Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

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

Reviewer #1:

Major comments:

1. • The relevance of these findings to human biology remains unclear. In Figures 1-4, the authors present data showing that AATBC is enriched in thermogenic fat, and they argue that it regulates thermogenesis and mitochondrial biology. However, in Figures 6-7, where the authors look at AATBC in different human cohorts, they actually find that it is enriched in visceral fat, which is thought of as being the least thermogenic fat depot. The authors do not explain this seeming paradox, and thus, the role of AATBC in fat remains uncertain. *

RESPONSE: We thank the reviewer for this comment and have clarified the discussion to address this point. It has been recently shown (PMID: 28529941) that the pattern of browning genes in human white adipose tissue depots is actually inverted to mice, making visceral adipose tissue in humans actually more thermogenic than subcutaneous. This aligns well with our findings of AATBC is predominantly expressed in thermogenic adipose tissue.

• In many of the experiments, insufficient controls are provided, or the data are not at all convincing. For example:*

(a) The first four figures rely on in vitro adipocyte models, but the authors do not present data to show these cells differentiate properly and equally. This is especially relevant for the gain and loss of function studies.

RESPONSE: We agree with the reviewer that equal differentiation is necessary for in vitro adipocyte models. Therefore, we added Oil-red-O stainings and the corresponding quantifications to Supp. Fig. 4 (see below) for the differentiation of hMADS in the absence of AATBC. We also want to emphasize, that the expression levels of PLIN1, a surrogate marker for differentiation was unchanged in our experiments, as already shown in the initial draft of the manuscript. On top of that, in all experiments presented in the original draft of the manuscript, AATBC gene expression was only altered in mature adipocytes.

(b) Some of the experiments in Figure 1 (K-L) seem to only show an N of 1.

RESPONSE: Figure 1 highlights a screening process to find new lncRNA regulated during thermogenesis. The forskolin sample was included to achieve an additional dimension in the filtering process. The displayed values in K&L demonstrate the validity of the sample. The validation of AATBC as a target was performed with statistical power in the work displayed in the following figures.

(c) The RNAscope data in Figure 2 is not at all convincing for nuclear localization

RESPONSE: We respectfully disagree. In our opinion, the RNAScope is convincing for nuclear localization of the lncRNA. However, we have repeated the experiments with different probes that strengthen our data (see figure for the reviewer)

(d) The ASO mediated knockdown of AATBC in Figure 3 only reduced expression slightly. A more complete knockdown or deletion may elicit a stronger phenotype.

RESPONSE: We thank the reviewer for the feedback. We have repeated the knockdown experiments but were not able to reduce the expression further, even after designing additional ASOs. However, already with current approach, the reduction in AATBC expression elicited a phenotype, highlighting the importance of AATBC in a dose-dependent manner.

(e) In Figure 4, OPA1 is shown as a single band in panel E and a doublet in panel N. Based on this, are the authors certain they are detecting OPA,1 or could this be a nonspecific band?

RESPONSE: We thank the reviewer for this comment. Protein extraction has been performed at different research institutes with slightly different buffers. Multiple bands (cleaved/uncleaved) have been described for OPA1 in the past, therefore we are certain that the correct protein has been detected.

*(f) The correlations in Figure 6 I-L and Figure 7 do not include any statistical analysis. *

REPONSE: For better readability, the statistical analysis is being mentioned in the figure legend. The reviewer might have overlooked this information.

• The gain of function studies in mice are problematic. The authors have performed a large amount of invasive studies in a short period of time. The animals will undoubtedly lose weight after each study and with insufficient time to recover, this could influence the subsequent studies.*

RESPONSE: These general concerns are valid, but all controls are in place and the animals gained weight during the experiments, as one would have been expected with animals of that age (see below).

*In addition, since the authors present data in Figures 1-4 arguing that AATBC overexpression is associated with increased thermogenesis, it is surprising that the authors never looked at this in Figure 5 (aside from measuring Ucp1 mRNA). It would be interesting to measure energy expenditure by indirect calorimetry and cold tolerance. *

RESPONSE: We agree with the reviewer on this point but are due to animal protocol limitations in conjunction with the viral approach are unable to perform these experiments.

• The authors do not provide any mechanistic insights into how AATBC may be acting.*

RESPONSE: Certainly, more mechanistic insight into the direct mode of action of AATBC would be interesting. To address this point, over the past year we performed multiple attempts to perform pulldown of AATBC using the ChIRP technology. However, we were unable to achieve a sufficient enrichment, which would have allowed us to give further information about direct interaction partners of AATBC. However, we believe that our data regarding mitochondrial dynamics, which we now also have confirmed in in vivo experiments, explain the connection of AATBC and thermogenicity. In future, we aim to work on this point further but for multiple reasons have decided to close this chapter here.

Minor comments:

• The introduction is rather long and would benefit from being condensed.*

RESPONSE: We have edited the text for better readability.

Reviewer #2:

Major Comments:

1. • The key conclusion that AATBC is a novel obesity-linked regulator of adipocyte plasticity is made relatively clear with the comparison between various stages of adipocytes and the loss and gain of function with AATBC. - Figure 1 H and J do not seem to be consistent with the data in Figure 1F in LINC00473 level-There is no difference in Control vs NE in the heatmap but in Figure1J, the difference seems to be quite obvious; Figure 1K does not seem to be consistent with AATBC level-The measurement in Control VS Fsk group showed no difference in AATBC in heatmap, but in Figure K, there seem to be a dramatic increase. Therefore, the claims that there is a difference in these two lncRNA expression in these cell groups needs further clarification. *

RESPONSE: To combine the different approaches to identify novel lncRNA into one heatmap the data need to be normalized over experiments. As the fold change of the expression of AATBC in BAT compared to WAT (on average ~100x) is higher than with forskolin (~4x), this will stand out in the heatmap and will to some extent overshadow the smaller fold changes. The same holds true for LINC00473, which is drastically induced with forskolin, which to some extent masks the higher expression in the other approaches. Therefore, we decided to show both the heatmap to represent the general approach and the “zoomed in” versions to show the consistent increases. We are confident this clarifies the issue.

• Figure 4H and I, the difference in the representative immunoblot seem to be minimal and inconsistent with the decrease shown in the bar graph. *

RESPONSE: We agree with the reviewer and have removed the claim from our manuscript.

• In Figure 5, after overexpressing human AATBC in murine adipose tissue , is it possible to look at the mitochondria changes that were seen before in cell lines? If there are similar changes in murine adipose tissue, then it would prove the changes in vitro hold up with the in vivo model. But if the mitochondria changes were not seen, then it would indicate the changes in leptin, triglyceride levels may due to other mechanisms. The length of the suggested experiment to look into the mitochondrial differences in mice may vary depending on whether there are preserved samples from previous experiments. If there are, then the time period would be couple of weeks for immunblot and analysis. If there are no samples preserved, then the estimated period for the suggested experiments may be around 1.5 to 2 months at least .*

RESPONSE: We thank the reviewer for the suggestion. We performed Western Blot analysis on the tissues from the in vivo study and have included them in Fig. 5, further strengthening the link between AATBC and mitochondrial dynamics (please see figure on the right).

• The data are convincing overall in that the replicates are clearly marked with dots in many figures. Some immune blot and expression level are inconsistent with other data showing the same results however. *

RESPONSE: We thank the reviewer and have removed the necessary quantifications.

• Figure 6 and 7 are provocative and significant, reporting strong associations of AATBC with well-known markers of metabolism in adipocytes. The sex difference for adiponectin and AATBC expression is particularly intriguing. Further discussion of this point would be interesting. However, there is no information provided about the medication status of the obese subjects that were consented for samples used in the analysis. Specifically, many of the obese subjects (mean BMI 45 or more with a range going up to 97.3) would be expected also to have metabolic diagnoses and to be treated with numerous medications, including Metformin, GLP1 agonists, Orlistat, Liraglutide, Bupropion/Naltrexone and combinations. It is unreasonable to ignore possible effects of major medications on AATBC expression. Please comment on the strengths and weaknesses of the analysis that ignores medications, or if some annotations of clinical data are available, perhaps to explain outliers in the plots, please discuss. *

RESPONSE: We thank the reviewer for this suggestion. Unfortunately, we are unable to exclude additional diagnoses and medication of our patients due to the points the reviewer stated. However, given the large size of the cohorts we are confident that such effects are being compensated for. We have added a part on weaknesses of the study in the discussion.

Minor Comments:

• The labeling of figure 2 A-K is not clear because the use of the same color of bars is easily misunderstood as the same source of cells, but it is in fact not. For example, the grey color that appeared in 2B and 2C are not the same source but can be misunderstood. *

RESPONSE: The coloring of Fig.2A&G has been changed.

• Figure 3 ASO-AATBC has two repeats #1 and #2, and over-expression of AATBC has one, even though there are enough repeats. It would be less confusing to present all of the repeats in ASO_AATBC together in one bar.*

RESPONSE: The two different ASO target different areas of AATBC. In line with general guidelines for ASO use, those are not pooled but used separately, which is why the results are also split up. As the overexpression is additional genomic information of AATBC, it is impossible to use different variants in this case, therefore only one bar for overexpression is shown.

• The experimental outline can be a bit more detailed and explain some of the words like Thermo versus Browning.*

RESPONSE: The manuscript has been revised regarding this point.

• Some of the panels in Figure 7 could be put into supplementary if space is at a premium, and present the representative graph would be enough*

RESPONSE: We think that all our data of Fig. 7 warrants enough attention to be considered in a main figure, but if space is sparse, we are very happy to oblige. We would kindly ask the editors for input on this matter.

Reviewer #3:

1. • Throughout the study, the data provided are mainly correlative and in some cases not robust. In Fig. 2, AATBC expression is described to be elevated in the so-called "thermogenic condition", which contained prolonged PPARg agonist treatment (rosiglitazone) known to promote adipogenesis. Consistent with this notion, adipogenic markers, such as PLIN1 and FABP4, are higher in "thermogenic adipocytes" (Suppl Fig. 2). As such, the result may only suggest that AATBC has higher expression in mature adipocytes vs pre-adipocytes. *

RESPONSE: We thank the reviewer for the suggestion. We have added Oil-Red-O-Stainings to Suppl. Fig. 2 to show unchanged lipid content upon modulation of AATBC gene expression, which can be seen as a surrogate for differentiation. Concerning the use of rosiglitazone as a browning agent, we want to emphasize that rosiglitazone was used during the entirety of differentiation until day 9, where it was removed in the “non-thermogenic” group. At this point we already observe fully differentiated adipocytes. This is an established protocol. Furthermore, the data is in line with using norepinephrine or forskolin as a short-term inducer of browning, making it very likely that the effect seen is due to the “more thermogenic” character of the adipocytes.

• Along the same vein, whether and how AATBC affects adipogenesis is unclear. Suppl Fig. 3H and 3L (misplaced as Suppl Fig. 4) show the adipocyte differentiation marker FABP4 is down-regulated by both ASO- and AV-AATBC. Since mitochondrial respiration (and other parameters including UCP1 expression) is tightly linked to adipogenic efficiency, the authors need to address whether these manipulations affect adipocyte differentiation. *

RESPONSE: We agree with the reviewer that differences in differentiation capacity would falsify our data on mitochondrial dynamics. We have added Oil-Red-O-Staining to Suppl. Fig. 2 to show that no significant difference in lipid content exists during modulation of AATBC gene expression, which can be seen as a surrogate for differentiation. Furthermore, in all experiments presented in the manuscript, the modulation of AATBC occurs in already fully differentiated adipocytes. Accordingly, we are confident that AATBC does not influence differentiation but mainly acts through the modulation of mitochondrial dynamics.

• The data in Fig. 4 supporting a role for AATBC in regulating mitochondrial dynamics are superficial and not robust. Fig. 4A/4J do not have high enough resolution to provide accurate assessment of the mitochondrial network.*

RESPONSE: We respectfully disagree with the reviewer on this point. State of the art methods and algorithms were used to image and analyze the mitochondrial network. Furthermore, we have used multiple established markers of mitochondrial dynamics in western blot analysis to further strengthen our assessments of the immunofluorescence. In summary, we feel like have given enough evidence for an accurate assessment of the mitochondrial network.

• The level of loading control TUBB is clearly lower in siAATBC in Fig. 4H. In addition, OPA1 should have multiple isoforms and Fig. 4E/4N show inconsistent patterns. As such, mitochondrial dynamics is not likely an underlying mechanism. *

RESPONSE: We agree with the reviewer on the assessment of the expression of complex 5 and have removed this claim from the manuscript. Regarding the expression of OPA1, protein extraction has been performed at different research institutes with slightly different buffers. Multiple bands (cleaved/uncleaved) have been described for OPA1 in the past, therefore we are certain that the correct protein has been detected.

• Notably, RNAseq data in Suppl Fig. 4 (misplaced as Suppl Fig. 3) seem to indicate that AATBC over-expression promotes TG synthesis, while AATBC knockdown modulates cell death. The authors should consider exploring the leads from RNAseq analysis?*

RESPONSE: We thank the reviewer for the feedback. The small number of altered genes in the RNASeq make us believe in a rather post-transcriptional role of AATBC. We investigated cell death and oxidative stress response as GO terms were highlighted in the analysis, but we were unable to detect any differences in the absence of AATBC, pointing to a minimal effect on transcriptional level (See figure below for the reviewers).

• In Fig. 5, the AV-AATBC transduction in WAT/BAT is localized, transient and not homogeneous. Not surprisingly, this manipulation does not produce any robust effects. The difference in circulating leptin/leptin expression appears to be driven by 4-5 mice in the control group (Fig. 5H/5N). The correlation data in Fig. 6 and Fig. 7, although relevant, do not provide additional mechanistic insights. Unfortunately, the efforts in Fig. 5-7 fail to lead to information related to the biological function of adipose AATBC.*

RESPONSE: We agree with the reviewer on the limitations of the AV model, but we have performed these experiments with the highest technical standard. As the reviewer states, the overexpression, especially in WAT, has different magnitudes depending on the individual mouse, but the overexpression is present and consistently high in every animal. We would expect even bigger alterations in a genetic model, which, however, is beyond the scope of this first manuscript on AATBC in adipocytes. We are disappointed that the reviewer does not value the human data presented, as it very strongly hints to a relevant function of our human lncRNA in vivo by robust correlations with established biomarkers mirroring the effects seen in vitro and in the mouse model. A limitation of human studies is in virtually every case that it is based on correlations, as manipulation of gene expression, which would be necessary to delineate a biological process as requested by the reviewer, is not possible in humans. We do not concur on dismissing our human data on that behalf.

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 #3

Evidence, reproducibility and clarity

In the current work, the authors have characterized a human IncRNA AATBC whose expression is up-regulated in conditions favoring a "thermogenic phenotype" of fat cells. Results derived from transient knockdown/over-expression approaches indicate that AATBC may play a role in modulating mitochondrial functions. In addition, gene expression analyses have demonstrated that AATBC is positively correlated with thermogenic genes, such as UCP-1 and PGC-1a and negatively correlated with adipogenic genes, including PPARg, FABP4 and Leptin in humans.

The association between AATBC and BMI is of potential interest. However, restricted by the limitations of the employed approaches, the current study falls short of robust evidence supporting a role for AATBC in adipocyte plasticity and mitochondrial dynamics/respiration (and certainly a link between the two events). A substantial amount of work would be needed to tie up loose ends for a cohesive study with mechanistic insights.

Specific comments:

1. Throughout the study, the data provided are mainly correlative and in some cases not robust. In Fig. 2, AATBC expression is described to be elevated in the so-called "thermogenic condition", which contained prolonged PPARg agonist treatment (rosiglitazone) known to promote adipogenesis. Consistent with this notion, adipogenic markers, such as PLIN1 and FABP4, are higher in "thermogenic adipocytes" (Suppl Fig. 2). As such, the result may only suggest that AATBC has higher expression in mature adipocytes vs pre-adipocytes.
2. Along the same vein, whether and how AATBC affects adipogenesis is unclear. Suppl Fig. 3H and 3L (misplaced as Suppl Fig. 4) show the adipocyte differentiation marker FABP4 is down-regulated by both ASO- and AV-AATBC. Since mitochondrial respiration (and other parameters including UCP1 expression) is tightly linked to adipogenic efficiency, the authors need to address whether these manipulations affect adipocyte differentiation.
3. The data in Fig. 4 supporting a role for AATBC in regulating mitochondrial dynamics are superficial and not robust. Fig. 4A/4J do not have high enough resolution to provide accurate assessment of the mitochondrial network. The level of loading control TUBB is clearly lower in siAATBC in Fig. 4H. In addition, OPA1 should have multiple isoforms and Fig. 4E/4N show inconsistent patterns. As such, mitochondrial dynamics is not likely an underlying mechanism. Notably, RNAseq data in Suppl Fig. 4 (misplaced as Suppl Fig. 3) seem to indicate that AATBC over-expression promotes TG synthesis, while AATBC knockdown modulates cell death. The authors should consider exploring the leads from RNAseq analysis.
4. In Fig. 5, the AV-AATBC transduction in WAT/BAT is localized, transient and not homogeneous. Not surprisingly, this manipulation does not produce any robust effects. The difference in circulating leptin/leptin expression appears to be driven by 4-5 mice in the control group (Fig. 5H/5N). The correlation data in Fig. 6 and Fig. 7, although relevant, do not provide additional mechanistic insights. Unfortunately, the efforts in Fig. 5-7 fail to lead to information related to the biological function of adipose AATBC.

Significance

Thermogenic adipocytes are thought to be a druggable target to combat obesity and related metabolic diseases. The current study aims to identify genes associated with thermogenic capacity in human adipocytes. To this end, the authors have characterized a human IncRNA AATBC whose expression is up-regulated by thermogenic stimulations in cultured adipocytes. Of potential interest is the association between adipose AATBC expression and BMI in human samples.

Reviewer's field of expertise: metabolic regulation, obesity and related metabolic diseases, molecular physiology.

Referee Cross-commenting

I think there is a good consensus about the strengths and weaknesses of the study. My comments were very similar to those of reviewer 1. My view is the authors did not provide sufficient evidence to support their claims. Some of the data are also not robust enough to reach meaningful conclusions.

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

This investigation deals with a significant problem: how to discover and understand the signals that regulate white and brown or beige adipogenesis, which is central to energy balance. New approaches to promote brown, thermogenic adipogenesis have been considered as potential metabolic therapies to combat the current epidemic of obesity and obesity-related complications. Interestingly, the authors have identified a long non-coding RNA that associates with the thermogenic phenotype, and then through functional studies offer strong evidence that this lncRNA (AATBC) has potential as a regulator of adipocyte plasticity. The investigation is rigorously conducted, novel and important. The report may be suitable for publication upon completion of the improvements suggested here.

Major Comments:

• The key conclusion that AATBC is a novel obesity-linked regulator of adipocyte plasticity is made relatively clear with the comparison between various stages of adipocytes and the loss and gain of function with AATBC.
• Figure 1 H and J do not seem to be consistent with the data in Figure 1F in LINC00473 level-There is no difference in Control vs NE in the heatmap but in Figure1J, the difference seems to be quite obvious; Figure 1K does not seem to be consistent with AATBC level-The measurement in Control VS Fsk group showed no difference in AATBC in heatmap, but in Figure K, there seem to be a dramatic increase. Therefore, the claims that there is a difference in these two lncRNA expression in these cell groups needs further clarification.
• Figure 4H and I, the difference in the representative immunoblot seem to be minimal and inconsistent with the decrease shown in the bar graph.
• In Figure 5, after overexpressing human AATBC in murine adipose tissue, is it possible to look at the mitochondria changes that were seen before in cell lines? If there are similar changes in murine adipose tissue, then it would prove the changes in vitro hold up with the in vivo model. But if the mitochondria changes were not seen, then it would indicate the changes in leptin, triglyceride levels may due to other mechanisms.
• The length of the suggested experiment to look into the mitochondrial differences in mice may vary depending on whether there are preserved samples from previous experiments. If there are, then the time period would be couple of weeks for immunblot and analysis. If there are no samples preserved, then the estimated period for the suggested experiments may be around 1.5 to 2 months at least.
• The data are convincing overall in that the replicates are clearly marked with dots in many figures. Some immune blot and expression level are inconsistent with other data showing the same results however.
• Figure 6 and 7 are provocative and significant, reporting strong associations of AATBC with well-known markers of metabolism in adipocytes. The sex difference for adiponectin and AATBC expression is particularly intriguing. Further discussion of this point would be interesting.

However, there is no information provided about the medication status of the obese subjects that were consented for samples used in the analysis. Specifically, many of the obese subjects (mean BMI 45 or more with a range going up to 97.3) would be expected also to have metabolic diagnoses and to be treated with numerous medications, including Metformin, GLP1 agonists, Orlistat, Liraglutide, Bupropion/Naltrexone and combinations. It is unreasonable to ignore possible effects of major medications on AATBC expression. Please comment on the strengths and weaknesses of the analysis that ignores medications, or if some annotations of clinical data are available, perhaps to explain outliers in the plots, please discuss.

Minor Comments:

• The labeling of figure 2 A-K is not clear because the use of the same color of bars is easily misunderstood as the same source of cells but it is in fact not. For example, the grey color that appeared in 2B and 2C are not the same source but can be misunderstood.
• Figure 3 ASO-AATBC has two repeats #1 and #2, and over-expression of AATBC has one, even though there are enough repeats.
• It would be less confusing to present all of the repeats in ASO_AATBC together in one bar.
• The experimental outline can be a bit more detailed and explain some of the words like Thermo versus Browning.
• Some of the panels in Figure 7 could be put into supplementary if space is at a premium, and present the representative graph would be enough.

Significance

• Adipocyte plasticity and physiology has been linked to a lot of diseases such as diabetes, cardiovascular diseases and cancer. Therefore the conclusion coming from this paper reveals other aspects when looking at adipocytes in healthy or disease conditions-long noncoding RNA. lncRNA can be seen as a biomarker of an indicator of the physiology of adipose tissue, and may be able to account for changes that cannot be explained by cell genome analysis. There has been a surge of interest in RNA containing in exosomes, which serves as vesicles that travel between cells. Some studies have also shown that the content of exosomes arrives at the nucleolus of the recipient cell. The overlap of the location of exosomal RNA and lncRNA is a representative of a whole set of regulation of genes and expressions that was not noticed before.

• The audience of the paper may be interested in the outcome of the changes of mitochondria change in the context of disease such as obesity, diabetes, etc. For example, if there is a casual relationship between AATBC level and the status of obesity.

Referee Cross-commenting

Thank you, I concur.

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

The manuscript by Giroud et al. describes a role for the human-specific lncRNA AATBC in adipocyte plasticity. By overlaying datasets from tissues (white vs. brown fat) and cell lines (treated with norepinephrine or forskolin), the authors identified a limited number of lncRNAs demonstrating coordinate regulation. One of these lncRNAs is AATBC, which has not previously been studied in adipocytes. The authors show that AATBC is enriched in thermogenic adipose tissues/cells. They then perform gain and loss of function studies in cellular models and argue that AATBC is involved in thermogenesis and appears to be associated with the state of the mitochondrial network. The authors then explain that modulating AATBC has minimal effects on global transcription, and so they argue it mainly works via post-transcriptional mechanisms, though these are not defined. The authors then expressed AATBC in adipose tissue of mice and observed a decrease in plasma leptin levels and an increase in triglyceride levels, while other metabolic phenotypes were unchanged. Finally, the authors analyzed associations between adipose tissue AATBC and a variety of metabolic parameters in a few human cohorts. While the identification of a novel lncRNA involved in adipocyte biology and systemic metabolism would be of great interest, the data presented here does not convincingly support the conclusions made. Substantial additional experiments are needed to support the claims in this paper.

