based on deletion strains, as we did in this study, and the readouts of these studies have been well accepted in the field without provision of the physiological conditions that would mimic the situation of these knock-out strains.

   Collectively, we truly believe that there are still many understudied physiological conditions for V. cholerae; however, finding the right conditions could take years and is therefore beyond the scope of the current study.

   ---

   Major points:

   *Comment #19: *- Figs. 1 and S1: I think it is interesting that the varS mutant strain does not share the cell shape phenotype with the varA mutant. As pointed out by the authors, this result indicates that varA activity is controlled by another histidine kinase. While I believe it might be beyond the scope of this manuscript to determine which other histidine kinases signal towards VarA, I think it would be useful to measure and compare CsrB/C/D levels in WT, DvarA, and DvarS cells.

   __Authors’ reply #19: __Thanks for this comment. We fully agree that finding the secondary histidine kinase is beyond the scope of this study. In the revised manuscript, we will, however, include the CsrB/C/D levels of the WT, ∆varA, and ∆varS strains, as suggested by the reviewer.

   *Comment #20: *- Figs. 1, S1, and 4C: The regulatory logic implied by these results suggest that deletion of varA results in reduced CsrB/C/D levels, which in turn leads to higher activity of CsrA in the cell. Thus, it would be useful to test if A) over-production of CsrB, CsrC, or CsrC can rescue the phenotype of an varA mutant and if B) combined deletion of csrB/C/D will phenocopy the mutation of varA.

   Authors’ reply #20: These are also very good suggestions. Notably, this has been done in the past for the V. cholerae system (Lenz et al., 2005, Mol. Microbiol.). Indeed, after receiving the reviewer’s comment, we immediately asked these authors to kindly share their csrB,C,D overproduction plasmids as well as the triple knock-out strain with us (as all of these constructs have been extensively verified in their published work). Unfortunately, we are not entirely sure whether it will be possible to receive these constructs any time soon, as we were told that such shipment might take >1 year (though, upon further discussion, this timeframe was lowered to ~3 months). If we manage to receive these published constructs in a reasonable timeframe, we will certainly perform the suggested experiments.

   * *

   *Comment #21: *- Figs. 4B & 5C: I was somewhat surprised by these results. Given that AspA overproduction is suggested to cause cell shape abnormalities in the varA mutant, I would have expected additional transposon insertion in aspA. The fact that mutations only occurred in csrA could indicate that additional (CsrA-controlled) could be involved in the phenotype.

   __Authors’ reply #21: __See authors’ reply #5 above, where we explain that the ∆varA∆aspA strain has a slight growth disadvantage. For this reason, any ∆varA-aspA::Tn transposon mutant would likely be outcompeted by the csrAsuppressor mutants in our genetic screen.

   ---

   __Minor points: __

   *Comment #22: *- Throughout study: italicize gene names

   __Authors’ reply #22: __Gene names have been italicized in the initial manuscript; however, strain names - such as strain ∆varA – haven’t been italicized, in accordance with several of our previous publications. However, for the revision, we will italicize all strain names to accommodate the reviewer’s request.

   *Comment #23: *- Figs. 1C and S3C: please quantify the results of these western blots and indicate how many replicates were performed.

   __Authors’ reply #23: __We apologize for this oversight – indeed, all Western Blot were performed three independent times, as is good scientific practice. We will add this information into the methods section of the revised manuscript.

   Concerning the quantification: the primary claim of these figures is that the HapR protein is still produced in the ∆varA mutant in the different pandemic strain backgrounds, while the luxO-mutated strains have a significant defect in HapR production (as we have previously reported; Stutzmann and Blokesch, 2016, mSphere). These data are qualitatively very clear in the Western Blots and can be considered as “black or white” results.

   However, for the revision we will quantify the bands’ intensities of the performed Western blots and provide these quantitative data, as requested by the reviewer.

   *Comment #24: *- Fig. 5A and B: in order to properly quantify the levels of AspA in the cell (and link them to CsrA activity in the transposon mutants), I think it would be better to add a tag to the chromosomal aspA gene and perform quantitative Western blot analysis.