Major comments:

1. The relevance of these findings to human biology remains unclear. In Figures 1-4, the authors present data showing that AATBC is enriched in thermogenic fat, and they argue that it regulates thermogenesis and mitochondrial biology. However, in Figures 6-7, where the authors look at AATBC in different human cohorts, they actually find that it is enriched in visceral fat, which is thought of as being the least thermogenic fat depot. The authors do not explain this seeming paradox, and thus, the role of AATBC in fat remains uncertain.
2. In many of the experiments, insufficient controls are provided or the data are not at all convincing. For example: (a) The first four figures rely on in vitro adipocyte models, but the authors do not present data to show these cells differentiate properly and equally. This is especially relevant for the gain and loss of function studies. (b) Some of the experiments in Figure 1 (K-L) seem to only show an N of 1. (c) The RNAscope data in Figure 2 is not at all convincing for nuclear localization. (d) The ASO mediated knockdown of AATBC in Figure 3 only reduced expression slightly. A more complete knockdown or deletion may elicit a stronger phenotype. (e) In Figure 4, OPA1 is shown as a single band in panel E and a doublet in panel N. Based on this, are the authors certain they are detecting OPA1 or could this be a nonspecific band?( f) The correlations in Figure 6 I-L and Figure 7 do not include any statistical analysis.
3. The gain of function studies in mice are problematic. The authors have performed a large amount of invasive studies in a short period of time. The animals will undoubtedly lose weight after each study, and with insufficient time to recover, this could influence the subsequent studies. In addition, since the authors present data in Figures 1-4 arguing that AATBC overexpression is associated with increased thermogenesis, it is surprising that the authors never looked at this in Figure 5 (aside from measuring Ucp1 mRNA). It would be interesting to measure energy expenditure by indirect calorimetry and cold tolerance.
4. The authors do not provide any mechanistic insights into how AATBC may be acting. The manuscript contains some potentially interesting observations, but without some mechanistic insight, it is hard to understand how AATBC might regulate adipocyte plasticity.

Minor comments:

1. The introduction is rather long and would benefit from being condensed.

Significance

This manuscript may represent an interesting advance in terms of highlighting a new lncRNA with a role in adipocyte biology. These findings would be of broad interest to researchers interested in obesity and metabolism. I myself am in this field of research, so feel quite qualified to evaluate this manuscript. However, as noted above, major concerns would need to be addressed in order to justify the conclusions made here.

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

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

*The current manuscript by Shiryaev et al describes their observation of the new function of zika NS2B-NS3 proteases. They have shown that NS2B-NS3 protease lacking the helicase domain binds to RNA and the interaction can be affected by protease inhibitors. Main two new findings are presented in the manuscript: super open conformation of the protease; RNA binding activity of the protease region. Although the manuscript is interesting, the design of the experiments is not convincing. *

Major issues:

1. • the claim of a super open confirmation is problematic. Using an artificial construct lacking the C-terminal portion of NS2B will of course generate the open conformation. This is a wrong definition unless you observe such a conformation in living cells.*

2. We understand the skepticism towards a less known super-open confutation of flavivirus NS2B-NS3pro complex. In addition to our own structure of ZIKV NS2B-NS3pro (PDB ID 7M1V), the crystal structure of another orthologous flavivirus Japanese encephalitis virus (JEV) NS2B-NS3pro (PDB ID 4R8T) was discovered in 2015 1. However, no functional analysis was provided for this crystal structure resulting in the lack of attention paid by the research community. We computed the overlay of the ZIKV NS2B-NS3 protease structures in the super-open conformation (PDB ID 7M1V, deposited by us in 2021) with the crystal structure of JEV protease (PDB ID 7M1V ) (Rebuttal Figure 1). We observed an almost identical organization of the critical NS3pro C-terminal loop between these two structures (RMSD 0.6A). Polypeptides with over 35% identity are very likely to have a similar fold2. Given over 50% identity(!) between flaviviral proteases across the family3,4, we posit that the super-open conformation demonstrated for JEV and ZIKV NS2B-NS3pro is a common feature of the Flaviviridae family. Further, NS2B peptide is always tightly associated with NS3pro via a three-strand beta-barrel (aa 49-58 of NS2B), which remains intact in all NS3Pro conformations. The C-terminal portion of NS2B progressively loses association with NS3pro, being mostly associated in the closed conformation, less so in the open, and even less in the super-open conformation. The G4SG4 linker between NS2B and NS3pro remains unstructured in all conformations. The native C-terminal portion of NS2B (TGKR) is equally unstructured when competed out of the protease active site by another substrate. It is unclear to us why “lacking the C-terminal portion of NS2B will of course generate the open conformation”.

3. It is odd that authors made homology model to generate open conformation structures. the authors did not cite the two papers of eZiPro (Phoo et al 2016 NC) and bZiPro (Zhang et al 2016, Science). these two structures show the closed conformation of protease in the absence and presence of a natural substrate.*

4. We agree with the reviewer that in both constructs eZiPro5 and bZiPro6 of ZIKV NS2B-NS3pro are likely to exist in the closed conformation as documented by the crystal structures. However, in both cases, the active center of ZIKV NS2B-NS3pro is occupied with a short peptide fragment, which is sufficient to induce the closed conformation of NS2B-NS3 protease. We superimposed eZiPro (PDB ID 5GJ4) with bZiPro (PDB ID 5GPI) to better demonstrate that the active center in both structures is occupied either by tetrapeptide TGKR (T127-G128-K129-R130 ) originating from the NS2B C-terminus (eZiPro) or by a tetrapeptide KKGE (K14-K15-G16-E17) originating from a neighboring NS3 molecule (bZiPro) (Rebuttal Figure 2). Indeed, Zheng et al., 2016 6 stated that: “the structure (bZiPro) does capture the protease in complex with a reverse peptide. The tetrapeptide K14K15G16E17 folds into a small hairpin loop to occupy the active site.” Further, Phoo et al., 2016 5 stated that “binding of the ‘TGKR’ peptide to the catalytic site stabilizes the protease (eZiPro)”. To the best of our knowledge, so far there are no crystal structures of flaviviral NS2B-NS3 proteases in the closed conformation without peptide/inhibitor in the active center. We take it as a hint that the closed conformation is always induced by a substrate present in the active center.

Finally, we would like to draw the attention of this reviewer to the fact that the 15N R2 NMR signal from NS2B residues 65-85 is missing in bZiPro alone but re-appears when AcKR is added. This is consistent with the idea that without AcKR, bZiPro exists in the open conformation where much of the C-terminal part of NS2B is dissociated from NS3Pro and remains unstructured, thus resulting in the lack of NMR signal.

• RNA binding is novel, but is it observed in cells? only one method was used for testing the interactions, not other biophysical methods are used.*

• Given a complex network of protein-RNA interactions and the fact that NS3pro and NS3hel are connected by a single polypeptide, separating dynamically bound 11kB RNA to NS3pro from that to NS3hel in a native cell is a major technical challenge beyond the scope of this work. We employed a fluorescent polarization assay to demonstrate ssDNA and ssDNA binding to ZIKV NS2B-NS3pro. Subsequently, we employed a proteolytic activity assay with labeled peptide mimicking natural substrate for protease to demonstrate that the presence of ssRNA and ssDNA can efficiently inhibit proteolytic activity. To the best of our knowledge, this is the first indication that ssRNA or ssDNA could block proteolytic activity for any serine proteases, let alone a viral protease. Therefore, we consider the proteolytic activity assay used in the current work an orthogonal biochemical method supporting ssRNA binding to ZIKV NS2B-NS3pro.

• binding studies with RNA used artificial construct, how about the one with KTGR present like eZiPro. Keep in mind that the P1-P4 residues are present under native conditions.*

__- __As mentioned by the reviewer, TGKR peptide was found in the active center in the eZiPro crystal. Indeed, the junction region between NS2B and NS3 protease contains native cleavage sites for the NS2B-NS3Pro and is naturally cleaved by protease during the viral polyprotein processing. However, the TGKR peptide representing P1-P4 positions will have to leave the active center after the cleavage to ensure enzyme processivity/cleaving additional targets (otherwise, the protease would get stacked after the first cleavage). Proteolytic activity assay utilizes the fluorogenic peptide labeled with FAM (such as TGKR-FAM; where FAM is a group representing P1’ position in this case). TGKR-FAM peptide will compete and easily replace cleaved TGKR peptide from the active center in proteolytic activity assay. In sum, the C-terminal end of NS2B will be competed out of the protease active center by the next substrate, and there is no evidence that it will be naturally placed back in the active center after each round of protease proteolytic activity. Indeed, several crystal structures of flaviviral NS2B-NS3Pro in open conformation lack the C-terminal part of NS2B in the active center. Our unpublished NMR studies demonstrated that the C-terminal part of NS2B is unstructured in solution if the substrate peptide or small molecule inhibitor are not present in the active center of the protease.

• authors built up nice models, it is great to consider the full length NS2B, but authors haven't taken into account the effect of NS2B on the open or closed conformation of the protease. *

- __ All crystal structures of flavivirus NS2B-NS3pro in the closed, open, or super-open conformations have NS2B associated withNS3pro via a beta-barrel (__Rebuttal Figure 3), which is located at the opposite side from the RNA binding site. The transition from the closed to the open and to the super-open conformation is associated with the progressive dissociation of NS2B from NS3pro. Therefore, the effect of NS2B on NS3Pro is progressively diminished. In the closed conformation of NS3Pro, the negatively charged C-terminal part of NS2B is associated with the same positively charged grove as the RNA in the open conformation of NS3Pro. The C-terminal part of NS2B is dissociated from NS3Pro in the open conformation.

Minor issues:

*This manuscript shows the novel function of zika protease and conclude that protease binds to RNA. This is a novel finding, but the conclusion needs to be further confirmed, to avoid misinterpretations by future readers *

• closed, and super open conformations. But the definition was not carefully compared with current literatures. I am surprised that the two important papers are not cited. It is well known the G4SG4 linker affect the conformation of the protease.*

• The crystal structures and the proteolytic activities of gZiPro, eZiPro, and bZiPro are rather similar. In fact, Km (μM) are 2.86 ± 0.90 for gZiPro, 6.332 ± 2.41 for bZiPro, and the IC 50s of BPTI inhibition for gZiPro, eZiPro and bZiPro are 350, 76 and 12 nM respectively. NS2B and NS3pro have a large binding area in the closed conformation. Upon changing the conformation to the open conformation (and even more so to the super-open conformation), the C-terminal part of NS2B is progressively dissociated from NS3Pro. Therefore, possible minor effects introduced by the G4SG4 linker is unlikely to affect any of the conclusions in our work.

• Authors need to show super open conformation is present in nature e.g. the model in which full length NS2B and NS3pro.*

• A full-length NS2B has 2 transmembrane domains, which tether the NS2B-NS3pro complex to the cell membrane (we have modeled the presence of such transmembrane domains to account for the orientation of NS2B-NS3pro with respect to the cell membrane). The full-length complex has never been crystallized or tested in any assay due to the major technical challenges associated with the modeling of complex transmembrane proteins.

• RNA is a charged molecule under some conditions, NS3 also have charged residues, it is important to show whether the binding between RNA-protease is relevant to the function{Luo, 2010 #9270;Chernov, 2008 #9275;Xu, 2019 #10006}, or is this due to the application of the artificial constructs used in this study. Why so many mutants are used? *

• The requirement of NS3pro for the helicase function was shown by several investigators 7–9. Given the structural independence of NS3pro and NS3hel, which mostly rules out the allosteric effect, RNA binding by NS3pro is a newly proposed function of NS3pro for the helicase activity. We demonstrated biochemically that RNA-bound to NS3pro inhibits its protease function. A variety of mutants were used to constrain the conformations of NS2B-NS3pro (e.g. enforce the super-open confirmation) for crystallization studies.

• Using a construct close to the native protease, at least the P1-P4 residues should be present. Using a peptide in the assay is also useful.*

• We were unable to interpret this critique.

• Test binding of RNA with protease using another method such as biophysical methods, or even gel shift assay*

• We thank the reviewer for this suggestion. Although the gel-shift assay seems to be a reasonable method to test the binding, given the ease of spontaneous conformational change (i.e. into the super-open conformation), this assay could result in a progressive loss of bound RNA during migration in the gel.

• I don't know the correlation between Figure 7 and Figure 6. The authors describe ploy A binding to protease, while Figure 7 is talking about Helicase binds to dsRNAs. *

• There is no correlation. Figure 6 describes the models for NS2B-NS3pro binding to ssRNA. Figure 7 describes a separate point, the direction of dsRNA processing by NS3hel.

• I am glad to see the consideration of full length NS2B, NS3 in the models Figure 8, 9 and 11, but there is no data to support any of the model proposed. *

• There is no experimental data. We have modeled the N-terminal and C-terminal parts of full NS2B, which are predicted to be inserted into the cell membrane due to their characteristic amphipathic helical structure.

• Is the linker a ploy G not G4SG4? *

The linker is GGGGSGGGG (G4SG4) as stated in Materials and Methods of the manuscript.

• Do the mutant sustain their protease activity? *

• All mutants with intact catalytic centers have protease activity, except the mutants with a disulfide bridge that fixes the polypeptides in the super-open conformation.

---

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

*The manuscript by Shiryaev et al., submitted to BioRXiv is an exploration of the ability of NS2B-NS3protease to bind RNA and its subsequent role in NS3 helicase processivity. The authors first utilize fluorescence polarization assays to demonstrate that NS2B-NS3protease can bind ssRNA with a strong affinity (and also ssDNA with lower affinity). They subsequently utilize mutational and small molecule inhibitor strategies in these assays to force the NS2B-NS3protease into different conformations, with the associated results inferring that the "open" conformation is responsible for ssRNA binding affinity. Furthermore, they demonstrate that ssRNA binding impairs protease activity, suggesting these roles may be exclusive in the viral life cycle. They also identified a number of small molecule ligands that target the putative ssRNA binding channel, and demonstrate that these ligands inhibit ssRNA binding by NS2B-NS3protease, providing potential inhibitor candidates for ZIKV. Finally, the authors utilized their crystal structures and others for the various conformations of NS2B-NS3protease to model ssRNA binding by the domain and the full NS3 protein, and used these models to propose a reverse inchworm model for NS3 travelling along ssRNA as it unwinds the dsRNA duplex. Overall, the authors utilize a comprehensive approach to demonstrate a number of novel findings (ssRNA binding by NS2B-NS3protease, small molecule ligands that inhibit this interaction) that would be of interest to both virologists and structural biologists. However, there are some important experimental design limitations and viral life cycle considerations that the authors should address before acceptance of the manuscript. Major and minor comments intended to improve the manuscript are outlined in more detail below. *

Major Comments:

1. • While the quantity of indirect data (ruled out closed and super-open, inhibitors of ssRNA binding pocket) suggest that the open conformation of NS2B-NS3protease is associated with ssRNA binding, the argument would be greatly strengthened by direct experimental data. Is there a mutational or small molecule approach to locking the NS2B-NS3 protease in the open conformation? If so, the authors should perform such experiments to strengthen the foundation of their argument.*

2. Unfortunately, despite significant efforts, mutations or small molecules locking the NS2B-NS3 protease in the open conformation have not been identified for the ZIKV protease. However, several structures for NS2B-NS3 proteases have been documented in other flaviviruses (i.e., DENV PDB IDs 2FOM and 5T1V; WNV PDB ID 2GGV). Polypeptides with over 35% identity are very likely to have a similar fold2. Given over 50% identity(!) between flaviviral proteases across the family3,4, there is little doubt that ZIKV NS2-NS3 protease adopts an open conformation similar to all flaviviral proteases. Our modeling demonstrated that there are no sterically/structural problems in folding NS2B-NS3 protease into the open conformation.

3. A negative control should be used in Figure 4A to strengthen the claim that ssRNA binding in the open conformation impairs protease activity (ie. include a curve for dsRNA). Such an experiment would lend support to ssRNA inhibition being due to specific binding instead of some other non-specific effect of increasing local nucleic acid concentration.*

4. To address this critique, we have conducted the modeling of dsRNA binding to the open conformation of NS2B-NS3Pro. The model revealed that dsRNA could not be accommodated by the open conformation of the NS2B-NS3Pro complex (Rebuttal Figure 4). Indeed, dsRNA has a very different rigid structure compared to the extended form of the ssRNA chain. The dsRNA is unable to provide continuous interactions between negatively RNA backbone and positively charged side chain amino acids in NS3pro. The continuous interface on NS2B-NS3 protease interacting with ssRNA is an extension of the exit groove for one of the ssRNA strands exiting the NS3 Helicase after unwinding. Therefore, the ssRNA, but not dsRNA is naturally always present in close proximity of the NS2B-NS3Pro complex.

5. *

6. Due to the highly coupled roles of NS5 and NS3 in replication, the authors should include some more consideration of the role of NS5 in their complex. They very briefly address this interplay in the fifth paragraph of the discussion, but then neglect to discuss the implications any further. In particular (perhaps in a brief comparison to an NS3/NS5 modeling approach such as Brands et al., 2017; WIRES), the authors should consider some of the following questions: could the channel on protease domain lead to ssRNA entry site on RdRp?*

7. Indeed, our model suggests that the negative strand (-)ssRNA exits from NS2B-NS3protease facing the ER membrane in the area where the protease is anchored to the ER membrane via the NS2B transmembrane domains. It is possible that NS3pro interacts with NS5 polymerase and “handles” (-)ssRNA to the NS5 polymerase. This scenario would modify Brands et al., 2017 model to add NS2B-NS3Pro complex between NS3Hel and NS5. However, at present, the NS3-NS5 (or NS2B-NS3-NS5) complex together has not been crystallized. It would be logical for NS5 polymerase to access the (-)ssRNA strand after it is released from NS2B-NS3Pro since the (-)ssRNA strands are used as a template for the (+)ssRNA which is used for polyprotein synthesis and packaging into viral particles.

8. would NS5 interaction constrain or augment inchworm model of NS2B/NS3 translocation? *

9. Yes, integrating NS5 interaction with the NS2B-NS3pro handling (-)ssRNA will augment the utility of the suggested reverse inchworm model.

10. how does increased activity of NS3 when complexed with NS5 (**Xu et al. 2019) align with proposed inchworm model? *

11. We appreciate the reviewer's question. We think that NS2, NS3, NS4, and NS5 work in concert as one coordinated complex where various subunits of NS2 and NS4 may provide anchoring of the entire complex to the ER membrane. Indeed, such a complex has recently been proposed6. Also, see our response to the previous reviewer’s point (#4). We have incorporated this discussion into the revised manuscript.

Minor Comments: 1. Introduction, 4th paragraph, NS3-NS4 should read NS3-NS4A.

• We corrected this sentence.

* ** Throughout the manuscript, the authors should denote some key amino acid residues in each figure to help orient the reader better to the observed structural changes and rotations. Inclusion, at least in the supplement, of the crystal structures of mutants solved herein should **also be included. *

• We annotated the key residues in all figures (e.g. catalytic residues, loop interacting with the membrane, position of NS2B, and other elements) and kept the same orientation of complexes in all figures.

• Section: RNA binding inhibits the proteolytic activity of ZIKV NS2B-NS3pro, last sentence, NS2N-NS3pro should be NS2B-NS3pro*.

• We corrected this sentence.

• Section: Allosteric inhibitors of NS2B-NS3 protease interfere with RNA binding- first sentence: "The open conformation of NS2B-NS3pro is achieved by the rearrangement of NS2B cofactor (its dissociation from the C-terminal half of NS3pro) leading to a loss of proteolytic activity [32]. - the reference is not correct. I could not find the reference the authors refer to here and had not heard before that NS2B cofactor was able to disassociate from the C-terminal half of NS3pro; hence, this really needs to be appropriately referenced. *

• We have revised this sentence and added additional references. “The open conformation of NS2B-NS3pro is achieved by the rearrangement of NS2B cofactor (partial dissociation from NS3pro), leading to a loss of proteolytic activity4,11.”

• Section: Modeling RNA binding to ZIKV NS2B-NS3, first sentence - unwinds should be unwind*.

• We corrected this sentence.

---

• With respect to the results of Figure 3A, the authors should address that adding the linker alone to the NS3 protease may not be an accurate examination of its role/importance. The linker in this scenario is only constrained at its N-terminus, while it is always constrained at both termini during infection (and even more so by the interactions of those two linked domains [protease and helicase] with each other). As such, the authors statement that "observations suggests that the 12-aa linker region modulates RNA binding to NS2B-NS3pro" should be more strongly qualified to this effect. In addition, it would be interesting to see the effects of the linker mutations on ssRNA binding in the context of the full NS3 protein, albeit admittedly more complex due to the confounding ssRNA binding by the helicase domain.*

• We agree with this reviewer that the protease-helicase linker is also restrained at both termini. We have rephrased the statement in the revised manuscript. The goal of the experiment shown in Figure 3A was to examine whether a negatively charged linker is able to compete with ssRNA binding as we expected from the structural model. The mutational analysis of the protease helicase linker is, indeed, a very interesting subject that is, however, beyond the scope of this work.

7. The NS#hel should be changed to NS3hel in part (C) of figure legend for Figure 11. - We corrected this mishap.

• The authors data in Figure 4A (and even more so the nature of the viral life cycle where 1000s of viral polyproteins are created from the first genome during infection) disputes the depiction in the inchworm model of how NS3 protease cleaves the polyprotein while the helicase binds ssRNA. At minimum, the authors need to discuss this discrepancy, and it is recommended that they modify the cartoon in their model to not include the ssRNA binding on the protease side of the equation (or show as alternative on that side to the existing cartoon).*

• Indeed, as proposed by our reverse inchworm model, ssRNA is not bound to NS3Pro in the closed conformation, while NS2B-NS3pro has a protein substrate in the active center (Figure 11A). We agree that NS2B-NS3Pro in the closed conformation cannot bind ssRNA as we demonstrated in competitive cleavage assay. Only large amounts of ssRNA can shift the balance towards the open conformation which binds ssRNA. We think that most of the time NS2B-NS3Pro cycles between the open and the super conformations handling ssRNA (Figure 11(B-C_D), but as soon as protein substrate becomes available (typically a loop from a transmembrane viral polypeptide), NS2B-NS3Pro quickly switches to the closed proteolytically active conformation to act as protease.

• In the third paragraph of the discussion, the authors state "An alternative model of coupled transcription and translation where viral RNA is associated with ribosomes right after the release from NS2B-NS3 is also possible". Considering there is abundant evidence that translation and replication are exclusive and that translation does not take place in ROs, it would be prudent to remove such statements from the discussion. Without any supporting evidence, these statements will be misleading to readers by providing a false equivalency. The preceding discussion of RFs would be sufficient to contextualize your inchworm model in the broader viral life cycle (which was done quite well). *

• We have adjusted the discussion in the revised manuscript to avoid a false equivalency.

10. There were a number of aspects I appreciated about the manuscript and will briefly list a few here: ** i) the focus on how different non-structural proteins effect the structure and function of ** each other during the viral life cycle, which forms a more comprehensive and informative model ** ii) the use of structural and functional assays as complementary approaches to studying the intra- and inter-protein relationships of NS3 ** iii) the depiction of the forks in Figure 10, which effectively demonstrated the channels and oriented the reader to the conservation data ** *iv) the use of small molecule inhibitors to modify structure and function of NS3, which greatly deepened the richness of the story from both a basic and applied science view point *

• We are very grateful to the Reviewer for these kind remarks.

Reviewer #2 (Significance (Required)): ** Strengths and limitations: ** - provides some experimental and modeling data to provide a new model for RNA interactions with the NS3pro-hel; may help inform models for enzyme function, mostly consistent with previous literature ** - leaves out the NS5 RdRp, known to contribute to NS3 activity. ** - some suggestions are made which might strengthen the conclusions and inclusions of additional controls would improve the data. ** Advance ** - conceptual, perhaps may provide some insight into mechanism; although limited by the lack of NS5 RdRp which is crucial to helicase activity. It is unclear if the ssRNA would be oriented this way given interactions with NS5 RdRp and MT domains (is the ssRNA routed to NS5 or along NS3, or potentially are both happening?) ** Audience: ** - quite specialist, but may include structural biologists and virologist alike. ** Expertise of the reviewer(s): ** *- molecular virologists, RNA viruses - including flaviviruses; replication complex biogenesis, protein-RNA and RNA-RNA interactions. While comfortable with the concepts regarding complex formation, the appropriateness of computational modeling and RNA docking tools as well as structural biology is out of our area of expertise. *

---

---

---

---

---

Reviewer #3 (Evidence, reproducibility and clarity (Required)):

*This paper investigates the nucleic acid binding properties of zika virus protease. In particular the data suggest that single stranded RNAs and DNAs are capable of binding to and inhibiting ZIKV protease at micromolar concentrations. With the use of active site inhibitors and mutants that lock the protease in closed and super-open conformation, the authors concluded that RNA binds to the open conformation. Through extensive modeling of the protease and helicase domains, this manuscript provides a model of how ssRNAs can bind to all conformations of the proteas, but the open conformation provides two positively charged forks that should be available to bind RNA. *

* SECTION A - Evidence, reproducibility, and clarity ** Major comments: **

*·The main conclusions of this paper rely on the existence of the super-open conformation, however this conformation has not been reported in the scientific literature previously. Structures deposited in the pdb are referenced in this manuscript, however no citation for an accompanying publication is provided. This calls into question the biological relevance of this super open conformation. This is of particular concern because in other highly-homologous flaviviral proteases, structures that have been observed crystallographically (e.g. the open conformation of dengue virus protease) appear to be only very sparsely populated in solution. What is the evidence that the super-open conformation exists in solution.

• Please, see our reply to question #1 from Reviewer 1.

• The activity of each of the constructs used was not reported making it impossible to directly compare the impact of these changes on intrinsic activity. In particular, the NS2B-NS3 long construct is predicted to exist in the super-open conformation. If this is correct, it should show no activity against a peptide substrate. *

• We appreciate these concerns. The NS2B-NS3pro-long construct is proteolytically active (only NS2B-NS3pro-short construct is proteolytically inactive because its NS3pro C-terminal part is too short to fold into the closed conformation). It is unconstrained and likely capable of adopting all possible conformations (closed, open, super open). As we suspected, the negatively charged linker interferes with RNA binding, potentially via direct competition. Investigating the role of the protease-helicase linker is an exciting subject of a separate manuscript in preparation.

• This paper reports that the IC50 is much weaker than the Kd for binding of ssRNA to ZIKV NS2B-NS3pro. Are orthogonal assays, such as thermal shift assay, available which could distinguish between the reported IC50 and the Kd. *

• Binding of ssRNA occurs in an area distinct from the protease active center. We think that there is a constant competition between C-terminal NS2B binding/release versus ssRNA binding/release from NS3pro. We think that ssRNA “catches” the moment when protease has the open conformation and freezes that conformation by blocking the C-terminal of NS2B from binding to NS3Pro. In terms of thermal shift assay, the structure of NS3Pro is changed, only the C-terminal of NS2B is affected. Note that the 15N R2 NMR signal from NS2B residues 65-85 is missing in bZiPro alone but re-appears when AcKR is added6. This is consistent with the idea that without AcKR, bZiPro exists in the open conformation where much of the C-terminal part of NS2B is dissociated from NS3Pro and remains unstructured, thus resulting in the lack of NMR signal. Taken together, these observations suggest that thermal shift assay is unlikely to be of much help.

• *This paper suggests that ssRNA binds to the open conformation of ZIKV NS2B-NS3pro, however no experimental evidence, only modeling has been used to suggest binding to the open conformation. In Dengue virus protease, the M84P variant has been reported to lock the protease into the open conformation. How does the F84P variant of ZIKV NS2B-NS3pro impact ssRNA binding? *

• We appreciate this question. Indeed, M84P mutation shifts Dengue NS3Pro to the open conformation, which is proteolytically inactive12, consistent with our reverse inchworm model. We have not investigated the effect of this mutation on ZIKV NS3pro. We expect this mutation has a similar effect in ZIKV NS3pro in Dengue NS3Pro.

• The relevance of the discussion on the co-crystallization of NSC86314 with the Mut7was not clear. What point was being made?

• We provide a proof-of-principle for a novel class of allosteric inhibitors that specifically target newly identified druggable pockets present in the open and super-open conformations of ZIKV NS2B-NS3pro. Our results suggest that such allosteric inhibitors can interfere with the RNA-binding activities of NS2B-NS3pro in addition to blocking the protease activity. The co-crystallization of NSC86314 with the Mut7 confirms a novel pocked bound by NSC86314.

*- These data show that both active site and allosteric inhibitors block binding of ssRNA to the protease. The paper also suggests that ssRNA only binds to the open conformation. What is the evidence that the allosteric inhibitors do not enable or promote formation of the open conformation? *

• We thank this reviewer for an interesting question. Indeed, we have no evidence of whether allosteric inhibitors enable or promote the formation of the open conformation. This is formally possible and will need to be investigated.

• This paper makes two claims about the function of the protease. The title should specify what those dual functions are (proteolytic activity and ssRNA-recruitment).*

• We appreciate this reviewer's suggestions for the title.

• The discussion of Figures 6 and 9 are highly similar. The main takeaway points for both figures seem to be nearly identical: the presence of two positively charged pitchfork on the open conformation. The distinction between these two figures should be more significantly and explicitly stated. *
• Figure 6 presents several models that provide evidence for the open conformation of ZIKV NS2B-NS3pro being uniquely suitable to bind RNA. Figure 9 presents several models of the entire RNA-NS2B-NS3pro-NS3hel complex anchored into the ER membrane. Figure 9 illustrates that the open conformation of NS2B-NS3pro provides two positively charged/polar forks, contiguous with the positively charged groove on NS3hel. Figure 6 does not illustrate that point.

---

*- Mention explicitly in the materials and methods if the 12-amino acid linker is present in all the mutants used. *

• This is mentioned explicitly and shown in Supplementary Figure 2A.

Minor comments: ** · Figure 1. The rotation that promotes the transitions from orientation in panel A to that in panel B should be drawn. ** · FAM should be defined in the legend of Figure 2. ** · The term Cold should be changed to unlabeled. ** · Please check labels for the supplementary Figure 2. For example one label states 1-1 but it ** should be 1-170. ** · Figure 1C does not exist and it is referenced in the results section under "NS2B-NS3pro substrate-mimicking inhibitors compete with RNA binding." ** · As discussed above, if the super open conformation is going to be addressed in this paper, then either a reference for the manuscript describing those structures should be included, or this manuscript should include in the materials and methods the procedure on crystallization, data collection, structure determination, refinement, and analysis as well as a table for crystallographic data and refinement statistics. ** · Adjust figure arrangement (ABCED to ABCDE) in Figure 11.

• We thank this reviewer for all minor comments. We corrected the above-mentioned errors in the manuscript.

Reviewer #3 (Significance (Required)): ** It is well established that the flaviviral proteases exist in different conformations but most of the structures published are concentrated on the closed conformation which is the one required for effective substrate processing. The open conformation has recently been the subject of increased interest, especially with the discovery of allosteric inhibitors for which modeling suggests that these compounds result in the dissociation of the C-terminal region of NS2B from the NS3. This paper adds important insights into the function of the open conformation and in general implicitly shows the importance of the dynamic nature of ZIKV NS2B-NS3pro. In addition to these insights, this paper aptly demonstrates that ssRNA can bind and inhibit these proteases as has not been shown previously. ** I am a senior graduate student working on characterizing and understanding the mechanism of action of allosteric compounds against viral proteases, specifically proteases from Zika and dengue viruses.

References.

1. Weinert T, Olieric V, Waltersperger S, Panepucci E, Chen L, Zhang H, Zhou D, Rose J, Ebihara A, Kuramitsu S, Li D, Howe N, Schnapp G, Pautsch A, Bargsten K, Prota AE, Surana P, Kottur J, Nair DT, Basilico F, Cecatiello V, Pasqualato S, Boland A, Weichenrieder O, Wang BC, Steinmetz MO, Caffrey M, Wang M. Fast native-SAD phasing for routine macromolecular structure determination. Nat Methods. nature.com; 2015 Feb;12(2):131–133. PMID: 25506719
2. Solis AD, Rackovsky SR. Fold homology detection using sequence fragment composition profiles of proteins. Proteins. 2010 Oct;78(13):2745–2756. PMCID: PMC2933786
3. Brinkworth RI, Fairlie DP, Leung D, Young PR. Homology model of the dengue 2 virus NS3 protease: putative interactions with both substrate and NS2B cofactor. J Gen Virol. 1999 May;80 ( Pt 5):1167–1177. PMID: 10355763
4. Aleshin AE, Shiryaev SA, Strongin AY, Liddington RC. Structural evidence for regulation and specificity of flaviviral proteases and evolution of the Flaviviridae fold. Protein Sci. 2007 May;16(5):795–806. PMCID: PMC2206648
5. Phoo WW, Li Y, Zhang Z, Lee MY, Loh YR, Tan YB, Ng EY, Lescar J, Kang C, Luo D. Structure of the NS2B-NS3 protease from Zika virus after self-cleavage. Nat Commun. 2016 Nov 15;7:13410. PMCID: PMC5116066
6. Zhang Z, Li Y, Loh YR, Phoo WW, Hung AW, Kang C, Luo D. Crystal structure of unlinked NS2B-NS3 protease from Zika virus. Science. science.org; 2016 Dec 23;354(6319):1597–1600. PMID: 27940580
7. Luo D, Wei N, Doan DN, Paradkar PN, Chong Y, Davidson AD, Kotaka M, Lescar J, Vasudevan SG. Flexibility between the protease and helicase domains of the dengue virus NS3 protein conferred by the linker region and its functional implications. J Biol Chem. 2010 Jun 11;285(24):18817–18827. PMCID: PMC2881804
8. Chernov AV, Shiryaev SA, Aleshin AE, Ratnikov BI, Smith JW, Liddington RC, Strongin AY. The two-component NS2B-NS3 proteinase represses DNA unwinding activity of the West Nile virus NS3 helicase. J Biol Chem. 2008 Jun 20;283(25):17270–17278. PMCID: PMC2427327
9. Xu S, Ci Y, Wang L, Yang Y, Zhang L, Xu C, Qin C, Shi L. Zika virus NS3 is a canonical RNA helicase stimulated by NS5 RNA polymerase. Nucleic Acids Res. 2019 Sep 19;47(16):8693–8707. PMCID: PMC6895266
10. Klema VJ, Padmanabhan R, Choi KH. Flaviviral Replication Complex: Coordination between RNA Synthesis and 5’-RNA Capping. Viruses. 2015 Aug 13;7(8):4640–4656. PMCID: PMC4576198
11. Shiryaev SA, Aleshin AE, Muranaka N, Kukreja M, Routenberg DA, Remacle AG, Liddington RC, Cieplak P, Kozlov IA, Strongin AY. Structural and functional diversity of metalloproteinases encoded by the Bacteroides fragilis pathogenicity island. FEBS J. 2014 Jun;281(11):2487–2502. PMCID: PMC4047133
12. Lee WHK, Liu W, Fan JS, Yang D. Dengue virus protease activity modulated by dynamics of protease cofactor. Biophys J. 2021 Jun 15;120(12):2444–2453. PMCID: PMC8390872

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 #3

Evidence, reproducibility and clarity

This paper investigates the nucleic acid binding properties of zika virus protease. In particular the data suggest that single stranded RNAs and DNAs are capable of binding to and inhibiting ZIKV protease at micromolar concentrations. With the use of active site inhibitors and mutants that lock the protease in closed and super-open conformation, the authors concluded that RNA binds to the open conformation. Through extensive modeling of the protease and helicase domains, this manuscript provides a model of how ssRNAs can bind to all conformations of the proteas, but the open conformation provides two positively charged forks that should be available to bind RNA.

SECTION A - Evidence, reproducibility, and clarity

Major comments:

• The main conclusions of this paper rely on the existence of the super-open conformation, however this conformation has not been reported in the scientific literature previously. Structures deposited in the pdb are referenced in this manuscript, however no citation for an accompanying publication is provided. This calls into question the biological relevance of this super open conformation. This is of particular concern because in other highly-homologous flaviviral proteases, structures that have been observed crystallographically (e.g. the open conformation of dengue virus protease) appear to be only very sparsely populated in solution. What is the evidence that the super-open conformation exists in solution.
• The activity of each of the constructs used was not reported making it impossible to directly compare the impact of these changes on intrinsic activity. In particular, the NS2B-NS3 long construct is predicted to exist in the super-open conformation. If this is correct, it should show no activity against a peptide substrate.
• This paper reports that the IC50 is much weaker than the Kd for binding of ssRNA to ZIKV NS2B-NS3pro. Are orthogonal assays, such as thermal shift assay, available which could distinguish between the reported IC50 and the Kd.
• This paper suggests that ssRNA binds to the open conformation of ZIKV NS2B-NS3pro, however no experimental evidence, only modeling has been used to suggest binding to the open conformation. In Dengue virus protease, the M84P variant has been reported to lock the protease into the open conformation. How does the F84P varian of ZIKV NS2B-NS3pro impact ssRNA binding?
• The relevance of the discussion on the co-crystallization of NSC86314 with the Mut7was not clear. What point was being made?
• These data show that both active site and allosteric inhibitors block binding of ssRNA to the protease. The paper also suggests that ssRNA only binds to the open conformation. What is the evidence that the allosteric inhibitors do not enable or promote formation of the open conformation?
• This paper makes two claims about the function of the protease. The title should specify what those dual functions are (proteolytic activity and ssRNA-recruitment).
• The discussion of Figures 6 and 9 are highly similar. The main takeaway points for both figures seem to be nearly identical: the presence of two positively charged pitchfork on the open conformation. The distinction between these two figures should be more significantly and explicitly stated.
• Mention explicitly in the materials and methods if the 12-amino acid linker is present in all the mutants used.

Minor comments:

• Figure 1. The rotation that promotes the transitions from orientation in panel A to that in panel B should be drawn.
• FAM should be defined in the legend of Figure 2.
• The term Cold should be changed to unlabeled.
• Please check labels for the supplementary Figure 2. For example one label states 1-1 but it should be 1-170.
• Figure 1C does not exist and it is referenced in the results section under "NS2B-NS3pro substrate-mimicking inhibitors compete with RNA binding."
• As discussed above, if the super open conformation is going to be addressed in this paper, then either a reference for the manuscript describing those structures should be included, or this manuscript should include in the materials and methods the procedure on crystallization, data collection, structure determination, refinement, and analysis as well as a table for crystallographic data and refinement statistics.
• Adjust figure arrangement (ABCED to ABCDE) in Figure 11.

Significance

It is well established that the flaviviral proteases exist in different conformations but most of the structures published are concentrated on the closed conformation which is the one required for effective substrate processing. The open conformation has recently been the subject of increased interest, especially with the discovery of allosteric inhibitors for which modeling suggests that these compounds result in the dissociation of the C-terminal region of NS2B from the NS3. This paper adds important insights into the function of the open conformation and in general implicitly shows the importance of the dynamic nature of ZIKV NS2B-NS3pro. In addition to these insights, this paper aptly demonstrates that ssRNA can bind and inhibit these proteases as has not been shown previously.

I am a senior graduate student working on characterizing and understanding the mechanism of action of allosteric compounds against viral proteases, specifically proteases from Zika and dengue viruses.

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

The manuscript by Shiryaev et al., submitted to BioRXiv is an exploration of the ability of NS2B-NS3protease to bind RNA and its subsequent role in NS3 helicase processivity. The authors first utilize fluorescence polarization assays to demonstrate that NS2B-NS3protease can bind ssRNA with a strong affinity (and also ssDNA with lower affinity). They subsequently utilize mutational and small molecule inhibitor strategies in these assays to force the NS2B-NS3protease into different conformations, with the associated results inferring that the "open" conformation is responsible for ssRNA binding affinity. Furthermore, they demonstrate that ssRNA binding impairs protease activity, suggesting these roles may be exclusive in the viral life cycle. They also identified a number of small molecule ligands that target the putative ssRNA binding channel, and demonstrate that these ligands inhibit ssRNA binding by NS2B-NS3protease, providing potential inhibitor candidates for ZIKV. Finally, the authors utilized their crystal structures and others for the various conformations of NS2B-NS3protease to model ssRNA binding by the domain and the full NS3 protein, and used these models to propose a reverse inchworm model for NS3 travelling along ssRNA as it unwinds the dsRNA duplex. Overall, the authors utilize a comprehensive approach to demonstrate a number of novel findings (ssRNA binding by NS2B-NS3protease, small molecule ligands that inhibit this interaction) that would be of interest to both virologists and structural biologists. However, there are some important experimental design limitations and viral life cycle considerations that the authors should address before acceptance of the manuscript. Major and minor comments intended to improve the manuscript are outlined in more detail below.

Major Comments:

1. While the quantity of indirect data (ruled out closed and super-open, inhibitors of ssRNA binding pocket) suggest that the open conformation of NS2B-NS3protease is associated with ssRNA binding, the argument would be greatly strengthened by direct experimental data. Is there a mutational or small molecule approach to locking the NS2B-NS3 protease in the open conformation? If so, the authors should perform such experiments to strengthen the foundation of their argument.
2. A negative control should be used in Figure 4A to strengthen the claim that ssRNA binding in the open conformation impairs protease activity (ie. include a curve for dsRNA). Such an experiment would lend support to ssRNA inhibition being due to specific binding instead of some other non-specific effect of increasing local nucleic acid concentration.
3. Due to the highly coupled roles of NS5 and NS3 in replication, the authors should include some more consideration of the role of NS5 in their complex. They very briefly address this interplay in the fifth paragraph of the discussion, but then neglect to discuss the implications any further.

In particular (perhaps in a brief comparison to an NS3/NS5 modeling approach such as Brands et al., 2017; WIRES), the authors should consider some of the following questions: - could the channel on protease domain lead to ssRNA entry site on RdRp? - would NS5 interaction constrain or augment inchworm model of NS2B/NS3 translocation? - how does increased activity of NS3 when complexed with NS5 (Xu et al. 2019) align with proposed inchworm model?

Minor Comments:

1. Introduction, 4th paragraph, NS3-NS4 should read NS3-NS4A.
2. Throughout the manuscript, the authors should denote some key amino acid residues in each figure to help orient the reader better to the observed structural changes and rotations. Inclusion, at least in the supplement, of the crystal structures of mutants solved herein should also be included.
3. Section: RNA binding inhibits the proteolytic activity of ZIKV NS2B-NS3pro, last sentence, NS2N-NS3pro should be NS2B-NS3pro
4. Section: Allosteric inhibitors of NS2B-NS3 protease interfere with RNA binding- first sentence: "The open conformation of NS2B-NS3pro is achieved by the rearrangement of NS2B cofactor (its dissociation from the C-terminal half of NS3pro) leading to a loss of proteolytic activity [32]. - the reference is not correct. I could not find the reference the authors refer to here and had not heard before that NS2B cofactor was able to disassociate from the C-terminal half of NS3pro; hence, this really needs to be appropriately referenced.
5. Section: Modeling RNA binding to ZIKV NS2B-NS3, first sentence - unwinds should be unwind
6. With respect to the results of Figure 3A, the authors should address that adding the linker alone to the NS3 protease may not be an accurate examination of its role/importance. The linker in this scenario is only constrained at its N-terminus, while it is always constrained at both termini during infection (and even more so by the interactions of those two linked domains [protease and helicase] with each other). As such, the authors statement that "observations suggests that the 12-aa linker region modulates RNA binding to NS2B-NS3pro" should be more strongly qualified to this effect.

In addition, it would be interesting to see the effects of the linker mutations on ssRNA binding in the context of the full NS3 protein, albeit admittedly more complex due to the confounding ssRNA binding by the helicase domain. 7. The NS#hel should be changed to NS3hel in part (C) of figure legend for Figure 11. 8. The authors data in Figure 4A (and even more so the nature of the viral life cycle where 1000s of viral polyproteins are created from the first genome during infection) disputes the depiction in the inchworm model of how NS3 protease cleaves the polyprotein while the helicase binds ssRNA. At minimum, the authors need to discuss this discrepancy, and it is recommended that they modify the cartoon in their model to not include the ssRNA binding on the protease side of the equation (or show as alternative on that side to the existing cartoon). 9. In the third paragraph of the discussion, the authors state "An alternative model of coupled transcription and translation where viral RNA is associated with ribosomes right after the release from NS2B-NS3 is also possible". Considering there is abundant evidence that translation and replication are exclusive and that translation does not take place in ROs, it would be prudent to remove such statements from the discussion. Without any supporting evidence, these statements will be misleading to readers by providing a false equivalency. The preceding discussion of RFs would be sufficient to contextualize your inchworm model in the broader viral life cycle (which was done quite well). 10. There were a number of aspects I appreciated about the manuscript and will briefly list a few here:

- i) the focus on how different non-structural proteins effect the structure and function of

each other during the viral life cycle, which forms a more comprehensive and informative model - ii) the use of structural and functional assays as complementary approaches to studying the intra- and inter-protein relationships of NS3 - iii) the depiction of the forks in Figure 10, which effectively demonstrated the channels and oriented the reader to the conservation data - iv) the use of small molecule inhibitors to modify structure and function of NS3, which greatly deepened the richness of the story from both a basic and applied science view point

Significance

Strengths and limitations:

• provides some experimental and modeling data to provide a new model for RNA interactions with the NS3pro-hel; may help inform models for enzyme function, mostly consistent with previous literature
• leaves out the NS5 RdRp, known to contribute to NS3 activity.
• some suggestions are made which might strengthen the conclusions and inclusions of additional controls would improve the data.

Advance

• conceptual, perhaps may provide some insight into mechanism; although limited by the lack of NS5 RdRp which is crucial to helicase activity. It is unclear if the ssRNA would be oriented this way given interactions with NS5 RdRp and MT domains (is the ssRNA routed to NS5 or along NS3, or potentially are both happening?)