   __Authors’ reply #24: __We respectfully disagree. Firstly, this is not a subtle difference that we observe in these cell lysates/the corresponding stained gel bands but a rather strong difference when WT is compared to the mutants (see, for instance, a copy of panel 5B below as a kind reminder). Together with the genetic experiments that follow afterwards, the link seems very solid to us. Secondly, adding a tag could change the proteins abundance (change of the protein’s production/degradation dynamics) and/or activity, which could cause more confusion than needed (and a loss of the spherical cell shape if the enzyme loses its activity through the tagging).

   However, as mentioned above under authors’ reply#12, we meanwhile observed that the aspA mRNA levels also increase in the ∆varA mutant. Thus, we will provide qRT-PCR data in the revised manuscript (and discuss all options on how the increase of the transcript and subsequently the protein might be caused, as mentioned above under authors’ reply #12), which we truly believe will fulfill the reviewer’s request for quantification.

   [figure not shown in online form]

   Reviewer #2 (Significance (Required)):

   __*Comment #25: *__I think this manuscript starts with an interesting observation, which is that varA mutant cells of V. cholerae display an aberrant cell shape. The manuscript also provides several important findings explaining the molecular basis of this phenotype.

   __Authors’ reply #25: __Once again, we thank the reviewer for the kind words.

   __*Comment #26: *__However, as pointed out in my report, I think the manuscript is yet incomplete in connecting this information to identify the underlying regulatory mechanism.

   __Authors’ reply #26: __As mentioned above, the focus of this study was never on the elucidation of the entire regulatory pathway. Instead, we aimed at deciphering the molecular mechanism behind an observed phenotype - that is, the cell wall modification in the varA-deficient strain that leads to the bacterium’s spherical shape, which can be restored to the WT Vibrio shape by peptidoglycan precursor cross-feeding from neighboring cells – followed by the identification of several regulators and enzymes that trigger these phenotypes. Overall, we consider this a very complete study. However, as mentioned above, we will certainly discuss in the revised manuscript that the step between CsrA and AspA could be indirect in V. cholerae, in contrast to what was experimentally shown for *Salmonella *and E. coli.

   **Referee Cross-commenting**

   __*Comment #27: *__As pointed out in my review, I think this manuscript is well written and easy to follow. However, I agree with reviewer #1 that the underlying phenotype is most likely an artifact, which limits the biological relevance of this study. In addition, I am missing the molecular mechanism that connects CsrA with AspA production in V. cholerae.

   __Authors’ reply #27: __See authors’ reply #18 above. We disagree that there is any strong indication that the observed phenotype is an artifact. Given that it is state-of-the-art to study TCS by deleting their genes, our study isn’t any more prone to being an artifact than any other study on TCSs.

   We truly believe that it is also important to not take reviewer #1’s comment out of context by stating “I agree with reviewer #1 that the underlying phenotype is most likely an artifact”. Indeed, he/she provided a rather encouraging statement in which he/she mentions the possibility of an artifact but also clearly states that this study is interesting and builds the foundation to further investigate the newly observed phenotype(s): “I think the finding is quite interesting, even though it is not clear__ to me__ if this observed cell morphology has a biological function or if it is an artifact of completly removing VarA. However, this manuscript builds the foundation to further test this question.”

   Moreover, whether there is a direct (as in *Salmonella *and E. coli) or indirect connection between CsrA and AspA production is not a key aspect of the current study, as discussed above.

   Read the original source
3. 