Audience:

• quite specialist, but may include structural biologists and virologist alike.

Expertise of the reviewer(s):

• molecular virologists, RNA viruses - including flaviviruses; replication complex biogenesis, protein-RNA and RNA-RNA interactions. While comfortable with the concepts regarding complex formation, the appropriateness of computational modeling and RNA docking tools as well as structural biology is out of our area of expertise.

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

The current manuscript by Shiryaev et al describes their observation of the new function of zika NS2B-NS3 proteases. They have shown that NS2B-NS3 protease lacking the helicase domain binds to RNA and the interaction can be affected by protease inhibitors. Main two new findings are presented in the manuscript: super open conformation of the protease; RNA binding activity of the protease region. although the manuscript is interesting, the design of the experiments is not convincing.

Major issues

1. the claim of a super open conformation is problematic. Using an artificial construct lacking the C-terminal portion of NS2B will of course generate the open conformation. This is a wrong definition unless you observe such a conformation in living cells.
2. It is odd that authors made homology model to generate open conformation structures. the authors did not cite the two papers of eZiPro (phoo et al 2016 NC) and bZiPro (Zhang et al 2016, Science). these two structures show the closed conformation of protease in the absence and presence of natural substrate.
3. RNA binding is novel, but is it observed in cells? only one method was used for testing the interactions, not other biophysical methods are used.
4. binding studies with RNA used artificial construct, how about the one with KTGR present like eZiPro. Keep in mind that the P1-P4 residues are present under native conditions.
5. authors built up nice models, it is great to consider the full length NS2B, but authors haven't taken into account the effect of NS2B on the open or closed conformation of the protease.

Significance

This manuscript shows the novel function of zika protease and conclude that protease binds to RNA. This is a novel finding, but the conclusion needs to be further confirmed, to avoid misinterpretations by future readers

It is great to introduce the conformational changes of a protease through defining open, closed, and super open conformations. But the definition was not carefully compared with current literatures. I am surprised that the two important papers are not cited. It is well known the G4SG4 linker affect the conformation of the protease, it is problematic to introduce the super-open conformation here. Authors need to show super open conformation is present in nature e.g. the model in which full length NS2B and NS3pro. RNA is a charged molecule under some conditions, NS3 also have charged residues, it is important to show whether the binding between RNA-protease is relevant to the function, or is this due to the application of the artificial constructs used in this study. Why so many mutants are used?

Other minors

1. Using a construct close to the native protease, at least the P1-P4 residues should be present. Using a peptide in the assay is also useful.
2. Test binding of RNA with protease using another method such as biophysical methods, or even gel shift assay
3. I don't know the correlation between Figure 7 and Figure 6. The authors describe ploy A binding to protease, while Figure 7 is talking about Helicase binds to dsRNAs.
4. I am glad to see the consideration of full length NS2B, NS3 in the models Figure 8, 9 and 11, but there is no data to support any of the model proposed.
5. Is the linker a ploy G not G4SG4?
6. Do the mutant sustain their protease activity?

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

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

The authors do not wish to provide a response at this time.

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 #3

Evidence, reproducibility and clarity

The authors ask the important question whether post-embryonic organ formation follows the same mechanisms as during embryonic development. They focus on neuromasts of the lateral line system in the caudal fin of medaka. They used live imaging to find that the post-embryonic caudal-neuromast-cluster develops from organ-founder neural stem cells as a bud (and not as individual migrating stem cells as in zebrafish) that detaches and migrates from the founding neuromast (P0). They show that the formation of post-embryonic neuromasts does not require the lateral line nerve, which establishes another difference to the process in zebrafish. Artificial reduction in Cxcr4b chemokine signalling slows down stem cell delamination, which invariably occurs at the anterior aspect of the P0-neuromast. They then show (via changes in cadherin-type gene expression and cellular imaging) that stem cell delamination from the P0 neuromast involves an epithelial-to-mesenchymal transition. Forcing this process in the entire neuromast accelerates new organ formation, but directionality is maintained. Finally they ask whether the stem cells required for this process are pre-specified or are generated from the neuromast by ablating the pre-bud stem cells. They find that other stem cells within the same organ rearrange and re-establish organ founder cells. The authors liken this new mechanism of organ formation to pathological metastases in humans and name it metastatic-like organogenesis.

Major comments

• In Fig. 2D both P0- and PE1-neuromasts appear with fewer hair cells, the reasons for this should be explained. Did the ablation also damage the primordium or is the pLL nerve required for complete neuromasts to form?
• p. 6: The reasoning behind generating a cxcr4b l-o-f mutant does not become clear, since a mutant already existed with a non-migratory primordium. Why did the authors expect their mutant would have a different phenotype?
• p. 7: The authors state that K15::cxcr7 larvae lack secondary embryonic neuromasts, but it seems from Figs. 3B,D that they might simply be delayed (note that the last one is missing in 3B). This delay may have been taken over from the delay in embryonic primordium formation of P0 and this result (as shown in Fig. 3) would not contradict the assumption that PE1 can form in the absence of cxcr4b signalling. I suggest that the authors actually show and quantify that K15::cxcr7 adults have fewer CNC neuromast numbers, because this seems to be the definitive proof that overexpressing the "sink" may be enough to reduce cxcr4b signaling to a level where its requirement for the formation of the PE1-neuromast can be assayed.
• Fig. 6A': The scheme is at least ambiguous and interpretation of it requires more supporting information: In A the stippled lines represent position within the neuromast, but what do they represent in A', numbers of BrdU+ cells? So what does e.g. the grey peak at the anterior mean - position or number of cells? Is the area inside the coloured lines important or the edge-points? Stage III is the only distribution with a clear left-bias, the others are centered (with a left-extreme for stage II), so what in the figure is the anterior proliferation peak? These are just some questions this reviewer had. Maybe the problem lies in the octagonal lines, meaning different things in both images? It is further unclear how the means given in the text can be derived from the figure. Maybe it would be best to try to represent the data with a heat map-overlay of the image in A, one for each stage?
• The authors propose two competing models regarding the origin of founder stem cells (p. 9, first sentence: early determination vs. in situ generation). In the third sentence again two scenarios are presented as to why experimentally prompting EMT does not trigger organ founder cell migration. This paragraph would benefit from stating more precisely which of these questions is addressed by the BrdU- and ablation experiments, together with a clearer statement at the end of that section as to which hypothesis in each case (origin and migration) is preferred.

Minor comments

• K15+ cells are described as neuromast stem cells and Fig. 1 suggests that these are the mantle cells: Please comment on the question whether all mantle cells are stem cells.
• p. 5: The reference (Seleit, Krämer et al., 2017) is ambiguous, as there are 2 references listed that fit the abbreviation.
• Fig.1B-F': Even though, as the authors state, there is variation in the timing of the budding process, it would be helpful to add an exemplay time frame to stages I-V.
• Fig. 4A'-D': The cell bodies of the support cells should have a distinct colour, otherwise they are easily confused with the nuclei of the other cell types. This would make it easier to understand the schematic at first sight.
• p. 9: T2A and H2A should be explained.
• Nuclei are shown protruding posteriorly in wildtype neuromasts (Fig. 5A-A'), while P0 neuromasts stem cells protrude anteriorly. Please explain the significance of the difference.
• Fig. 5 legend: Quantification is "E"; "and increased" should probably read "an increased"?
• p. 10: Unconventionally, Fig 7 is mentioned prior to Fig. 6B-C, I suggest combining both figures into one.

Significance

The manuscript describes a new mechanism of post-embryonic organ formation. Investigating how accesory neuromasts are formed during growth of juvenile medaka, the authors find that stem cells from a founder neuromast undergo epithelial-mesenchymal transition and migrate away directionally to form a complete new organ. This new mechanism is likened to that of cancer metastases.

Importantly, and different from zebrafish, this process is not dependent on innervation of the neuromast and is not a budding process, but relies on neural stem cells leaving the organ.

The interesting question posed by the authors, and answered here positively for accessory neuromasts in juvenile fish, is whether the mechanisms of organ formation differ between embryonic and post-embryonic development. The reported findings should be of interest to the stem cell communities and to researchers interested in post-embryonic development.

This review was written by a developmental biologist.

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

Authors study how new sensory organs of the lateral line form post-embryonically in medaka fish.

Evidence, reproducibility, and clarity

Overall, I find the study very preliminary, since hardly any conclusions are sufficiently backed by quantifications and statistical analysis as detailed below.

Major comments:

Figure 1: Authors describe events as "detached cells migrate" in the results text and in the legend as "detached stem cells...move". They should clarify how these images were exactly generated, since this has implication on whether they can conclude that cells migrate. Are these frames from movies in which the entire process was seen to occur in individual larvae? Are they derived from repeated imaging of the same larvae? Or are they images from different larvae and authors assume that the sequence of events occurs as presented? While it's likely that cell migration occurs, a wave of de novo transgene expression could also travel through the tissue. Thus, authors can only conclude that it's migration if they have seen cells move relative to their surroundings using live imaging or via lineage-tracing. e.g. using a photoconvertible fluorophore.

Page 5: "The process of post-embryonic neuromast formation in the CNC displays variability in timing". Authors should add information on the actual timing. This relates to issue 1, maybe they don't know, but if they did live imaging or repeated imaging of the same larvae they should have information on how long it takes from delamination of cells to new neuromast formation.

Figure 2A: where is the data showing that the laser ablation actually worked and supporting the statement "All ablated fish showed escaping cells from the P0-neuromast (N=7/7 fish)"?

Figure 2B-D: I find it difficult to see how these data support the conclusions the authors draw. First of all, it would be helpful to show the structures labeled by the individual transgenes for the non-initiated readers, so readers can judge themselves which of the signals are derived from the kremen and which from the eya1 transgene. Second, in B' readers are supposed to "Notice the uninjured primordium that continued migration and embryonic neuromast formation", yet where is the uninjured primordium? There are a lot of signals in the image. And how do we know that it "continued migration?" Relative to what? Finally in D, the label "PE1 neuromast" is printed in between a structure that is weakly and one that is strongly labeled. Do the authors suggest that there are 2 PE1 neuromasts here? Both the P0 and the P1 neuromast(s?) on the experimental side seem to be much smaller that on the control side, but authors do not mention this. Does this show that neuron proliferation or differentiation are nerve dependent?

Cxcr4 mutant: it's unclear why the authors mention the new mutant, albeit they don't use it. The only reason to justify this would be if the community were to benefit from this new allele. However, this requires that authors clarify how it differs from the previously existing one (e.g by showing the predicted protein sequence) and whether it has any advantages, e.g. is it more likely to be null? Further, where are the data supporting this statement? "cxcr4bD625 larvae display the same phenotype as the previously published cxcr4b mutants".

Figure 3: authors should validate the k15:cxcr7 line by showing cxcr7 (over) expression in k15+ cells using in situ hybridization.

Figure 3B: how do the authors distinguish single and double transgenics? If they cannot and only assume that those that have a phenotype are the double transgenic they need to confirm this by genotyping embryos post imaging.

Figure 3D: sample size is missing. Significance should be tested.

"Adult Tg(K15::Cxcr7) fish display significantly lower CNC neuromast numbers compared to wild type fish (N= 5 WT fish, 3.8 organs per CNC; N= 4 Tg(K15:Cxcr7), 2.2 organs per CNC". Which statistical test has been used to support the use of the term "significantly" in this statement? Images should be shown to support this conclusion.

Fig. 4C: authors should show YFP and CFP channels to allow readers to see why authors have false-colored 2 cells. What do they mean by the term "repolarization" that they put in the figure label?

Fig. 4D: "Subsequently, the organ-founder stem cells start elongating lamellipodia-like processes in the anterior direction". How do the authors know that this is "subsequently"? These are obviously not frames from a movie, so how do they know that the cells the point their arrow at delaminated from the neuromast considering that there are also quite a few other YFP+ cells in this frame?

Fig. 4C, D: These observations must be supported by a least rudimentary information on reproducibility. How many neuromasts have the authors analyzed? How often did they see this?

Fig. 4F, G: These conclusions must be supported by quantifications and statistics (e.g. E- and N-cadherin staining intensities).

Fig. 5: is the number of neuromasts with protruding cells significantly different between the control and experimental group? % should be rounded to significant figures.

Fig. 5D: these data are not convincing since it is unclear how authors identify the neuromast, considering that almost all cells seem to contain H2B-GFP.

Fig. 5E: is there a significant difference?

Fig. 6A: Exemplatory images of BrdU+ cells should be shown.

Fig. 6A: "This approach revealed a proliferation peak at stage I-II (mean: 6 and 5.4 BrdU+ cells, respectively". This statement does not support that this is a peak in the absence of information on how much proliferation there is at other stages. Why don't the authors show a graph plotting this with variation and statistical analyses to support that the numbers differ at different stages.

Fig.6A: also the conclusion that more proliferation happens in anterior positions must be backed by statistics.

Fig. 6: ablation data. Text mentions that PE1 formed in 7/8 ablated larvae, but images show differences between the control side, where PE1 has not yet formed and the ablated side where it has... Is this a non-representative image? Authors should clarify.

Fig. 7: Supplementary Movie 2 does not work, so I could not review it.

Fig. 7: Conclusions need to be supported by quantification and statistics, it's not sufficient to show only frames from one movie.

Fig. 7: Where is the data supporting this statement "(mean: 0.05 BrdU+ stem cells; N=8 P0-neuromasts)"? How was the BrdU incorporation experiment performed?

This statement needs to be supported by quantifications and statistical analysis: "We noticed that in most cases, the ablated P0-founder neuromast was considerably smaller than the non-ablated founder organ in the contralateral side (Fig.6 B',C)(N=4/7 larvae)"

This statement needs to be supported by quantifications and statistical analysis: The PE1-neuromast, however, reached the regular size in all cases, regardless of the status of the P0-founder neuromast.

Minor:

I think the author pitch the fact that they are examining "organ-founder stem cells" a bit too aggressively. It would be more appropriate to stick to the term "mantle" cells in all figures that describe data and reserve the "organ-founder stem cells" term to text where they interpret their results. It's particularly strange that Fig. 4C is labelled with such an interpretation.

Significance

There are many interesting open questions about post-embryonic development and in particular about how novel structures/organs form in those species where this happens. Thus, the overall research topic of this study is very interesting. The model is also well suited to derive novel mechanistic insight into these questions. The problem with the study in its current form is that I find it quite anecdotal, since hardly any conclusions are sufficiently backed by thorough data, in particular there is a rather shocking lack of quantifications and statistical analyses throughout. However, if the authors back their conclusions with such data, it'll certainly make a very interesting paper.

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

Groß and colleagues explore the mechanisms by which post-embryonic organogenesis occurs. To do so, they use medaka caudal fin neuromasts as a model, in which they follow the genesis of a new neuromast, PE1, from an existing neuromast, P0. Previously the group has established that ablation of the P0 neuromast results in the absence of PE1 neuromast formation, and lineage tracing confirmed that the P0 neuromast gives rise to all the neuromasts in the caudal-neuromast-cluster.

To dynamically assess PE1 formation, the authors use a broad lateral line transgene reporter, labeling keratin 15-expressing cells (K15+). Through this method, they observe that anterior K15+ cells of the P0 neuromast detach from the organ, migrate anteriorly, and give rise to the PE1 neuromast. By ablating the posterior lateral line (pLL) nerve prior to PE1 formation, which innervates the P0 neuromast initially, the authors report that the pLL is not necessary for proper PE1 formation, as it still develops in the absence of this nerve. Further, the authors explore chemokine signals which may underlie this process. The authors discuss a previous finding that the Cxcr4b/Cxcr7/Cxcl12a signaling pathway has been shown to regulate primordium migration, which is an organ required for neuromast formation. Therefore, the authors modulated the signaling pathway by first generating a missense cxcr4b mutant, only to find that this resulted in the absence of P0 formation. To circumvent this issue, the authors instead overexpressed Cxcr7, a non-signaling receptor that competes with Cxcr4b for the Cxcl12a ligand, in K15+ cells. Results from these experiments displayed a role for the Cxcr4b/Cxcr7/Cxcl12a signaling pathway in the temporal regulation of post embryonic neuromast formation. At the developmental time that wild type fish develop post embryonic neuromasts, Tg(K15:Cxcr7) animals have reduced numbers of post embryonic neuromasts. However, the authors find that post embryonic neuromasts do form in these animals at a later stage of development. Finally, the authors show throughout the report that the organ founder stem cells, which go on to form the PE1 neuromast, arise exclusively from the anterior side of the P0 neuromast. They questioned whether some cells were pre-specified to become organ-founder stem cells or if all stem cells have the same capacity to generate a new organ. To understand this, the authors used laser ablation on the anterior side of the P0 neuromast and found that the neuromast cells were capable of rearranging and initiating migration from the anterior side.

In all, this report highlights a potentially novel cellular mechanism by which a post embryonic neuromast is generated from a pre-established neuromast. By using longitudinal imaging, the authors were able to observe this process in developing animals and begin to probe its molecular mechanism. Several aspects of the manuscript should be strengthened prior to publication to more rigorously support the authors' model.

Major Comments:

A key aspect of the current manuscript is whether the morphogenetic process that they report is fundamentally different from the "budding" mechanism previously observed in zebrafish during neuromast stitch formation (Wada et al., 2010; Wada et al., 2013) or simply represents a different description of an analogous process in medaka. Inspection of the previous work in zebrafish reveals several similarities at the cellular level that leave this reviewer unconvinced that this is really a different process, rather than differences due to reporters used (membrane and nuclear here vs cytoplasmic previously). Specifically, Figures 2 and 3 in Wada et al., (2010) and Figure 2 in Wada et al., (2013) look quite similar to the morphogenetic process herein. Wada et al., (2013) describe stitch formation as: "a budding process that begins when a few cells elongate away from the founder neuromast". Here, on pg 4 Groß and colleagues describe a process that "is executed by a few organ-founder neural stem cells that detach from their original organ and migrate to generate a new organ remotely." How exactly are these two processes different? One key mechanistic difference is potentially the nerve dependence, but there are concerns with the interpretation of this experiment (see comment below). Unless the authors can definitively demonstrate that what they describe is a distinct process, they might be better served by reframing their results in light of previous findings by Wada and colleagues.

A second important thrust of the paper is that migrating K15+ cells act as "organ founder stem cells." However, the authors do not show conclusive evidence of K15+ cells acting as stem cells during PE1 formation. To support their claim, the authors should more rigorously define the K15+ organ-founder cells as stem cells. This could be done by several approaches (e.g., via lineage tracing, molecular analysis, or live-cell imaging). Importantly, it has previously been shown that zebrafish hair cell regeneration is predominantly driven by surrounding support cell proliferation (reviewed by ​​Lush and Piotrowski, 2014 doi: 10.1002/dvdy.24167). Therefore, it is important for the authors to determine whether their organ-founder cells are in fact stem cells or, alternatively, migrate along with K15- support cells from the P0 neuromast.

The rates of post-embryonic development in teleost fish, including medaka (Iwamatsu et al., 2003 doi: 10.2108/zsj.20.607), depend on several housing conditions (e.g., rearing density and feeding). It was surprising to this reviewer, therefore, that the authors performed all of their staging by days post-fertilization rather than standard length. Results from most experiments, especially those in Figures 1 and 3, would be more easily interpretable (and reproducible across different labs) if authors were to report standard length of the animals being used. For example, are Tg(K15:Cxcr7) animals smaller than their wildtype counterparts? If so, could this explain why PE1 neuromast formation is delayed? This goes for the interpretation of Figure 5E as well.

The authors should include statistical analyses of all their quantification. For example, it would be appropriate to use a statistical test to compare groups in Figures 3D and 5E.

The authors use the transgene Eya1:mCFP to visualize and laser ablate the pLL nerve, stating that Eya1 labels the nerve. However, a few sentences after introducing this transgene, the authors now use the transgene Kremen:mYFP to label the pLL nerve, and in double transgenic Kremen:mYFP; Eya1:mCFP fish, the Eya1 transgene is used to "assess the differentiation state of the putative newly formed PE1 neuromast". In the figure legend, the authors explain that Tg(Eya1:mCFP) labels the pLL in Figure 2A, but in Figure 2B it is used to visualize the primordium. The primordium is not the pLL nerve, and if the Eya1 transgene is labeling the primordium, and not the pLL (the text suggests it fulfills both roles in different scenarios) then the authors should repeat this experiment using the Kremen transgene to appropriately label the pLL nerve for accurate ablation. If, however, the Eya1 transgene labels both the primordium and the pLL (it is hard to tell from the double labeled images, as they are in grayscale), then this should be made explicitly clear.

Consider tempering the interpretation of the pLL nerve ablation experiment. Since the neurons and associated Schwann cells are still present following the severing of the nerve, albeit at a distance, could the nerve not still signal via a diffusible paracrine molecule? Alternatively, the authors could ablate the pLL ganglion. Since pLL ablation is performed 18 days prior to imaging, it may be possible that there is nerve regeneration occurring. To make the authors' findings more convincing, use of a secondary reporter, or antibody staining with a pan-neuronal antibody to confirm the absence of neuromast innervation at 21 dpf should be considered. Additionally, in Wada et al., (2013), neuromasts still extend cellular processes after nerve ablation. Since Wada et al. uses a membrane reporter, and here we see the use of a nuclear reporter, is it possible that we are seeing the same results as Wada et al.? To enhance the strength of this figure, authors should also consider using a membrane reporter or performing IHC similar to that in Figure 4D, so that readers can visualize there is no cellular process connecting the P0 and PE1 neuromasts in this condition.

Supplemental Figure 1 is a colorimetric in-situ hybridization image which depicts the localization of cxcl12 transcripts. The authors state that there is transcript expression in the vicinity of the P0 neuromast, however, the image is of poor quality and the expression does not appear spatially restricted to the anterior portion of the neuromast. Consider either discussing how a uniformly localized chemokine cue fits with their model and/or providing more detailed evidence of the expression pattern, e.g., by co-staining with a neuromast-specific marker and performing high-resolution imaging. Ideally the authors could quantify cxcl12 expression relative to the A/P axes of the P0 neuromast.

In Figure 4, the authors provide evidence that: 1) K15+ cells in the anterior of the P0 neuromast change shape and extend invasive, lamellipodia-like protrusions, and 2) that a cluster of cells to the anterior of the P0 neuromast show decreased E-cadherin and increased N-cadherin staining. These results are consistent with a subset of K15+ cells undergoing a MET. However, the authors could strengthen this portion of the manuscript in several ways. First, is it possible to live-image delamination from the neuromast? This would provide unambiguous evidence in support of their model. Second, is it possible to combine E- and N-cadherin staining with visualization of the K15+ population? The authors state that "N-cadherin is clearly upregulated in migrating organ-founder stem cells", but at the moment the evidence is circumstantial that K15+ cells switch cadherin expression. Co-visualization of the K15+ population would strengthen this point. Ideally, the authors could quantify E- and N-cadherin levels in K15+ cells relative to neighboring cells to support their claim. Minor point: have the E- and N-cadherin antibodies previously been validated in medaka? If so, the authors should cite the relevant work. If not, the authors should provide evidence of antibody specificity.

In Figure 5, the authors use Tg(K15:snail1b-T2A-H2A-mCherry) to address whether snail1b expression is sufficient to drive ectopic exit of cells from the P0 neuromast. The argument that snail1b is functional is an increase in neuromasts with protruding cells, but it is not demonstrated that any cells actually undergo EMT. A simple explanation for the observed lack of ectopic delamination is that snail1b is not expressed at sufficient levels from the transgene. With the current data, this reviewer would suggest tempering the claims related to interpretation of this experiment. On a related point, is snail1b normally expressed in delaminating cells? If so, this would provide further evidence to support the authors' EMT model in Figure 4. Minor point: the authors state that "not a single case of ectopic cell migration was observed when we analyzed the P0-neuromast by live imaging" - please clarify: what was the imaging time window and at what stage was the imaging performed?

Minor Comments:

Abstract, first sentence. This sentence is confusing since mammalian organs certainly grow at post-embryonic stages.

On page 6 of the manuscript, in the "pLL nerve is dispensable for organ-founder stem cell migration and PE1-neuromast formation" section, the authors describe the transgenic animals used in their pLL ablation studies as, "double Tg(Eya1:mCFP) - to label the pLL nerve - (K15:H2B-RFP) - to visualize migrating neuromast stem cells". The phrasing of the transgenic description is cluttered and confusing to read. Authors should consider rewriting this description as something such as, "double Tg(Eya1:mCFP); (K15:H2B-RFP) animals were used to visualize the pLL nerve as well as migrating neuromast stem cells, respectively".

On page 6 of the manuscript, in the "pLL nerve is dispensable for organ-founder stem cell migration and PE1-neuromast formation" section, the authors use the term escaping point. This is the first and only time the term is used, and it is not well defined. Presumably this refers to the anterior side of the P0 neuromast. This should be rewritten to more clearly articulate the meaning of this term.

Could the authors refer the reader to Figure 1A in the corresponding section of the introduction? It might also be helpful to the reader to label the primary vs secondary neuromasts on the schematic diagram.

In figure 2, the authors conclude that PE1 neuromast formation is not hindered by pLL ablation, however, in Figure 2C and D, it is apparent that the resulting PE1 neuromast post-pLL ablation is significantly smaller in size. The authors should address this, especially since they refer to the new PE1 neuromast in this condition as "mature". Is an organ mature if it is substantially smaller than in control conditions? Are the other resident cell types present in the proper proportions? Does the size of this organ grow to become wild-type further along in development?

On page 6, in the results section for Figure 2C and D, Eya1:GFP is being used to visualize "post-mitotic neurons" in the P0 and PE1 neuromasts, however, Figures 2C and 2D look more representative of the K15 reporter. Also this reviewer is not aware of post-mitotic neurons within neuromasts. The corresponding figure legend states that Tg(Eya1:EGFP) labels neuromast hair cells. Could the authors please clarify what is being labeled?

"Although specific for neural stem cells in the mature neuromast, the K15 promoter drives expression at earlier stages, after primary neuromasts were deposited by the primordium". This statement leads to confusion about the specificity of the K15 promoter, indicating it may be more broadly expressed than the authors state. The K15 promoter should be more rigorously described in the text, and evidence for its specificity should be clearly cited/provided.

The last sentence of the first paragraph of the Discussion is unnecessary and possibly overstating the findings within the report, as no evidence for "hijacking" of this post embryonic neuromast formation process was assayed.

The authors might consider discussing similarities and differences between their work and anchor cell invasion in C. elegans, which also involves post-embryonic organ remodeling by an invasive cellular behavior.

Methods transgene construction - please provide concentrations of nucleotides and proteins injected.

Methods Live-imaging section - "tranquilized" should probably read "anesthetized". More details on the imaging are needed. e.g., at what temperature was the imaging performed? What objective(s) was used?

Methods BrdU section - how were animals fixed? Please also describe the antigen retrieval step in detail.

It would be helpful for the supplemental movies to have labels for the transgenes, axes, and timestamps (as appropriate).

Supplemental Movie 2 shows significant xy movement between timepoints. Perhaps registration of the timepoints would help eliminate this and make the movie easier to interpret?

Significance

Groß and colleagues present an intriguing new model for post-embryonic morphogenesis of neuromasts in medaka. However, in its current state it is unclear whether these findings truly represent a new model for organ morphogenesis, rather than an alternative description of a previously described process. If it is the latter, the manuscript still has new cellular and molecular insights, but should be reframed. This work is likely to appeal to basic scientists, e.g., developmental biologists interested in organogenesis and neurobiologists interested in cell-cell interactions. This reviewer has expertise in teleost development and organogenesis.

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

Dear Editor and reviewers,

Thank you very much for the thorough assessment of our manuscript. We have carefully considered the comments and reflected most of them in the new version. We recognized the need to shorten and clarify the manuscript. Therefore, we have omitted particularly the less important passages concerning metabolism and the loss of genes encoding mitochondrial proteins, which cut the text by six pages in the current layout. We have also removed the text relating this model to eukaryogenesis. Finally, we have slightly changed the structure and linked the different sections to improve the flow of the story and to emphasize the key messages, which are the absence of mitochondria in a large proportion of oxymonads and the impact of this loss, loss of Golgi stacking and transformation to endobiotic lifestyle on selected gene inventories. We hope the manuscript is now clear and more concise and will be of interest to a broad readership interested in the evolution of eukaryotes, mitochondria and protists.

1. Point-by-point description of the revisions

Reviewer #1 (Evidence, reproducibility and clarity):

This is a very interesting paper that investigates through detailed comparative genomics the tempo and mode of the evolution of microbial eukaryotes/protists members of the Metamonada with a focus on Preaxostyla, currently the only known lineage among eukaryotes to have species that have lost, by all accounts, the mitochondria organelle all together. Notably, it includes a free-living representative of the lineage allowing potential interesting comparison between lifestyles among the Preaxostyla. This is a generally nicely crafted manuscript that presents well supported conclusions based on good quality genome sequence assemblies and careful annotations. The manuscript presents in particular (i) additional evidence for the common role of LGT from various bacterial sources into eukaryotic lineages and (ii) more details on the transition from a free-living lifestyle to an endobiotic one and (iii) the related evolution of MROs and associated metabolism.

Thank you very much for the positive assessment.

I have some comments to improve a few details:

In the introduction, lines 42-43, the last sentence should be more conservative by replacing "whole Oxymonadida" with "...all known/investigated Oxymonadida".

The sentence has been changed to: "Our results provide insights into the metabolic and endomembrane evolution, but most strikingly the data confirm the complete loss of mitochondria and every protein that has ever participated in the mitochondrion function for all three oxymonad species (M. exilis, B. nauphoetae, and Streblomastix strix) extending the amitochondriate status to all investigated Oxymonadida."

Similarly on line 62, the sentence could state "... contain 140 described...".

The sentence has been changed to: "Oxymonadida contain approximately 140 described species of morphologically divergent and diverse flagellates exclusively inhabiting digestive tracts of metazoans, of which none has been shown to possess a mitochondrion by cytological investigations (Hampl 2017)."

When discussing the estimated completeness of the genome are discussed (lines 117-120) and contrasted with the values for Trypanosoma brucei and other genomes, the author should explicitly state that these genomes are considered complete, which seems is what they imply, is that the case? If so, please provide more details to support this idea.

We have elaborated on this part also in reaction to comments of other reviewers. The text now reads: "It should be noted that, despite their wide usage, BUSCO values are not expected to reach 100% in lineages distant from model eukaryotes simply due to the true absence (or high sequence divergence) of some of the assessed marker genes. For example, various Euglenozoa representatives with highly complete genome sequences, including Trypanosoma brucei, have BUSCO completeness estimates in the range of 71-88% (Butenko et al. 2020), and representatives of Metamonada fall within the range of 60-91% (Salas-Leiva et al. 2021). Specifically in the case of oxymonad M. exilis, the improvement of the genome assembly using long-read resequencing from 2092 scaffolds to 101 contigs led to only a marginal increase of BUSCO value from 75.3 to 77.5 (Treitli et al. 2021). "

Also please see the detailed table prepared in response to reviewers 2 and 3 summarizing the presence/absence of genes from BUSCO set in the selected representatives of Metamonada and Trypanosoma brucei. The table is commented in the answer to Reviewer 3 comment (page 18)

The supplementary file named "132671_0_supp_2540708_rmsn23" is listed as a Table SX? (note: I found it rather difficult to establish exactly what file corresponds to what document referred in the main text)

We apologize for this mistake. We have checked and corrected references to tables, figures and supplementary material throughout the manuscript and hope it now does not contain any errors.

Lines 243-245, where 46 LGTs are discussed, it is relevant that the authors investigate their functional annotations. Indeed, it is suggested that these could have adaptive values, hence investigating their functional annotation will allow the authors to comment on this possibility in more details and precision. When discussing LGTs it would also be very useful to cite relevant reviews on the topic - covering their origins, functional relevance when known, distribution among eukaryotes. This is done when discussing the evolution and characteristics of MROs but not when discussing LGTs, with several reviews cited and integrated in the discussion of the data and their interpretation.

Available annotations of all putative LGT genes are provided in Supplementary_file_3 and also in the Supplementary_file_6 if the gene belongs to a manually annotated cellular system. Although we agree with the reviewer that the discussion of 46 species-specific LGTs might be interesting, for the sake of conciseness and brevity of the manuscript, we have decided not to expand the discussion further. However, note that we discuss selected cases of P. pyriformis-specific LGTs in the part “P. pyriformis possesses unexpected metabolic capacities” which follows right after the lines reviewer is referring to.

The sentence, lines 263-265, where the distribution of some LGTs are discussed, needs to be made more precise. When using the work "close" the authors presumably refer to shared/similar habitat,s or else? Entamoeba is not a close relative to the other listed taxa.

The “close relatives” mentioned in the text were meant as close relatives of all p-cresol-synthesizing taxa discussed in the paragraph, including Mastigamoeba, i.e. a specific relative of Entamoeba. We have modified the text such as to make the intended meaning easier to follow.

Lines 346-348, that sentence needs to end with a citation (e.g. Carlton et al. 2007).

The citation proposed by the reviewer has been added. The sentence was changed to: " The most gene-rich group of membrane transporters identified in Preaxostyla is the ATP-binding cassette (ABC) superfamily represented by MRP and pATPase families, just like in T. vaginalis (Carlton et al. 2007). "

In the paragraph (line 580-585) discussing ATP transporters, note that Major et al. (2017) did not describes NTTs but distantly related members of MSF transporter, shared across a broader range of organisms then the NTTs. Did the authors check if the genome of interest encoded homologues of these transporters too?

The citation has been removed; we admit that it was not the most appropriate one in the given

context. Concerning the NTT-like transporters, encouraged by the reviewer we searched for them in the Preaxostyla genome and transcriptome assemblies and found no candidates. This is not explicitly stated in the revised manuscript. The paragraph now reads: “MROs export or import ATP and other metabolites typically using transporters from the mitochondrial carrier family (MCF) or sporadically by the bacterial-type (NTT-like) nucleotide transporters (Tsaousis et al. 2008). We did not identify any homolog of genes encoding proteins from these two families in any of the three oxymonads investigated. In contrast, MCF carriers, but not NTT-like nucleotide transporters, were recovered in the number of four for each P. pyriformis and T. marina (Supplementary file 6).“

Line 920-921, I don't understand how the number 30 relates to "guarantee" inferring the directionality of LGTs events. This will be very much dataset dependent, 100 sequences might still not allow to infer directionality of LGT events. The authors probably meant to "increase the possibility to infer directionality".

We agree the original wording has not been particularly fortunate, so the sentence has changed to: "Files with 30 sequences or fewer were discarded, as the chance directionality of the transfer can be determined with any confidence is low when the gene family is represented by a small number of representatives."

Reviewer #2 (Evidence, reproducibility and clarity):

Using draft genome sequencing of the free-living Paratrimastix pyriformis and the sister lineage oxymonad Blattamonas nauphoetae, Novack et al. infer the metabolic potential of the two protists using comparative genomics. The authors conclude that the common oxymonad ancestor lost the mitochondrion/mitosome and discuss general strategies for adapting to commensal/symbiotic life-style employed by this taxon. Some elaborations on pathways go on for several paragraphs and feel unnecessarily stretched, which made those sections of the paper rather difficult to digest.

Having seen reflections on the manuscript by three reviewers we carefully reconsidered its content and attempted to make it shorter and more compact by removing some of the less substantial material. Namely, we have dispensed completely with the original last section of Results and Discussion (“No evidence for subcellular retargeting of ancestral mitochondrial proteins in oxymonads”) and made various cuts throughout other sections. We hope that the revised version makes a substantially better job of delivering the key messages of our study to the readers compared to the original submission.

This might be also be because the work, and all conclusions drawn, depend entirely on incomplete (ca. 70-80%) genome data and simple similarity searches, and e.g. no kind of biochemistry or imaging is presented to underpin the manuscripts discussion.

This is a very crude and superficial assessment of our data. We have actually good reasons to believe that the genome assemblies are close to complete. Please see the discussion on this topic below and an answer to a particular comment from reviewer 3 (page 18).

This is noteworthy in light of other protist genome reports published in the last few years that differ in this respect, including previous work by this group. And for sequencing-only data, this paper - https://doi.org/10.1016/j.dib.2023.108990 - might offer an example of where we are at in 2023.

Frankly, we do not think it is fair or relevant to compare our study to the paper pointed to by the reviewer, as that paper reports on a metagenomic study that delivers a set of metagenomically assembled genomes (MAGs) of varying quality retrieved from environmental DNA samples without providing any in-depth analysis of the gene content. Our study is very different in its scope and aims, and we are not certain what lesson we should take from this reviewer’s point. We have good reasons to believe that the datasets are close to complete. Please see the discussion on this topic below and answer to comment of reviewer 3 (page 18).

With respect to previous work of the group (Karnkowska et al. 2016 and 2019), this submission is very similar (analysis pattern, even some figures and more or less the conclusion), i.e. to say, the overall progress for the broader audience is rather incremental. Then there are also some incidents, where the data presented conflicts with the author‘s own interpretation.

It was our intention to use the previous analytical experiences and approaches, which at the same time makes the new results comparable with those published before. Although the format is intentionally similar, this work is a substantial step forward because only with our present study the amitochondrial status of the large part of Oxymonadida group can be considered solidly established. This in turn allows us to estimate the timing of the loss of mitochondrion (more than 100 MYA) demonstrating that the absence of mitochondrion in this group is not an episodic transient state but a long-established status. We do not understand what exactly the reviewer had in mind when pointing to “incidents, where the data presented conflicts with the author‘s own interpretation” – we are not aware of such cases.

The text (including spelling and grammar) needs some attention and the choice of words is sometimes awkward. The overuse of quotation marks ("classical", "simple", "fused", "hits", "candidate") is confusing (e.g. was the BLAST result a hit or a "hit").

The whole text has been carefully checked and the language corrected whenever necessary by a one of the co-authors, who is a native English speaker. The use of quotation marks has been restricted as per the reviewer’s recommendation.

In its current formn the manuscript is, unfortunately, very difficult to review. This reviewer had to make considerable efforts to go through this very large manuscript, mainly because of issues affecting to the presentation and the lack of clarity and conciseness of the text. It would be greatly appreciated if the authors would make more efforts upfront, before submission, to make their work more easily accessible both to readers and facilitate the task of the reviewers.

We admit that the story we are trying to tell is a complex one, consisting of multiple pieces whose integration into a coherent whole is a challenging task. As stated above, the reports provided by the reviewers provided us with an important stimulus, leading us to substantially modify the manuscript to make it more concise, less ambiguous when it comes to particular claims, and easier to read. We hope this intention has been fulfilled to a larger degree.

About a fifth of the two genome is missing according the authors prediction (table 1). Early on they explain the (estimated) incompleteness of the genomes to be a result from core genes being highly divergent. In light of this already suspected high divergence, using (the simplest NCBI) sequence similarity approach to call out the absence of proteins (for any given lineage) may need lineage-specific optimization. The use of more structural motif-guided approaches such as hidden Markov models could help, but it is not clear whether it was used throughout or only for the search for (missing) mitochondrial import and maturation machinery. The authors state that the low completeness numbers are common among protists, which, if true, raises several questions: how useful are then such tools/estimates to begin with and does this then not render some core conclusions problematic? The reader is just left with this speculation in the absence of any plausible explanation except for some references on other species for which, again, no context is provided. Do they have similar issues such as GC-content, same core genes missing, phylogenetic relevance?, etc.. No info is provided, the reader is expected to simply accept this as a fact and then also accept the fact that despite this flaw, all conclusions of the paper that rests on the presence/absence of genes are fine. This is all odd and further skews the interpretations and the comparative nature of the paper.

The question of the completeness of the data sets was raised also by reviewer 3 and we would like to provide an explanation at this point. First, it should be stated that there is no ideal and objective way how to measure the completeness of the eukaryotic genomic assembly. In the manuscript, we have used the best established method, adopted by the community at large, which is based on the search for a set of „core eukaryotic genes“ using a standardized pipeline BUSCO or previously popular CEGMA. The pipeline uses its own tools to identify the homologues of genes/proteins which ensures standardization of the procedure. This answers the question of reviewer 2, why we have not used more sensitive tools for these searches. We did not use them, because we followed the procedure that is the gold standard for such assessments, for comparability with other genomes and to make this as clear to the reader as possible. Although the result of the pipeline is usually interpreted as the completeness of the assembly, this is a simplification. Strictly speaking, the result is a percentage of the genes from the set of 303 core eukaryotic genes (in our case) which were detected in the assembly by the pipeline. Even in complete assemblies, the value is usually below 100% because some of the genes are not present in the organism and some diverged beyond recognition. We do not see any other way how to deal with this drawback than to compare with related complete genome assemblies acting as standards. This we have done in Supplementary file 11, where we list the presence/absence of each gene for Preaxostyla species and three highly complete assemblies of Trypanosoma brucei, Giardia intestinalis and Trichomonas vaginalis. T. brucei and G. intestinalis are assembled into chromosomes. As you can see, in these three „standards“ 63, 148 and 77 genes from the core were not detected resulting in BUSCO completeness values of 79%, 51% and 75%, respectively. 18 of the non-detected genes function in mitochondria (shown in red), which are highly reduced in some of these species, so the absence of the respective genes is therefore expected. Simply not considering these genes would increase the “completeness measure” for oxymonads by 6%. The values for our standards are not higher than the values for Preaxostyla (69-82%). In summary, the BUSCO incompleteness measure is far from ideal, particularly in these obscure groups of eukaryotes. The values received for Preaxostyla give no reason for concern about their incompleteness. See also our answer to reviewer 3 (page 18).

At the same time, we admit that the BUSCO values do not confirm the high completeness of our assemblies. So, why do we think they are highly complete? One reason is that we do not see suspicious gaps in any of the many pathways which we annotated but the main reason is the high contiguity of the assemblies. Thanks to Nanopore long read sequencing, the assembly of P. pyriformis and B. nauphoetae compose of 633 and 879 scaffolds, suggesting that there are “only” hundreds of gaps. Although this may still sound too much, it is a relatively good achievement for genomes of this size and the experience shows that a further decrease in the number of scaffolds would allow the detection of additional genes but not in huge numbers. As we have shown for M. exilis (Treitli et al. 2021, doi:10.1099/mgen.0.000745) the decrease from 2 092 scaffolds to 101 contigs, i.e., filling almost 2 000 gaps, allowed the prediction of additional 1 829 complete gene models, of which 1 714 were already present in the previous assembly but only partially and just 115 were completely new. None of these newly predicted genes was functionally related to the mitochondrion. Thus, we infer the chance that all mitochondrion-related genes are hidden in the gaps of assemblies is very low.

We have provided these arguments in a condensed form in the text following the description of genome assemblies: “It should be noted that, despite their wide usage, BUSCO values are not expected to reach 100% in lineages distant from model eukaryotes simply due to the true absence (or high sequence divergence) of some of the assessed marker genes. For example, various Euglenozoa representatives with highly complete genome sequences, including Trypanosoma brucei, have BUSCO completeness estimates in the range of 71-88% (Butenko et al. 2020), and representatives of Metamonada fall within the range of 60-91% (Salas-Leiva et al. 2021). Specifically in the case of oxymonad M. exilis, the improvement of the genome assembly using long-read resequencing from 2092 scaffolds to 101 contigs led to only a marginal increase of BUSCO value from 75.3 to 77.5 (Treitli et al. 2021).”

As a side note, this will also influence the number of proteins absent in other lineages and as such has consequences on LGT calls versus de novo invention. For the cases with LGT as an explanation, it would help to briefly discuss the candidate donors and some details of the proteins in the eco-physiological context (e.g. lines 263-268 suggest that HPAD may have been acquired by EGT which was facilitated by a shared anaerobic habitat and also comment on adaptive values for acquiring this gene). Exchanging metabolic genes via LGT (Line 163) blurs the differences between roles and extent of LGT in prokaryote vs eukaryote, and therefore is exciting and could use support/arguments other than phylogenies. I guess the number of reported LGTs among protists (whatever the source) over the last decade has by now deflated the novelty of the issue in more general; a report of the numbers is expected but they alone won't get you far anymore in the absence of a good story (such as e.g. work on plant cell wall degrading enzymes in beetles).

We agree with the reviewer that the cases of LGT involving Preaxostyla would deserve more discussion in the manuscript. On the other hand, we also agree that none of them provides such a “cool” story that would deserve a special chapter or even a separate paper. Therefore, we have decided, also with regard to keeping the text in a reasonable dimension, not to expand the discussion of LGTs with the exception of HgcAB, where some new information has been included and the phylogeny of the genes updated. Please note that we had discussed in the original manuscript the donor lineages and ecological/biochemical context in the cases of GCS-L2, HPAD, UbiE, and NAD+ synthesis and this material has been kept also in the revised version.

It would help to clarify which parts of the mitochondrial ancestor were reduced during the process of reductive evolution at what time in their hypothesized trajectory. For instance, loosing enzymes of anaerobic metabolism conflicts with the argued case of an aerobic (as opposed to facultative anaerobic) mitochondrial ancestor followed by gains of anaerobic metabolism in the rest of the eukaryotes via LGT, and some papers the authors themselves cite (e.g. the series by Stairs et al.). There is no coherent picture on LGT and anaerobic metabolism, although a reader is right to expect one.

These are very interesting questions, that would fill a separate article. In the manuscript, we focus on the Preaxostyla lineage only and there the trajectory seems relatively simple: replacement of the mitochondrial ISC by cytosolic SUF in the common ancestor of Preaxostyla, loss of methionine cycle and in in consequence mitochondrial GCS and the mitochondrion itself. We have modified the first conclusion paragraph in this sense and it now reads the following:

“The switch to the SUF pathway in these species has apparently not affected the number of Fe-S-containing proteins but led to a decrease in the usage of 2Fe-2S clusters. The loss of MRO impacted particularly the pathways of amino acid metabolism and might relate also to the loss of large hydrogenases in oxymonads. “

It is not clear to us how to understand the reviewer’s remark concerning the conflict between loss of enzymes of anaerobic metabolism and the (presumed) aerobic nature of the mitochondrial ancestor. Provided that we read the reviewer’s rationale correctly, is it really so implausible that the anaerobic metabolism gained laterally by a particular lineage was then secondarily lost in specific descendant lineages? As a clear example demonstrating the feasibility of such an evolutionary pattern consider the evolution of plastids. There is no doubt these organelles move across eukaryotes by secondary or higher-order endosymbiosis or kletoplastidy, establishing themselves in lineages where there was no plastid before. Secondary simplification of such plastids, e.g. by the loss of photosynthesis, in its extreme form culminating in the complete loss of the organelle, has been robustly documented from several lineages, such as Myzozoa (e.g., https://pubmed.ncbi.nlm.nih.gov/36610734/). Hence, we see absolutely no reason to rule out the possibility that the ancestral mitochondrion was obligately aerobic and enzymes of anaerobic metabolism spread secondarily by eukaryote-to-eukaryote LGT, with their secondary loss in particular lineages. We really do not see any conflict here and we do not agree with the interpretation provided by the reviewer. That said, we admit that the discussion on the earliest stages of mitochondrial evolution is not an essential ingredient of the story we try to tell in our manuscript, so to avoid any unnecessary misunderstanding we have removed the original last sentence of Conclusions (“Thorough searches revealed …”) from the revised manuscript.