   Review Commons

   Dec 14, 2021

   Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Referee #2

   Evidence, reproducibility and clarity

   In this manuscript, Rocha et al studied the effect of the VarA response regulator on cell shape of Vibrio cholerae. VarA is part of a two-component system that also includes the histidine kinase VarS. It has previously been shown that VarA activates the expression of three redundantly acting regulatory RNAs called CsrB, CsrC, and CsrD. All three Csr RNAs share the same regulatory principle, which is to sequester the activity of the RNA-binding protein CsrA. CsrA in turn can bind hundreds of mRNA species in the cell, which in the majority of cases results in reduced translation of these mRNAs (in addition to various other modes of …

   Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Referee #2

   Evidence, reproducibility and clarity

   In this manuscript, Rocha et al studied the effect of the VarA response regulator on cell shape of Vibrio cholerae. VarA is part of a two-component system that also includes the histidine kinase VarS. It has previously been shown that VarA activates the expression of three redundantly acting regulatory RNAs called CsrB, CsrC, and CsrD. All three Csr RNAs share the same regulatory principle, which is to sequester the activity of the RNA-binding protein CsrA. CsrA in turn can bind hundreds of mRNA species in the cell, which in the majority of cases results in reduced translation of these mRNAs (in addition to various other modes of action that have been reported). Here, the authors discovered that deletion of the varA gene results in an abnormal, spherical cell shape in stationary phase grown V. cholerae cells. Biochemical analysis revealed an unusual peptidoglycan composition in varA-deficient cells, showing increased levels of dipeptides, a reduction of tetrapeptides, and an overall decrease in peptide-cross linkage. Interestingly, the varA phenotype was complemented by the addition of conditioned medium from wild-type cells, which are likely to provide peptidoglycan building blocks in trans. The authors further discover that varA-deficiency results in AspA over-production, which could be linked to the activity of CsrA. The authors speculate that high AspA levels deplete the cell of aspartate, which is required to produce peptidoglycan precursors. The manuscript is interesting, well-written, and the rationale of the experiments is easy to follow. However, I have two major points of criticism, which reduce my enthusiasm for this work. First, the molecular pathways that links varA-deficiency to increased AspA levels is incomplete: please clarify how CsrA activates AspA levels and if this phenotype is linked to direct binding of CsrA to the aspA mRNA and if so how is activation is achieved at the molecular level. Second, I am not convinced about the biological relevance of the findings. The authors speculate in the discussion section that the VarA-pathway could modulate cell shape under physiological conditions, however, I am not sure such conditions exist given that VarA activity is not only controlled by VarS, but rather integrates information from multiple histidine kinases. I have a several additional comments, which I listed below.

   Major points:

   • Figs. 1 and S1: I think it is interesting that the varS mutant strain does not share the cell shape phenotype with the varA mutant. As pointed out by the authors, this result indicates that varA activity is controlled by another histidine kinase. While I believe it might be beyond the scope of this manuscript to determine which other histidine kinases signal towards VarA, I think it would be useful to measure and compare CsrB/C/D levels in WT, DvarA, and DvarS cells.
   • Figs. 1, S1, and 4C: The regulatory logic implied by these results suggest that deletion of varA results in reduced CsrB/C/D levels, which in turn leads to higher activity of CsrA in the cell. Thus, it would be useful to test if A) over-production of CsrB, CsrC, or CsrC can rescue the phenotype of an varA mutant and if B) combined deletion of csrB/C/D will phenocopy the mutation of varA.
   • Figs. 4B & 5C: I was somewhat surprised by these results. Given that AspA overproduction is suggested to cause cell shape abnormalities in the varA mutant, I would have expected additional transposon insertion in aspA. The fact that mutations only occurred in csrA could indicate that additional (CsrA-controlled) could be involved in the phenotype.

   Minor points:

   • Throughout study: italicize gene name
   • Figs. 1C and S3C: please quantify the results of these western blots and indicate how many replicates were performed.
   • Fig. 5A and B: in order to properly quantify the levels of AspA in the cell (and link them to CsrA activity in the transposon mutants), I think it would be better to add a tag to the chromosomal aspA gene and perform quantitative Western blot analysis.

   Significance

   I think this manuscript starts with an interesting observation, which is that varA mutant cells of V. cholerae display an aberrant cell shape. The manuscript also provides several important findings explaining the molecular basis of this phenotype. However, as pointed out in my report, I think the manuscript is yet incomplete in connecting this information to identify the underlying regulatory mechanism.

   Referee Cross-commenting

   As pointed out in my review, I think this manuscript is well written and easy to follow. However, I agree with reviewer #1 that the underlying phenotype is most likely an artifact, which limits the biological relevance of this study. In addition, I am missing the molecular mechanism that connects CsrA with AspA production in V. cholerae.