In light of their data the authors also discuss the importance of the mitochondrion with respect to the origin of eukaryotes:

First, the mitochondrion brought thousands of genes into the marriage with an archaeon, surely hundreds of which provided the material to invent novel gene families through fusions and exon shuffling and some of which likely went back and forth over the >billion years of evolution with respect to localizations. The authors look at a minor subset of proteins (pretty much only those of protein import, Fig. 6) to conclude, in the abstract no less: „most strikingly the data confirm the complete loss of mitochondria and every protein that has ever participated in the mitochondrion function for all three oxymonad species." I do not question the lack of a mitochondrion here, but this abstract sentence is theatrical in nature, nothing that data on an extant species could ever proof in the absence of a time machine, and is evolutionary pretty much impossible. A puzzling sentence to read in an abstract and endosymbiont-associated evolution.

We feel that the reviewer is putting too much emphasis on an aspect of our original manuscript that is rather peripheral to its major message. Indeed, the manuscript is not, and has never been thought to be, primarily about eukaryogenesis and the exact role the mitochondrion played in it. We are, therefore, somewhat reluctant to react in full to the very long and complex argument the reviewer has raised in his/her report, so we keep our reaction at the necessary minimum. Concerning the criticized sentence in the original version of the abstract, it alluded to a section of the manuscript (“No evidence for subcellular retargeting of ancestral mitochondrial proteins in oxymonads”) that we have removed from the revised version, and hence we have modified also the abstract accordingly by removing the sentence. We still think our original arguments were valid, but apparently, much more space and more detailed analyses are required to deliver a truly convincing case, for which there is no space in the manuscript.

Second, using oxymonads as an example that a lineage can present eukaryotic complexity in the absence of mitochondria and conflating it with eukaryogenesis is a logical fallacy. This issue already affected the 2019 study by Hampl et al.. We have known that a eukaryote can survive without an ATP-synthesizing electron transport chain ever since Giardia and other similar examples and the loss of Fe-S biosynthesis and the last bit of mitosome (secondary loss) doesn't make a difference how to think about eukaryogenesis. It confuses the need and cost to invent XYZ with the need and cost of maintenance. How can the authors write "... and undergo pronounced morphological evolution", when they evidently observe the opposite and show so in their Fig. 1? The authors only present evidence for reductive evolution of cellular complexity with the loss of a stacked Golgi. What morphological complexity did oxymonads evolve that is absent in other protists? A cytosolic metabolic pathway doesn't count in this respect, because it is neither morphological, nor was it invented but likely gained through LGT according to the authors. This is quite confusing to say the least. A recent paper (https://doi.org/10.7554/eLife.81033) that refers to Hampl et al. 2019 has picked this up already, and I quote: "Such parasites or commensals have engaged an evolutionary path characterized by energetic dependency. Their complexity might diminish over evolutionary timescale, should they not go extinct with their hosts first." Here the authors raise a red flag with respect to using only parasites and commensals that rely on other eukaryotes with canonical mitochondria as examples. If we now look at Fig. 1 of this submission, Novak et al. underpin this point perfectly, as the origin of oxymonads is apparently connected to the strict dependency on another eukaryote (or am I wrong?), and they support the prediction with respect to complexity reducing after the loss of mitochondria - mitosome gone, Golgi almost gone. What's next? This is a good time to remember that extant oxymonads are only a single picture frame in the movie that is evolution, and their evolution might be a dead-end or result in a prokaryote-like state should they survive 100.000s to millions of years to come.

It seems that in this point the reviewer is particularly concerned with the following sentence that is part of the Introduction and which relates to the existence of amitochondrial eukaryotes we are studying: “The existence of such an organism implies that mitochondria are not necessary for the thriving of complex eukaryotic organisms, which also has important bearings to our thinking about the origin of eukaryotes (Hampl et al. 2018).“ Even after re-reading the sentence we confess we stay with it and find it perfectly logical. Nevertheless, we decided to omit it from the text so as not to distract from the main topic of the study.

Next, when mentioning “… pronounced morphological evolution” we mean the evolution of four oxymonad families (Streblomastigidae, Oxymonadidae, Pyrsonymphidae and Saccinobaculidae) comprising almost a hundred described species with often giant and morphologically elaborated cells that evolved from a simple Trimastix-like ancestor (Hampl 2017, Handbook of Protists, 0.1007/978-3-319-32669-6_8-1). This is a fact that can hardly be dismissed. Also, given the current oxymonad phylogenies (Treitli et al. 2018, doi.org/10.1016/j.protis.2018.06.005) and the reported absence of a mitochondrion in M. exilis, B. nauphoetae, and S. strix we can infer that the mitochondrion was lost in the common ancestor of the three species at latest. This organism must have lived more than 100 MYA, as at that time oxymonads were clearly diversified into the families (Poinar 2009, 10.1186/1756-3305-2-12). So, these organisms indeed have lived without mitochondria for at least 100 MY. We think that these facts and our inferences based on them are solid enough to keep in the conclusion the following statement: “This fact moves this unique loss to at least 100 MYA deep past, when oxymonads had been already diversified (Poinar 2009), and shows that a eukaryotic lineage without mitochondria can thrive for eons and undergo pronounced morphological evolution, as is apparent from the range of shapes and specialized cellular structures exhibited by extant oxymonads (Hampl 2017).” Furthermore, as documented in Karnkowska et al. 2019 (https://pubmed.ncbi.nlm.nih.gov/31387118/), apart the loss of the mitochondrion oxymonads are surprisingly “normal” and complex eukaryotes, in fact much less reduced than, e.g., Giardia, Microsporidia, or even S. cerevisiae (in terms of the number of genes, introns, etc.). We strongly disagree with the claim that “Golgi is almost gone” in oxymonads, and our manuscript shows exactly the opposite. Viewing oxymonads as a lineage heading towards a prokaryote-like simplicity is dogmatic and ignores the known biology of these organisms.

Some more thoughts: Line 47-52: Hydrogenosome or mitosome is a biological and established label as (m)any other and I find the use of the word "artificial" in this context strange. While the authors are correct to note that there is a (evolutionary) continuum in the reduction - obviously it is step by step - they exaggerate by referring to the existing labels as "artificial". You make Fe-S clusters but produce no ATP? Well, then you're a mitosome. It's a nomenclature that was defined decades ago and has proven correct and works. If the authors think they have a better scheme and definition, then please present one. Using the authors logic, terms such as amyloplast or the TxSS nomenclature for bacterial secretions systems are just as artificial. As is, this comes across as grumble for no good reason.

We agree that the original wording sounded like unwarranted grumbling and we have changed the sentence in the following way: "However, exploration of a broader diversity of MRO-containing lineages makes it clear that MROs of various organisms form a functional continuum (Stairs et al. 2015; Klinger et al. 2016; Leger et al. 2017; Brännström et al. 2022)."

Line 158: A duplication-divergence may also explain this since sequence similarity-based searches will miss the ancestral homologues.

We do not disagree about this, in fact, the gene the reviewer’s point is concerned with for sure is a result of duplication and divergence, as it belongs to a broader gene family (major facilitator superfamily, as stated in the manuscript) together with other distant homologs. Nevertheless, this is not in conflict with our conclusion that it “may represent an innovation arising in the common ancestor of Metamonada”.

Lines 201-202: Presence of GCS-L in amitochondriate should be explained in light of this group once having a mitochondrion, which then makes ancestral derivation and differential loss (as invoked for Rsg1) also a likely explanation along with eukaryote-to-eukaryote LGT.

Yes, this most likely holds for the standard paralogue GCS-L1 (in P. pyriformis PAPYR_5544), which has the expected distribution and phylogenetic relationships and is absent in oxymonads. The discussion is, however, mainly about the rare, divergent and until now overlooked paralogue GCS-L2 (in P. pyriformis PAPYR_1328), which we found only in three distantly related eukaryote groups, Preaxostyla, Breviatea, and Archamoebae, which strongly suggests inter-eukaryotic LGT.

Lines 356-392: Describes plenty of genomic signal for Golgi bodies but simultaneously cites literature suggesting the absence of a morphologically an identifiable Golgi in oxymonads. An explicit prediction regarding what to observe in TEM for the mentioned species might be nice to stimulate further work.

We thank the reviewer for their suggestion and are glad that they are enthusiastic about this aspect of the manuscript. Unfortunately, the morphology of unstacked Golgi ranges from single cisternae (yeast, Entamoeba), vesicles (Mastigamoeba), and a “tubular membranous structure” in Naegleria. Therefore, no strong prediction is possible of what the oxymonad Golgi might look like under light or TEM. However, the data that we have provided should lead to molecular cell biological analyses aimed at identifying the organelle, giving target proteins to tag or against which to create antibodies as Golgi markers. An additional sentence to this effect has been added to the manuscript, “They also set the stage for molecular cell biological investigations of Golgi morphological variation, once robust tools for tagging in this lineage are developed.”

Lines 414: The preceding paragraphs in this result section describes only the distribution, without mentioning origins - a sweeping one-line summary that proclaims different origin needs some context and support. Furthermore, the distribution of glycolytic enzymes might indeed be patchy, but to suggest it represents an 'evolutionary mosaic composed of enzymes of different origins' without discussing the alternative of a singular origin and different evolutionary paths (including a stringer divergence in one vs. another species) discredits existing literature and the authors own claim with respect to why BUSCO might fail in protists.

The part of the text about glycolysis the reviewer alluded to has been removed while shortening the manuscript.

Line 486: How uncommon are ADI and OTC in lineages sister to metamonada?

This is an interesting but difficult question. Firstly, we are uncertain what is the sister lineage to Metamonada. Discoba, maybe, but a recent unpublished rooting of the eukaryotic tree does not support it (https://pubmed.ncbi.nlm.nih.gov/37115919/). Generally, the individual genes of the pathway (ADI, OTC and CK) are quite common in eukaryotes, but the combination of all three is rare (Metamonada, the heterolobosean Harpagon, the green algae Coccomyxa and Chlorella, the amoebozoan Mastigamoeba, and the breviate Pygsuia), see figure 1 in Novak et al 2016, doi: 10.1186/s12862-016-0771-4.

Line 504: It might help an outside reader to include a few lines on consequences and importance of having 2Fe-S vs 4Fe-S clusters and set an expectation (if any) in Oxymonads.

We apologize for omitting this explanation. The 2Fe-2S proteins are more common in mitochondria where 2Fe-2S clusters are synthesized in the early pathway of FeS cluster assembly, while the cytosolic CIA pathways produce 4Fe-4S clusters (https://pubmed.ncbi.nlm.nih.gov/33007329/). The original expectation therefore is that species without mitochondria should not have 2Fe-2S cluster proteins. Obviously, the switch to the SUF pathway affects this expectation as we do not know, what type of cluster this pathway produces in oxymonads (https://www.biorxiv.org/content/10.1101/2023.03.30.534840v1). For the sake of brevity, we have included a short statement as the beginning of the sentence in question, which now reads as follows: “As 2Fe-2S clusters are more frequent in mitochondrial proteins, the higher number of 2Fe-2S proteins in P. pyriformis compared to the oxymonads may reflect the presence of the MRO in this organism.”

Any explanations on what unique selection pressures and gene acquisition mechanisms may be operating in P. pyriformis which might allow for the unique metabolic potential?

Every species exhibits a unique combination of traits that results from changing selection pressures imposed on historical contingency (including neutral evolutionary processes such as genetic drift). We lack real understanding of these factors for a majority of taxa including the familiar ones, so we should not expect to have a good answer to the reviewer’s question. In fact, we do not know how unique is the particular combination of P. pyriformis traits discussed in our manuscript, as there has been no comprehensive comparative analysis that would include ecologically and evolutionarily comparable taxa. We note that Paratrimastix represents only a third free-living metamonad with a sequenced genome (together with Kipferlia and Carpediemonas), so more data and additional analyses are needed to be in a position when we may start hoping answers to questions like the one posed by the reviewer are in reach.

** Referees cross-commenting** To R3: Hampl et al. 2019, to which Novak et al. refer, is about eukaryogensis and that is exactly the context in which this is discussed again and what Raval et al. 2022 had decided to touch upon. If the authors do not bring this up in light of the ability to evolve (novel) eukaryote complexity, then what else? Maybe they can elaborate, especially with respect to energetics to which they explicitly refer to in 2019 (and here). And with respect to text-book eukaryotic traits (and the evolution of new morphological ones), I do not see any new ones evolving in any oxymonad, but reduction as Novak et al. themselves picture it in this submission. Is a change in the number of flagella pronounced morphological evolution? Maybe for some, but I believe this needs to be seen in light of the context of how they discuss it. I see a reduction of eukaryotic complexity and not a gain. They have an elaborate section on the loss of Golgi characteristics (and a figure), but I fail to read something along the same lines with respect to the gain of new morphological traits. Again, novel LGT-based biochemistry does not equal the invention of a new morphology such as a new compartment. Oxymonads depend on mitochondria-bearing eukaryotes for their survival or don't they? This is the main point, and if evidence show that I am wrong, then I will be the first to adapt my view to the data presented.

While we do see the logic of the reviewer’s point, a good reply would have to be too elaborate and certainly beyond the scope of the current manuscript. As the reviewers’ reports led us to reconsider the structure of the manuscript and to make it more focused and concise, we decided to simplify the matter by removing the allusions to eukaryogenesis, realizing that it is perhaps more suitable for a different type of paper (opinion, review). The comment on the evolution of complex morphology has been answered previously (see above).

I have concerns with the presentation of a narrative that in my opinion is too one-sided and that has been has been publicly questioned in the community (in press, at meetings, personally). For the benefit of science and of the young authors on this study, this reviewer feels strongly that these issues should be taken very seriously and discussed openly in a more balanced way. . We only truly move forward on such complex topics, if we allow an open and transparent discussion.

We agree that opinions on specific details of eukaryogenesis are divided in the community and that the topic requires a nuanced discussion for which there is perhaps no place in the current manuscript. As stated in the reply to the previous point, we have removed the discussion of the implications of our current study to eukaryogenesis from the revised manuscript.

Having said that, I am happy that R3 has picked up exactly the same major concerns as I did with respect to e.g. the phrasing on mito (gene) loss and the BUSCO controversy.

We appreciate these comments and hopefully have resolved the concern in the previous answers.

Reviewer #2 (Significance):

Using draft genome sequencing of the free-living Paratrimastix pyriformis and the sister lineage oxymonad Blattamonas nauphoetae, Novack et al. infer the metabolic potential of the two protists using comparative genomics. The authors conclude that the common oxymonad ancestor lost the mitochondrion/mitosome and discuss general strategies for adapting to commensal/symbiotic life-style employed by this taxon. Some elaborations on pathways go on for several paragraphs and feel unnecessarily stretched, which made those sections of the paper rather difficult to digest. This might be also be because the work, and all conclusions drawn, depend entirely on incomplete (ca. 70-80%) genome data and simple similarity searches, and e.g. no kind of biochemistry or imaging is presented to underpin the manuscripts discussion.

We have addressed the concern about the possible incompleteness of our genome data above, demonstrating it is not substantiated ad stems from an inadequate interpretation of quality measures we provide in the manuscript. We hope that the revised manuscript, which is streamlined and more concise compared to the initial submission, conveys the key messages in a substantially more persuasive way and will be appreciated by a broad community of readers.

Reviewer #3 (Evidence, reproducibility and clarity):

Summary: The genome sequences of two members of the protist group Preaxostyla are presented in this manuscript: Paratrimastix pyriformis and Blattamonas nauphoetae. The authors use a comparative genomics and phylogenetic approaches and compare the new genome datasets with three previously available genomes and transcriptomes from the group. The availability of genome-scale data from five Preaxostyla species is powerful to address interesting basic evolutionary questions. A substantial part of the manuscript is spent on testing the hypothesis of mitochondrial loss in the oxymonad lineage, which turns out to be supported. The datasets are also explored regarding the role of lateral gene transfer in the group, metabolic diversification and the evolution of Golgi.

Major comments: I find the manuscript very interesting with many different fascinating results presented. However, the manuscript is very long. Two genome sequences are presented and it is not clear to me what the main question was when this project was initiated and why these two species was selected to answer this question. I do not see an obvious reason for sequencing the P. pyriformis genome if the mitochondrial loss was the main question (given that a transcriptome was already available). Why not spend the time and resources on a member of Preoxystyla, which lacked previous data? The authors should more clearly state why these organisms were chosen to answer the main question or questions of the study.

We are sorry for having done a poor job when explaining the choice of the taxa for the comparison. The idea was to sample an outgroup of oxymonads (P. pyriformis) and a representative of other clades of oxymonads than M. exilis (B. nauphoetae and S. strix) for which it was feasible to obtain the data, or the data were already available. Obviously, more representatives of morphologically a probably also genetically diverse oxymonads should be investigated (e.g. Pyrsonympha, Oxymonas, Saccinobacullus) and we have such a plan but these organisms are difficult to work with. We considered it necessary to sequence the genome of P. pyriformis, and not rely on the transcriptome only, to avoid the issue of data set incompleteness (raised also by R2). Transcriptomes by nature provide an incomplete coverage of the full gene complement of the species, while our genome assemblies are close to complete, as we explain elsewhere.

The evolution of MROs have received substantial attention from the protist research community since the 1990's. During this period the mitochondrial organelle have been considered essential for eukaryotes. Therefore, the result presented in the manuscript has a high significance. However, I am not convinced that it is appropriate to use the term "evolutionary transition" for the mitochondrial loss. The loss of MRO is the endpoint of a gradual change of the internal organisation of the cell that probably started when the ancestor of these organism adapted to an anaerobic lifestyle. The last step described in the manuscript probably had little impact on how these organisms interacted with their environment. The presence or absence of biosynthesis of p-cresol by some, but not all, Preaxystyla probably is much more significant from an ecological point of view. My point is that the authors need to consider how they use the term evolutionary transition and be explicit about that.

We appreciate the comment concerning the use of the term “evolutionary transition”. Nevertheless, we believe there is no real consensus in the literature on what is and what is not an “evolutionary transition”, and the application of the term to specific cases is more or less arbitrary. For a lack of a standardized or better terminology, we have kept the term to refer to three evolutionary changes in the evolution of the Preaxostyla lineage that are particularly important from the cytological or ecological perspective, i.e. dispensing with the mitochondrion, reorganizing the Golgi apparatus by losing the stacked arrangement of the cisternae, and gaining the endobiotic life style.

In the abstract the main finding is describes as "the data confirm the complete loss of mitochondria and every protein that has ever participated in the mitochondrion function for all three oxymonad species (M. exilis, B. nauphoetae, and Streblomastix strix) extending the amitochondriate status to the whole Oxymonadida.". I find this a really interesting observation, but I do find the wording a bit too bold for several reasons: • Not every protein that has participated in the mitochondrial function is known. • Mitochondrial proteins could be present in oxymonads, but divergent beyond the detection limit for existing methods. • Genes for one or several mitochondrial proteins could be present in one or more oxymonad genomes, but remain undetected due to the incomplete nature of the datasets.

Although I do think that the authors' claim very well could be true, I don't think their data fully support it. Therefore, it needs to be rephrased.

As a result of our decision to streamline the manuscript by removing the final part of Results and Discussion (“No evidence for subcellular retargeting of ancestral mitochondrial proteins in oxymonads”, the revised manuscript no longer support the statement “the data confirm the complete loss of … every protein that has ever participated in the mitochondrion function for all three oxymonad species” that is criticized by the reviewer, and hence the statement has been removed from the abstract. This addresses bullet point 1. As for bullet points 2 and 3, the proof of absence is in principle impossible to deliver, and we have been fighting with this already in the Karnkowska et al. 2016 paper. Although our certainty will never reach 100% (this is in fact impossible for a scientific, i.e., falsifiable, hypothesis), the mounting of evidence through studies gives the hypothesis on the amitochodriate status of oxymonads more and more credit. The genes for mitochondrial marker proteins have not been detected by the most sensitive methods available neither in the first genome assembly of M. exilis (Karnkowska et al. 2016), nor in the improved M. exilis genome assembly composed of only 101 contigs (Treitli et al. 2021), nor in either of the other two oxymonad species investigated here. On the other hand, they were readily detected in the data sets of P. pyriformis and T. marina. What is the probability that these genes always hide in the assembly gaps, or that they have all escaped recognition? Obviously, this probability is not zero, but we believe it is approaching so low values that it is reasonably safe to make the conclusion on the amitochondriate status of these species.

The sentence was changed to: "Our results provide insights into the metabolic and endomembrane evolution, but most strikingly the data confirm the complete loss of mitochondria for all three oxymonad species investigated (M. exilis, B. nauphoetae, and Streblomastix strix), suggesting the amitochondriate status may be common to Oxymonadida."

The third point maybe could be analysed further. BUSCO scores are reported, but also argued not being reliable for this group of organisms (which is true). Would it, for example, be useful to analyse how large fraction of the BUSCO proteins found in all non-Preoxystyla metamonada genomes that are present in the various Preoxystyla datasets?

We provide a comprehensive answer to a similar comment of reviewer 2 above (page 6-8). We performed the requested analysis and provide the result in Supplementary file 11. In this table, we record presence/absence of each gene from the BUSCO set for our data sets and the highly complete “standard” datasets of Trypanosoma brucei, Giardia intestinalis and Trichomonas vaginalis. Of the 303 genes, 117 were present in all data sets and 17 in none (see column I). 20 were present only in Trypanosoma and not in metamonads. 6 were present in all Preaxostyla and absent in other metamonads (Trichomonas and Giardia), 44 were present in all Preaxostyla and Trichomonas and absent in Giardia, suggesting high divergence of this species. Only 23 (marked by *) were present in the three “standard” genomes and absent in one or more Preaxostyla species. Of those 8 and 8 were absent specifically in S. strix and P. pyriformis, respectively, but only 1 was absent specifically in M. exilis and no such case was observed in B. nauphoetae. We conclude that this non-random pattern argues for lineage-specific divergence rather than incomplete data sets, particularly in the case of M. exilis and B. nauphoetae.

Line 160-161: 15 LGT events specific for the Preaxostyla+Fornicata clade is reported. This is an exciting finding because it supports a phylogenetic relationship between these two groups. But such an argument is only valid if the observed pattern is more common than the alternative hypotheses (Preaxostyla+Parabasalids and Fornicata+Parabasalids). How many LGT events support each of these groupings? How are these observation affected by the current taxon sampling with the highest number of datasets from Fornicata? How were putative metamonada-to-metamonada LGTs treated in this context?

19 LGT are uniquely shared between Preaxostyla+Parabasalids, which is more than the number of shared LGTs between Preaxostyla and Fornicata. No common LGT was unique to Fornicata+Parabasalids. However, the latter is a direct consequence of our investigation method, which involved reconstruction phylogenies of genes present in Preaxostyla, and not across all metamonads. So, we do not have a way to investigate LGT gene families uniquely shared between Fornicata and parabasalids.

When it comes to the effect of taxon sampling, we agree that it is possible that the number of genes of horizontal origin shared between parabasalids and Preaxostyla is underestimated because of the lower taxon sampling in parabasalids. However, it is still larger (19) than the number of LGTs shared uniquely between fornicate and Preaxostyla (15). In addition, while the taxon sampling is larger in fornicates, it also contains some representatives of closely related lineages (e.g., Chilomastix caulleryi and Chilomastix cuspidate) which, while they increase the number of fornicate representatives, does not increase the detection of shared genes between fornicates and Preaxostyla. Altogether, it's difficult to estimate how the current taxon sampling is biasing the detection of LGTs one way or another.

Regarding metamonad-to-metamonad putative LGTs: we did not consider this possibility for the sake of not overestimating the number of gene transfers for two main reasons. First of all, our LGT detection relies on the incongruence between species tree and gene tree. The closer the lineages are in the species tree, the more difficult it is to interpret any incongruence in the gene tree as single protein phylogenies are notoriously poorly resolved because they rely on the little phylogenetic signal contained in few amino-acid positions. Because of this, small incongruences with the species tree could either reflect recent LGT events between metamonads, or simply blurry phylogenetic signal. Second, we can certainly use the argument that a limited taxonomic distribution among metamonads favors an LGT event between them. However, here again, the closer the lineages involved are, the more difficult it is to distinguish a scenario where one lineage was the recipient of an LGT from prokaryote before donating it to another metamonad, from a scenario involving a single ancestral LGT from prokaryotes to metamonads, followed by differential loss, leading to a patchy taxonomic distribution. Finally, we are working with both limited taxon sampling and incomplete genomic/transcriptomic data, which makes it more difficult to identify true absences. For all these reasons, we chose to be conservative and invoke the smallest number of LGT events.

The authors have used a large-scale approach to make single-gene trees for inferences of LGT. In other parts of the manuscript inferences of evolutionary origins of single genes are made without support of phylogenetic trees. I find this inconsistent and argue that the hypothesis of the origin of a specific protein should be tested with the same rigor whether it is a putative LGT, gene duplication, gene loss or an ancestral member of LECA. Specific cases where I think a phylogenetic analysis is needed includes: • Line 222-223: It is concluded that Rsg1 is a component of LECA. • Line 307: HgcAB are argued to be acquired by LGT of a whole opeon. • Lines 350-355: It is unclear how the different numbers of transporters are interpreted (loss or expansion by duplication). This could be address with phylogenetics. • Lines 407-408: A tree should support the claim of LGT origin. (PFP) • Lines 414-415: The different origins of glycolytic enzymes should be supported by data or references. • Line 486: Trees or a reference (if available) should support the claim for LGT.

As requested, trees were constructed for HgcA, HgcB, PFP and the transporters AAAP, CTL, ENT, pATPase, and SP. Citations were added for the glycolytic enzymes and the ADI pathway. No tree for Rsg1 is needed, as this is a eukaryote-specific protein lacking any close prokaryotic relatives. The inference on its presence in the LECA is based on the phylogenetically wide, however patchy, distribution across the eukaryote phylogeny. Testing possible eukaryote-eukaryote LGTs is hampered by a limited phylogenetic signal in the short and rapidly evolving Rsg1 sequences, resulting in very poorly resolved relationships among Rgs1 sequence in a tree we attempted to make (data not shown). For this reason, we opt for not presenting any phylogenetic analysis for Rsg1.

Lines 530-531 and 773-774: "The switch to the SUF pathway in these species has apparently not affected the number of Fe-S-containing proteins but led to a decrease in the usage of 2Fe-2S clusters." I find it difficult to evaluate if the data support this because no exact numbers or identities are given for 2Fe-2S and 4Fe-4S proteins in the various genomes in Suppl. Fig. S4 or Supplementary file 4.

The functional annotation of all detected FeS clusters containing proteins is provided in Supplementary Table S8 including the types of predicted clusters (columns G or F). Basically, the only putative 2Fe2S cluster containing proteins in species of oxymonad is xanthine dehydrogenase, while Paratrimastix and Trimastix contain also 2Fe2S cluster-containing ferredoxins and hydrogenases.

The method used in the paper varies between the different parts of the paper. One example is single gene phylogenies, which are described three times in the method section [Lines 959-973, lines 1011-1034, lines 1093-1101], in addition to the automated approach within the LGT detection pipeline lines 923-926]. The approaches are slightly different with, for example, different procedures for trimming. This makes it difficult to know how the different presented analyses were done in detail. No rationale for using different approaches is given. At the least, it should be clear in the method section which approach was used for which analysis.

The reviewer is correct, and we apologize for the inconsistency. The reason is only historical –the analyses were performed by different laboratories in different periods of time. We believe this fact does not make our results less robust, although it does not “look” nice and makes the description of the methods employed longer. We have double-checked the description and introduced slight changes as to make it maximally clear which method has been used for particular analyses presented in the Results and Discussion.

Specific comments on single gene phylogenies:

• Line 966-967: Why max 10 target sequences?

The limit of 10 was applied in order to keep the datasets in manageable dimensions. The sentence has been changed to: " In order to detect potential LGT from prokaryotes while keeping the number of included sequences manageable, prokaryotic homologues were gathered by a BLASTp search with each eukaryotic sequence against the NCBI nr database with an e-value cutoff of 10-10 and max. 10 target sequences.”

• Lines 996-998: Is it a problem that these are rather old datasets?

Although the publications are slightly older the set of queries is absolutely sufficient for the purpose.

Minor comments: I appreciate that many data is included as supplementary material. However, the organisation of the data could be improved. The numbering of the files is not included in their names or within the files, as far as I could find. Descriptions of the files are often missing and information on the annotation such as colour coding is not always included. These aspects of the supplementary material needs to be strengthened in order to make it more useful. Specific comments: • Supplementary file 1, Table 1: accession numbers are missing. Kipferlia bialta appears to have a much smaller number of sequences than reported in the publication. The file consists of three tables and it would be very helpful if the reference in the main manuscript indicate the table number. • Supplementary file 4: The trees lack proper species names and a documented colour coding. There are multiple trees in the file, which make it difficult to find the correct tree. I would appreciate if the different trees were labelled A, B, C, etc., and if these were used in the main text.

Supplementary file 1: Accession numbers were added.

Supplementary file 4: Species names and alphabetical labelling were added. Colour coding was explained in the text at the first mention of the file: "(Supplementary file 4 H; Preaxostyla sequences in red)."

o There is no HPAD-AE tree (as indicated on line 258), but a HPAD tree. Which part of the tree contain the described fusion protein?

Thank you for spotting the mistake. There should have been “HPAD” instead of “HPAD-AE” indicated in the text. The sentence has been changed to:" The P. pyriformis HPAD sequence is closely related to its homolog in the free-living archamoebid M. balamuthi (Supplementary file 4 K), the only eukaryote reported so far to be able to produce p-cresol (Nývltová et al. 2017)."

o Line 280-281: "UbiE homologs occur also in some additional metamonads, including the oxymonad B. nauphoetae and certain fornicates." These sequences should be clearly highlighted in the tree.

We discovered these additional UbiE homologs only after the tree presented in the supplement had been constructed, so these sequences are missing from it. To ensure consistency we have decided to remove the remark on the presence of UbiE homologs metamonads other than P. pyriformis, so it is no longer part of the revised manuscript.

o Lines 538-544: A three-gene system is mentioned, but only two AmmoMemoRadiSam trees are found.

This part has been removed while streamlining the manuscript.

• Supplementary file 6: I find it difficult to find the proteins discussed in the text, for example "the biosynthesis of p-cresol from tyrosine (line 254-255)".

Abbreviations identifying the different enzymes have now been added to all mentions in the text, facilitating their localization in the supplementary file: "P. pyriformis encodes a complete pathway required for the biosynthesis of p-cresol from tyrosine (Supplementary file 6), only the second reported eukaryote with such capability. This pathway consists of three steps of the Ehrlich pathway (Hazelwood et al. 2008) converting tyrosine to 4-hydroxyphenyl-acetate (AAT, HPPD, ALDH) and the final step catalyzed by a fusion protein comprised of 4-hydroxyphenylacetate decarboxylase (HPAD) and its activating enzyme (HPAD-AE)."

• Supplementary file 11: Which group of species are highlighted in red? How do I know from which species these sequences are (I can make educated guesses, but prefer full species names). I do not find any reference to this file in the main manuscript.

We apologise for this inconvenience. The taxon labels in the treed in this supplementary file have been corrected to contain full species names.

Line 227-228: "630 OGs seem to be oxymonad-specific or divergent, without close BLAST hits". It is unclear if BLAST searches includes only a representative of each 630 OGs, or every single protein in these OGs.

The BLAST searches include every single protein in the investigated OGs. We clarified it in the text: “Of these, 630 OGs seem to be oxymonad novelties or divergent ancestral genes, without close BLAST hits (e-value -15) to any of these sequences.”

Line 243: I think it is five LGT mapped to internal nodes of Preoxystyla in Figure 1 (1+3+1).

You are correct, we apologize for the mistake. The sentence has been changed to: "Also, 46 LGT events were mapped to the terminal branches and 5 to internal nodes of Preaxostyla, suggesting that the acquisition of genes is an ongoing phenomenon, and it might be adaptive to particular lifestyles of the species."

Lines 325-331: The argument would be stronger with a figure showing the fusion and the alignment indicating the conserved amino acids mentioned in the text.

We agree with the reviewer but for the sake of space, we finally decided not to include a new figure.

Lines 425: "none of the species encoded" should be replaced by something like "none of the enzyme could be detected in any of the species" (the datasets are incomplete).

The sentence has been changed to: "None of the alternative enzymes mediating the conversion of pyruvate to acetyl-CoA, pyruvate:NADP+ oxidoreductase (PNO) and pyruvate formate lyase (PFL), could be detected in any of the studied species."

Line 455: "suggesting a cytosolic localization of these enzymes in Preaxostyla." The absence of a phylogenetic affiliation with the S. salmonicida homolog does not preclude a MRO localisation.

The sentence was changed to: "Phylogenetic analysis of Preaxostyla ACSs (Supplementary file 4 B) shows four unrelated clades, none in close relationship to the S. salmonicida MRO homolog, consistent with our assumption that these enzymes are cytosolic in Preaxostyla."

Lines 570-571: "Manual verification indicated that all the candidates recovered in oxymonad data sets are false positives" Using which criteria?

The manual verification was based on the annotation of predicted proteins by BLAST and InterProScan. If the annotations did not correspond to the suggested function, they were considered false positives. For example, the protein BLNAU_15573 of Blattamonas nauphoetae was detected by Sam50 HMM profile and thus was considered a candidate for Sam50 proteins. Its functional annotation from BLAST was, however, unrelated to Sam50 (“putative phospholipase B”). Therefore, this candidate was concluded as a false positive hit of the HMM search resulting from the very high sensitivity of this method.

We clarified this in the Results

“Reciprocal BLASTs indicated that all the candidates recovered in oxymonad data sets are very likely to be false positives based on the annotations of their top BLAST hits (mainly vaguely annotated kinases, peptidases and chaperones) (Fig. 6, Supplementary file 9).”.

And Material and Methods

“Any hits received by the methods described above were considered candidates and were furter inspected as follows. All candidates were BLAST-searched against NCBI-nr and the best hits with the descriptions not including the terms 'low quality protein', 'hypothetical', 'unknown', etc. were kept. For each hit, the Gene Ontology categories were assigned using InterProScan-5.36-75.0. If the annotations received from BLAST or InterProScan corresponded to the originally suggested function, the candidates were considered as verified. Otherwise, they were considered as false positives.”

Lines 743-755: "Similar observations were made in other protists with highly reduced mitochondria, such as G. intestinalis or E. histolytica,..." References are needed.

This part of the manuscript has been removed while streamlining the text.

Line 849: How was the manually curation done for the gene models in the training set?

The sentence has been changed to: "For de novo prediction of genes, Augustus was first re-trained using a set of gene models manually curated with regard to mapped transcriptomic sequences and homology with known protein-coding genes."

Lines 853-856: It is a bit unclear which dataset was used for BUSCO and downstream analysis. Was it the Augustus-predicted proteins, or the EVM polished?

The sentence has been changed to: "The genome completeness for each genome was estimated using BUSCO v3 with the Eukaryota odb9 dataset and the genome completeness was estimated on the sets of EVM-polished protein sequences as the input."

Lines 858: What is it meant that KEGG and similarity searches was used in parallel (what if both gave a functional annotation?)?

A sentence has been added for clarity: "KEGG annotations were given priority in cases of conflict."

Lines 861-862 and 1007-1008: Which genes or sub-projects does this apply to? How many genes were detected in this procedure?

The sentence has been changed to make this clear: "Targeted analyses of genes and gene families of specific interest were performed by manual searches of the predicted proteomes using BLASTp and HMMER (Eddy 2011), and complemented by tBLASTn searches of the genome and transcriptome assemblies to check for the presence of individual genes of interest that were potentially missed in the predicted protein sets (single digits of cases per set). Gene models were manually refined for genes of interest when necessary and possible."

Lines 878-879: It is not clear to me why the sum of the two described numbers should be as high as possible and would appreciate an argument or a reference.

When optimizing the inflation parameter of OrthoMCL, we reasoned that the optimal level of grouping/splitting for our purpose should result in the highest number of orthogroups containing all representatives of the groups of interest (i.e. Preaxostyla) but no other species – pan-Preaxostyla orthogroups. When going down with the values, you observe more and more groupings of pan-Preaxostyla OGs with others (indication of overgrouping) in the opposite direction you observe splitting of pan Preaxostyla OGs which indicates oversplitting. Because we were optimizing the inflation parameter for Preaxostyla and Oxymonadida at the same time, we maximized the sum of pan-Preaxostyla and pan-Oxymonadida groups.

Lines 879-881: "Proteins belonging to the thus defined OGs were automatically annotated using BLASTp searches against the NCBI nr protein database (Supplementary file 1)." Why were these annotated in a different way (compare lines 857-859).

This little inconsistency resulted from the fact that these parts of the analyses were performed by different researchers who did not cross-standardize the procedures. This inconsistency has no effect on the downstream analyses and conclusions as the annotations from Supplementary file 1 were not used in any further analyses.

Lines 894-957: "Detection of lateral gene transfer candidates": • It is not clear which sequences were tested in the procedure. All Preaxostyla, or all metamonada? I think I am confused because in the result sections you only report numbers for Preaxostyla, but in the method section metamonada is mentioned repeatedly.

Thank you for noticing. There was indeed some inconsistency in our writing.

We did an all-against-all search using all metamonads. However, we filtered out all homologous families in which Preaxostyla were not present or that had no hit against GTDB. So in the end, the LGT search was restrained to protein families containing Preaxostyla homologues. We corrected the wording in our method section.

• It would be easier to follow the procedure if numbers are provided for the different steps.

We are not sure what numbers the reviewer refers to here.

• Why was only small oxymonad proteins discarded (line 900)?

This is indeed a mistake. We meant “Preaxostyla proteins”. This is because we only considered Preaxostyla sequences with significant hits against GTDB as a starting point, so we aimed to first remove those that might be too short to yield reliable phylogenies.

• Line 911: How many sequences were collected?

Up to 10,000 hits were retained. We have added that information to the text.

• Lines 916-919: What is the difference between the protein superfamilies (line 916) and the OGs (line 919)? Are the OGs the same orthogroups that is described earlier in the method section? How are the redundancy of NCBI nr entries retrieved in different searches dealt with?

We understand the confusion here. It primarily stemmed from two different ways to establish homologous families across the manuscript because of different researchers being responsible for different parts. Protein superfamilies that were used for reconstructing the single protein trees used for the LGT analyses were assembled based on the procedure describe line 916-919 (“Protein superfamilies were assembled by first running DIAMOND searches of all metamonad sequences against all (-e 1e-20 --id 25 --query-cover 50 --subject-cover 50). Reciprocal hits were gathered into a single FASTA file, as well as their NCBI nr homologues.”). However, this was a somewhat stricter procedure than the one used to establish the OGs that are discussed in the rest of the manuscript (because of the e-value and identity cut-off used), so we eventually enriched the datasets with the putatively missing metamonad sequences that were present in the OGs but not in the initial superfamily assembly. However, since these were often more divergent sequences, we did not use these as queries for our BLAST searches against prokaryotes.

Line 987-989: "...was facilitated by Rsg1 being rather divergent from other Ras superfamily members" This statement is vague. What does it mean in practise?

The sentence has been changed to: " The discrimination was facilitated by Rsg1 having low sequence similarity to other Ras superfamily members (such as Rab GTPases)."

Lines 1037-1038: Why were these proteins re-annotated?

They were not. We are sorry for this mistake, which has been fixed in the revised manuscript.

Figures: The figures would be easier to follow if the colour coding for the five different species were consistent between the figures.

This is a good point, the colour coding has been unified across all figures.

Figure 1: It appears that the Venn diagram in C only shows the Preaxostyla-specific protein in B, not all OGs for which contain Preaxostyla proteins. This is not clear from legend or from the figure itself. The same comment applies to D.

The interpretation of the figure by the reviewer is correct; we have modified the legend to make the meaning of the figure easier to understand.

Figures 2 and 6: It would be clearer with panel labels A, B, etc, instead of "upper" and "lower" panel, as in the other figures.

This is a fair point, we have added the alphabetical labels proposed by the reviewer to the figures.

Figure 6: What is the colour code in the figure? The numbers within the boxes are not aligned.

We have added an explanation of the color code to the legend and edited the figure to make it aesthetically more pleasing.

Supplementary figures 1-3: What do green and magenta indicate in the figure?

As with the previous figure, the color code is now explained in the revised legend.

** Referees cross-commenting** I agree with the other reviewers that the discussion of the functional and ecological implications of the LGTs could be developed.

We understand the reviewers but as already explained in response to Reviewer 1, we have decided not to extend the already rather long manuscript further. We believe that the several exemplar LGT cases that we do discuss in detail provide a good impression of the significance of LGT in the evolution of Preaxostyla.

In contrast to reviewer 2, I do not see that the authors discuss their result in the context of eukaryogenesis in this manuscript. Maybe the reference reviewer 2 mention could be cited in the introduction together with Hampl et al. 2018 to acknowledge that there are different views about the importance of secondarily amitochondrial eukaryotes on our thinking about the origin of eukaryotes. I disagree with reviewer 2's objection against the wording "... and undergo pronounced morphological evolution" because I think Fig. 4 in Hampl 2017 shows a large morphological diversity among oxymonads.

We are glad to see that our perspective is not shared by other colleagues in the field. Nevertheless, having carefully considered the case we have decided to remove any mentions of eukaryogenesis from the revised manuscript, as we admit this topic is peripheral to the key message of our present study. On the other hand, we appreciate very much the note by the reviewer on the large morphological diversity among oxymonads – we have now added a similar remark to the revised manuscript (the last sentence of Conclusions).

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 #3

Evidence, reproducibility and clarity

Summary:

The genome sequences of two members of the protist group Preaxostyla are presented in this manuscript: Paratrimastix pyriformis and Blattamonas nauphoetae. The authors use a comparative genomics and phylogenetic approaches and compare the new genome datasets with three previously available genomes and transcriptomes from the group. The availability of genome-scale data from five Preaxostyla species is powerful to address interesting basic evolutionary questions. A substantial part of the manuscript is spent on testing the hypothesis of mitochondrial loss in the oxymonad lineage, which turns out to be supported. The datasets are also explored regarding the role of lateral gene transfer in the group, metabolic diversification and the evolution of Golgi.

Major comments:

I find the manuscript very interesting with many different fascinating results presented. However, the manuscript is very long. Two genome sequences are presented and it is not clear to me what the main question was when this project was initiated and why these two species was selected to answer this question. I do not see an obvious reason for sequencing the P. pyriformis genome if the mitochondrial loss was the main question (given that a transcriptome was already available). Why not spend the time and resources on a member of Preoxystyla, which lacked previous data? The authors should more clearly state why these organisms were chosen to answer the main question or questions of the study.

The evolution of MROs have received substantial attention from the protist research community since the 1990's. During this period the mitochondrial organelle have been considered essential for eukaryotes. Therefore, the result presented in the manuscript has a high significance. However, I am not convinced that it is appropriate to use the term "evolutionary transition" for the mitochondrial loss. The loss of MRO is the endpoint of a gradual change of the internal organisation of the cell that probably started when the ancestor of these organism adapted to an anaerobic lifestyle. The last step described in the manuscript probably had little impact on how these organisms interacted with their environment. The presence or absence of biosynthesis of p-cresol by some, but not all, Preaxystyla probably is much more significant from an ecological point of view. My point is that the authors need to consider how they use the term evolutionary transition and be explicit about that.

In the abstract the main finding is describes as "the data confirm the complete loss of mitochondria and every protein that has ever participated in the mitochondrion function for all three oxymonad species (M. exilis, B. nauphoetae, and Streblomastix strix) extending the amitochondriate status to the whole Oxymonadida.". I find this a really interesting observation, but I do find the wording a bit too bold for several reasons: - Not every protein that has participated in the mitochondrial function is known. - Mitochondrial proteins could be present in oxymonads, but divergent beyond the detection limit for existing methods. - Genes for one or several mitochondrial proteins could be present in one or more oxymonad genomes, but remain undetected due to the incomplete nature of the datasets.

Although I do think that the authors' claim very well could be true, I don't think their data fully support it. Therefore, it needs to be rephrased.

The third point maybe could be analysed further. BUSCO scores are reported, but also argued not being reliable for this group of organisms (which is true). Would it, for example, be useful to analyse how large fraction of the BUSCO proteins found in all non-Preoxystyla metamonada genomes that are present in the various Preoxystyla datasets?

Line 160-161: 15 LGT events specific for the Preaxostyla+Fornicata clade is reported. This is an exciting finding because it supports a phylogenetic relationship between these two groups. But such an argument is only valid if the observed pattern is more common than the alternative hypotheses (Preaxostyla+Parabasalids and Fornicata+Parabasalids). How many LGT events support each of these groupings? How are these observation affected by the current taxon sampling with the highest number of datasets from Fornicata? How were putative metamonada-to-metamonada LGTs treated in this context?

The authors have used a large-scale approach to make single-gene trees for inferences of LGT. In other parts of the manuscript inferences of evolutionary origins of single genes are made without support of phylogenetic trees. I find this inconsistent and argue that the hypothesis of the origin of a specific protein should be tested with the same rigor whether it is a putative LGT, gene duplication, gene loss or an ancestral member of LECA. Specific cases where I think a phylogenetic analysis is needed includes: - Line 222-223: It is concluded that Rsg1 is a component of LECA. - Line 307: HgcAB are argued to be acquired by LGT of a whole opeon. - Lines 350-355: It is unclear how the different numbers of transporters are interpreted (loss or expansion by duplication). This could be address with phylogenetics. - Lines 407-408: A tree should support the claim of LGT origin. - Lines 414-415: The different origins of glycolytic enzymes should be supported by data or references. - Line 486: Trees or a reference (if available) should support the claim for LGT.