   Read the original source
4. 

   Review Commons

   Dec 14, 2021

   Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Referee #1

   Evidence, reproducibility and clarity

   In this manuscript, the authors follow up on an interesting finding that varA null mutants of V. cholerae form spherical cells in stationary phase. The authors determine that this cell rounding is due to weakening of the cell wall via less production of tetrapeptide cross links. Mutation of the regulator csrA and the enzyme aspA lead to a model in which a varA mutant cell lacks aspartate leading to low cross-linked cell wall that is unable to hold the typical curved V. cholerae shape. The data are robust, and the manuscript is clearly written. I think the finding is quite interesting, even though it is not clear to me if this …

   Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Referee #1

   Evidence, reproducibility and clarity

   In this manuscript, the authors follow up on an interesting finding that varA null mutants of V. cholerae form spherical cells in stationary phase. The authors determine that this cell rounding is due to weakening of the cell wall via less production of tetrapeptide cross links. Mutation of the regulator csrA and the enzyme aspA lead to a model in which a varA mutant cell lacks aspartate leading to low cross-linked cell wall that is unable to hold the typical curved V. cholerae shape. The data are robust, and the manuscript is clearly written. I think the finding is quite interesting, even though it is not clear to me if this observed cell morphology has a biological function or if it is an artifact of completly removing VarA. However, this manuscript builds the foundation to further test this question. The data all support the conclusions, but I do think the authors could have really confirmed their model by connecting mutations in csrA and aspA to restoration of high cross-linked cell well similar to the WT strain as done in Fig. 2. As it stands, this is still somewhat hypothetical and has not been directly demonstrated, although I do think their model is correct and these experiments will be conformation of it. I also have a few other suggestions to improve the manuscript, but in sum I think it is a well-done research study that will be interesting to research in V. cholerae and other gamma proteobacteria.

   Major comments:

   1. The enrichment for suppressors is very creative and connected the varA impact on cell morphology to misregulation of csrA as 10/10 mutants were ultimately linked to this gene. However, insertion in aspA should also suppress this phenotype, and I am curious why this gene was not identified in the transposon suppressor screen.
   2. The authors should complement at least one of their varA/csrA mutants with csrA.
   3. The changes in cell wall structure are not directly connected to the genetic identification of csrA and aspA. Yes, I agree their model makes sense, but to really nail it down they should analyze the cell wall composition in the varA/csrA and varA/aspA double mutants and show it returns to WT levels of crosslinking.
   4. Does deletion of aspA in the WT or varA mutant impact the growth rate?

   Minor comments

   1. Are the round cells able to divide? The data in Fig. S2 would suggest they can based on the increase in CFUs from hour 6 to hour 8, but the authors never comment on this point and it might be worth addressing in the discussion.
   2. Fig. S2-Why does the OD600 increase from 8 to 24 hours but the CFUs decrease in the varA mutant?
   3. Lines 309-A little bit more detail here would help the reader. Are the authors examining whole cell lysates or lysates from specific cellular components? I am actually very surprised this worked as there are so many proteins in crude cell lysates.
   4. Lines 320-321-I don't think there is evidence that CsrA enhances aspA RNA translation, merely that CsrA enhances AspA protein production. It is likely through increasing translation, but this cannot be concluded without direct evidence.
   5. Line 348-350-I do not understand the logic of this sentence stating that the "..until now, the signal that abrogates VarA phosphorylation..." as this manuscript does not contribute to our understanding of the VarS signal.
   6. I am curious if the total volume of the round versus curved cells is constant at 20 hours. This should be easy to determine using ImageJ and worth reporting.

   Significance

   Understanding changes to cell morphology and their biological implications is a growing area of microbiology. This study makes a new contribution to this area by demonstrating a round, spherical form of V. cholerae that is driven by alterations to the cell that decrease cell-wall cross linking.

   Read the original source
5. 

   Version published to 10.1101/2021.09.09.459595 on bioRxiv

   Sep 9, 2021