Lines 530-531 and 773-774: "The switch to the SUF pathway in these species has apparently not affected the number of Fe-S-containing proteins but led to a decrease in the usage of 2Fe-2S clusters." I find it difficult to evaluate if the data support this because no exact numbers or identities are given for 2Fe-2S and 4Fe-4S proteins in the various genomes in Suppl. Fig. S4 or Supplementary file 4.

The method used in the paper varies between the different parts of the paper. One example is single gene phylogenies, which are described three times in the method section [Lines 959-973, lines 1011-1034, lines 1093-1101], in addition to the automated approach within the LGT detection pipeline lines 923-926]. The approaches are slightly different with, for example, different procedures for trimming. This makes it difficult to know how the different presented analyses were done in detail. No rationale for using different approaches is given. At the least, it should be clear in the method section which approach was used for which analysis. Specific comments on single gene phylogenies: - Line 966-967: Why max 10 target sequences? - Lines 996-998: Is it a problem that these are rather old datasets?

Minor comments:

I appreciate that many data is included as supplementary material. However, the organisation of the data could be improved. The numbering of the files is not included in their names or within the files, as far as I could find. Descriptions of the files are often missing and information on the annotation such as colour coding is not always included. These aspects of the supplementary material needs to be strengthened in order to make it more useful. Specific comments: - Supplementary file 1, Table 1: accession numbers are missing. Kipferlia bialta appears to have a much smaller number of sequences than reported in the publication. The file consists of three tables and it would be very helpful if the reference in the main manuscript indicate the table number. - Supplementary file 4: The trees lack proper species names and a documented colour coding. There are multiple trees in the file, which make it difficult to find the correct tree. I would appreciate if the different trees were labelled A, B, C, etc., and if these were used in the main text. - There is no HPAD-AE tree (as indicated on line 258), but a HPAD tree. Which part of the tree contain the described fusion protein? - Line 280-281: "UbiE homologs occur also in some additional metamonads, including the oxymonad B. nauphoetae and certain fornicates." These sequences should be clearly highlighted in the tree. - Lines 538-544: A three-gene system is mentioned, but only two AmmoMemoRadiSam trees are found. - Supplementary file 6: I find it difficult to find the proteins discussed in the text, for example "the biosynthesis of p-cresol from tyrosine (line 254-255)". - Supplementary file 11: Which group of species are highlighted in red? How do I know from which species these sequences are (I can make educated guesses, but prefer full species names). I do not find any reference to this file in the main manuscript.

Line 227-228: "630 OGs seem to be oxymonad-specific or divergent, without close BLAST hits". It is unclear if BLAST searches includes only a representative of each 630 OGs, or every single protein in these OGs.

Line 243: I think it is five LGT mapped to internal nodes of Preoxystyla in Figure 1 (1+3+1).

Lines 325-331: The argument would be stronger with a figure showing the fusion and the alignment indicating the conserved amino acids mentioned in the text.

Lines 425: "none of the species encoded" should be replaced by something like "none of the enzyme could be detected in any of the species" (the datasets are incomplete).

Line 455: "suggesting a cytosolic localization of these enzymes in Preaxostyla." The absence of a phylogenetic affiliation with the S. salmonicida homolog does not preclude a MRO localisation.

Lines 570-571: "Manual verification indicated that all the candidates recovered in oxymonad data sets are false positives" Using which criteria?

Lines 743-755: "Similar observations were made in other protists with highly reduced mitochondria, such as G. intestinalis or E. histolytica,..." References are needed.

Line 849: How was the manually curation done for the gene models in the training set?

Lines 853-856: It is a bit unclear which dataset was used for BUSCO and downstream analysis. Was it the Augustus-predicted proteins, or the EVM polished?

Lines 858: What is it meant that KEGG and similarity searches was used in parallel (what if both gave a functional annotation?)?

Lines 861-862 and 1007-1008: Which genes or sub-projects does this apply to? How many genes were detected in this procedure?

Lines 878-879: It is not clear to me why the sum of the two described numbers should be as high as possible and would appreciate an argument or a reference.

Lines 879-881: "Proteins belonging to the thus defined OGs were automatically annotated using BLASTp searches against the NCBI nr protein database (Supplementary file 1)." Why were these annotated in a different way (compare lines 857-859).

Lines 894-957: "Detection of lateral gene transfer candidates": - It is not clear which sequences were tested in the procedure. All Preaxostyla, or all metamonada? I think I am confused because in the result sections you only report numbers for Preaxostyla, but in the method section metamonada is mentioned repeatedly. - It would be easier to follow the procedure if numbers are provided for the different steps. - Why was only small oxymonad proteins discarded (line 900)? - Line 911: How many sequences were collected? - Lines 916-919: What is the difference between the protein superfamilies (line 916) and the OGs (line 919)? Are the OGs the same orthogroups that is described earlier in the method section? How are the redundancy of NCBI nr entries retrieved in different searches dealt with?

Line 987-989: "...was facilitated by Rsg1 being rather divergent from other Ras superfamily members" This statement is vague. What does it mean in practise?

Lines 1037-1038: Why were these proteins re-annotated?

Figures: The figures would be easier to follow if the colour coding for the five different species were consistent between the figures.

Figure 1: It appears that the Venn diagram in C only shows the Preaxostyla-specific protein in B, not all OGs for which contain Preaxostyla proteins. This is not clear from legend or from the figure itself. The same comment applies to D.

Figures 2 and 6: It would be clearer with panel labels A, B, etc, instead of "upper" and "lower" panel, as in the other figures.

Figure 6: What is the colour code in the figure? The numbers within the boxes are not aligned.

Supplementary figures 1-3: What do green and magenta indicate in the figure?

** Referees cross-commenting**

I agree with the other reviewers that the discussion of the functional and ecological implications of the LGTs could be developed.

In contrast to reviewer 2, I do not see that the authors discuss their result in the context of eukaryogenesis in this manuscript. Maybe the reference reviewer 2 mention could be cited in the introduction together with Hampl et al. 2018 to acknowledge that there are different views about the importance of secondarily amitochondrial eukaryotes on our thinking about the origin of eukaryotes. I disagree with reviewer 2's objection against the wording "... and undergo pronounced morphological evolution" because I think Fig. 4 in Hampl 2017 shows a large morphological diversity among oxymonads.

Significance

The findings presented in this manuscript can be divided into two parts: the mitochondrial loss and the metabolic and Golgi analyses. The latter is a substantial contribution to the knowledge of metabolic adaptation in unicellular eukaryotes where it builds on previous similar works in other organismal groups. These findings should be of general interest for the protist field.

The loss of mitochondria in M. exilis has been reported by the authors in several previous publications (Karnkowska, et al. (2016, 2019), Treitli, et al. (2021)). Here they show that a distantly related oxymonad (B. nauphoetae) also lack all signs of the mitochondria, suggesting that all oxymonads might have lost the mitochondrion completely. This shows that M. elixis is not a weird lineage, which recently lost the organelle and therefore is on a fast evolutionary dead end. Rather, a whole group of microbial eukaryotes have lived for long evolutionary times without any organelle with mitochondrial ancestry.

This shows that eukaryotes can be successful without any kind of mitochondrial organelle. Such a conclusion should be of interest to a wide audience.

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

Using draft genome sequencing of the free-living Paratrimastix pyriformis and the sister lineage oxymonad Blattamonas nauphoetae, Novack et al. infer the metabolic potential of the two protists using comparative genomics. The authors conclude that the common oxymonad ancestor lost the mitochondrion/mitosome and discuss general strategies for adapting to commensal/symbiotic life-style employed by this taxon. Some elaborations on pathways go on for several paragraphs and feel unnecessarily stretched, which made those sections of the paper rather difficult to digest. This might be also be because the work, and all conclusions drawn, depend entirely on incomplete (ca. 70-80%) genome data and simple similarity searches, and e.g. no kind of biochemistry or imaging is presented to underpin the manuscripts discussion. This is noteworthy in light of other protist genome reports published in the last few years that differ in this respect, including previous work by this group. And for sequencing-only data, this paper - https://doi.org/10.1016/j.dib.2023.108990 - might offer an example of where we are at in 2023. With respect to previous work of the group (Karnkowska et al. 2016 and 2019), this submission is very similar (analysis pattern, even some figures and more or less the conclusion), i.e. to say, the overall progress for the broader audience is rather incremental. Then there are also some incidents, where the data presented conflicts with the authors own interpretation. The text (including spelling and grammar) needs some attention and the choice of words is sometimes awkward. The overuse of quotation marks ("classical", "simple", "fused", "hits", "candidate") is confusing (e.g. was the BLAST result a hit or a "hit").

In its current form the manuscript is, unfortunately, very difficult to review. This reviewer had to make considerable efforts to go through this very large manuscript, mainly because of issues affecting to the presentation and the lack of clarity and conciseness of the text. It would be greatly appreciated if the authors would make more efforts upfront, before submission, to make their work more easily accessible both to readers and facilitate the task of the reviewers.

About a fifth of the two genome is missing according the authors prediction (table 1). Early on they explain the (estimated) incompleteness of the genomes to be a result from core genes being highly divergent. In light of this already suspected high divergence, using (the simplest NCBI) sequence similarity approach to call out the absence of proteins (for any given lineage) may need lineage-specific optimization. The use of more structural motif-guided approaches such as hidden Markov models could help, but it is not clear whether it was used throughout or only for the search for (missing) mitochondrial import and maturation machinery. The authors state that the low completeness numbers are common among protists, which, if true, raises several questions: how useful are then such tools/estimates to begin with and does this then not render some core conclusions problematic? The reader is just left with this speculation in the absence of any plausible explanation except for some references on other species for which, again, no context is provided. Do they have similar issues such as GC-content, same core genes missing, phylogenetic relevance?, etc.. No info is provided, the reader is expected to simply accept this as a fact and then also accept the fact that despite this flaw, all conclusions of the paper that rests on the presence/absence of genes are fine. This is all odd and further skews the interpretations and the comparative nature of the paper.

As a side note, this will also influence the number of proteins absent in other lineages and as such has consequences on LGT calls versus de novo invention. For the cases with LGT as an explanation, it would help to briefly discuss the candidate donors and some details of the proteins in the eco-physiological context (e.g. lines 263-268 suggest that HPAD may have been acquired by EGT which was facilitated by a shared anaerobic habitat and also comment on adaptive values for acquiring this gene). Exchanging metabolic genes via LGT (Line 163) blurs the differences between roles and extent of LGT in prokaryote vs eukaryote, and therefore is exciting and could use support/arguments other than phylogenies. I guess the number of reported LGTs among protists (whatever the source) over the last decade has by now deflated the novelty of the issue in more general; a report of the numbers is expected but they alone won't get you far anymore in the absence of a good story (such as e.g. work on plant cell wall degrading enzymes in beetles). It would help to clarify which parts of the mitochondrial ancestor were reduced during the process of reductive evolution at what time in their hypothesized trajectory. For instance, loosing enzymes of anaerobic metabolism conflicts with the argued case of an aerobic (as opposed to facultative anaerobic) mitochondrial ancestor followed by gains of anaerobic metabolism in the rest of the eukaryotes via LGT, and some papers the authors themselves cite (e.g. the series by Stairs et al.). There is no coherent picture on LGT and anaerobic metabolism, although a reader is right to expect one.

In light of their data the authors also discuss the importance of the mitochondrion with respect to the origin of eukaryotes:

First, the mitochondrion brought thousands of genes into the marriage with an archaeon, surely hundreds of which provided the material to invent novel gene families through fusions and exon shuffling and some of which likely went back and forth over the >billion years of evolution with respect to localizations. The authors look at a minor subset of proteins (pretty much only those of protein import, Fig. 6) to conclude, in the abstract no less: „most strikingly the data confirm the complete loss of mitochondria and every protein that has ever participated in the mitochondrion function for all three oxymonad species." I do not question the lack of a mitochondrion here, but this abstract sentence is theatrical in nature, nothing that data on an extant species could ever proof in the absence of a time machine, and is evolutionary pretty much impossible. A puzzling sentence to read in an abstract and endosymbiont-associated evolution.

Second, using oxymonads as an example that a lineage can present eukaryotic complexity in the absence of mitochondria and conflating it with eukaryogenesis is a logical fallacy. This issue already affected the 2019 study by Hampl et al.. We have known that a eukaryote can survive without an ATP-synthesizing electron transport chain ever since Giardia and other similar examples and the loss of Fe-S biosynthesis and the last bit of mitosome (secondary loss) doesn't make a difference how to think about eukaryogenesis. It confuses the need and cost to invent XYZ with the need and cost of maintenance. How can the authors write "... and undergo pronounced morphological evolution", when they evidently observe the opposite and show so in their Fig. 1? The authors only present evidence for reductive evolution of cellular complexity with the loss of a stacked Golgi. What morphological complexity did oxymonads evolve that is absent in other protists? A cytosolic metabolic pathway doesn't count in this respect, because it is neither morphological, nor was it invented but likely gained through LGT according to the authors. This is quite confusing to say the least. A recent paper (https://doi.org/10.7554/eLife.81033) that refers to Hampl et al. 2019 has picked this up already, and I quote: "Such parasites or commensals have engaged an evolutionary path characterized by energetic dependency. Their complexity might diminish over evolutionary timescale, should they not go extinct with their hosts first." Here the authors raise a red flag with respect to using only parasites and commensals that rely on other eukaryotes with canonical mitochondria as examples. If we now look at Fig. 1 of this submission, Novak et al. underpin this point perfectly, as the origin of oxymonads is apparently connected to the strict dependency on another eukaryote (or am I wrong?), and they support the prediction with respect to complexity reducing after the loss of mitochondria - mitosome gone, Golgi almost gone. What's next? This is a good time to remember that extant oxymonads are only a single picture frame in the movie that is evolution, and their evolution might be a dead-end or result in a prokaryote-like state should they survive 100.000s to millions of years to come.

Some more thoughts:

Line 47-52: Hydrogenosome or mitosome is a biological and established label as (m)any other and I find the use of the word "artificial" in this context strange. While the authors are correct to note that there is a (evolutionary) continuum in the reduction - obviously it is step by step - they exaggerate by referring to the existing labels as "artificial". You make Fe-S clusters but produce no ATP? Well, then you're a mitosome. It's a nomenclature that was defined decades ago and has proven correct and works. If the authors think they have a better scheme and definition, then please present one. Using the authors logic, terms such as amyloplast or the TxSS nomenclature for bacterial secretions systems are just as artificial. As is, this comes across as grumble for no good reason.

Line 158: A duplication-divergence may also explain this since sequence similarity-based searches will miss the ancestral homologues.

Lines 201-202: Presence of GCS-L in amitochondriate should be explained in light of this group once having a mitochondrion, which then makes ancestral derivation and differential loss (as invoked for Rsg1) also a likely explanation along with eukaryote-to-eukaryote LGT.

Lines 356-392: Describes plenty of genomic signal for Golgi bodies but simultaneously cites literature suggesting the absence of a morphologically an identifiable Golgi in oxymonads. An explicit prediction regarding what to observe in TEM for the mentioned species might be nice to stimulate further work.

Lines 414: The preceding paragraphs in this result section describes only the distribution, without mentioning origins - a sweeping one-line summary that proclaims different origin needs some context and support. Furthermore, the distribution of glycolytic enzymes might indeed be patchy, but to suggest it represents an 'evolutionary mosaic composed of enzymes of different origins' without discussing the alternative of a singular origin and different evolutionary paths (including a stringer divergence in one vs. another species) discredits existing literature and the authors own claim with respect to why BUSCO might fail in protists.

Line 486: How uncommon are ADI and OTC in lineages sister to metamonada?

Line 504: It might help an outside reader to include a few lines on consequences and importance of having 2Fe-S vs 4Fe-S clusters and set an expectation (if any) in Oxymonads

Any explanations on what unique selection pressures and gene acquisition mechanisms may be operating in P. pyriformis which might allow for the unique metabolic potential?

** Referees cross-commenting**

To R3: Hampl et al. 2019, to which Novak et al. refer, is about eukaryogensis and that is exactly the context in which this is discussed again and what Raval et al. 2022 had decided to touch upon. If the authors do not bring this up in light of the ability to evolve (novel) eukaryote complexity, then what else? Maybe they can elaborate, especially with respect to energetics to which they explicitly refer to in 2019 (and here). And with respect to text-book eukaryotic traits (and the evolution of new morphological ones), I do not see any new ones evolving in any oxymonad, but reduction as Novak et al. themselves picture it in this submission. Is a change in the number of flagella pronounced morphological evolution? Maybe for some, but I believe this needs to be seen in light of the context of how they discuss it. I see a reduction of eukaryotic complexity and not a gain. They have an elaborate section on the loss of Golgi characteristics (and a figure), but I fail to read something along the same lines with respect to the gain of new morphological traits. Again, novel LGT-based biochemistry does not equal the invention of a new morphology such as a new compartment. Oxymonads depend on mitochondria-bearing eukaryotes for their survival or don't they? This is the main point, and if evidence show that I am wrong, then I will be the first to adapt my view to the data presented.

I have concerns with the presentation of a narrative that in my opinion is too one-sided and that has been has been publicly questioned in the community (in press, at meetings, personally). For the benefit of science and of the young authors on this study, this reviewer feels strongly that these issues should be taken very seriously and discussed openly in a more balanced way. . We only truly move forward on such complex topics, if we allow an open and transparent discussion.

Having said that, I am happy that R3 has picked up exactly the same major concerns as I did with respect to e.g. the phrasing on mito (gene) loss and the BUSCO controversy.

Significance

Using draft genome sequencing of the free-living Paratrimastix pyriformis and the sister lineage oxymonad Blattamonas nauphoetae, Novack et al. infer the metabolic potential of the two protists using comparative genomics. The authors conclude that the common oxymonad ancestor lost the mitochondrion/mitosome and discuss general strategies for adapting to commensal/symbiotic life-style employed by this taxon. Some elaborations on pathways go on for several paragraphs and feel unnecessarily stretched, which made those sections of the paper rather difficult to digest. This might be also be because the work, and all conclusions drawn, depend entirely on incomplete (ca. 70-80%) genome data and simple similarity searches, and e.g. no kind of biochemistry or imaging is presented to underpin the manuscripts discussion.

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

This is a very interesting paper that investigates through detailed comparative genomics the tempo and mode of the evolution of microbial eukaryotes/protists members of the Metamonada with a focus on Preaxostyla, currently the only known lineage among eukaryotes to have species that have lost, by all accounts, the mitochondria organelle all together. Notably, it includes a free-living representative of the lineage allowing potential interesting comparison between lifestyles among the Preaxostyla. This is a generally nicely crafted manuscript that presents well supported conclusions based on good quality genome sequence assemblies and careful annotations. The manuscript presents in particular (i) additional evidence for the common role of LGT from various bacterial sources into eukaryotic lineages and (ii) more details on the transition from a free-living lifestyle to an endobiotic one and (iii) the related evolution of MROs and associated metabolism. I have some comments to improve a few details: In the introduction, lines 42-43, the last sentence should be more conservative by replacing "whole Oxymonadida" with "...all known/investigated Oxymonadida". Similarly on line 62, the sentence could state "... contain 140 described...". When discussing the estimated completeness of the genome are discussed (lines 117-120) and contrasted with the values for Trypanosoma brucei and other genomes, the author should explicitly state that these genomes are considered complete, which seems is what they imply, is that the case? If so, please provide more details to support this idea. The supplementary file named "132671_0_supp_2540708_rmsn23" is listed as a Table SX? (note: I found it rather difficult to establish exactly what file corresponds to what document referred in the main text) Lines 243-245, where 46 LGTs are discussed, it is relevant that the authors investigate their functional annotations. Indeed, it is suggested that these could have adaptive values, hence investigating their functional annotation will allow the authors to comment on this possibility in more details and precision. When discussing LGTs it would also be very useful to cite relevant reviews on the topic - covering their origins, functional relevance when known, distribution among eukaryotes. This is done when discussing the evolution and characteristics of MROs but not when discussing LGTs, with several reviews cited and integrated in the discussion of the data and their interpretation. The sentence, lines 263-265, where the distribution of some LGTs are discussed, needs to be made more precise. When using the work "close" the authors presumably refer to shared/similar habitat,s or else? Entamoeba is not a close relative to the other listed taxa. Lines 346-348, that sentence needs to end with a citation (e.g. Carlton et al. 2007). In the paragraph (line 580-585) discussing ATP transporters, note that Major et al. (2017) did not describes NTTs but distantly related members of MSF transporter, shared across a broader range of organisms then the NTTs. Did the authors checked if the genome of interest encoded homologues of these transporters too? Line 920-921, I don't understand how the number 30 relates to "guarantee" inferring the directionality of LGTs events. This will be very much dataset dependent, 100 sequences might still not allow to infer directionality of LGT events. The authors probably meant to "increase the possibility to infer directionality".

Significance

This is a very interesting paper that investigates through detailed comparative genomics the tempo and mode of the evolution of microbial eukaryotes/protists members of the Metamonada with a focus on Preaxostyla, currently the only known lineage among eukaryotes to have species that have lost, by all accounts, the mitochondria organelle all together. Notably, it includes a free-living representative of the lineage allowing potential interesting comparison between lifestyles among the Preaxostyla. This is a generally nicely crafted manuscript that presents well supported conclusions based on good quality genome sequence assemblies and careful annotations.

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

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

This study evaluates the effect of fungal toxin candidalysin on neutrophils. The authors show that candidalysin induces NETosis when secreted by hyphae, but when candidalysin is added on its own, NLS are formed instead which are distinct from NETs. The authors have done lots of carefully controlled experiments, and delineated key components of the pathway inducing NLS, including the role of ROS and histone modifications. The data provided is high quality and well presented in figures.

Reviewer #1 (Significance (Required)):

Strengths are the depth of analysis - many different aspects of NETosis is assessed and robustly tested.

Comments: 1. I was a bit confused by what should be the main message of the paper - is it that candidalysin on its own doesn't induce NETosis but only NLS? The answer to this question wasn't well addressed in my opinion, but the paper switches between using live fungi and purified candidalysin so it became confusing at times.

*

Responses:

Thank you for this important comment. We have clarified our narrative on candidalysin throughout the manuscript to provide a red thread for the readership. Our message is that candidalysin alone has not the capacity to induce a full cycle of signalling events which result in canonical NET formation. Our data show that candidalysin alone falls short and can only produce NLS. On the other hand, our data show that in the context of growing C. albicans cells candidalysin is able to promote the release of NETs. This is important, since previously the hyphal form of C. albicans has been reported to be a formidable inducer of NETs, whereas the yeast form was not. Our data put candidalysin in the centre of this observation, showing that it indeed is a major contributor to NET formation when present with growing C. albicans cells. Since candidalysin expression and release is strictly connected to hyphal growth our new data agrees with previous assessment and provides new insight in how this hyphae-specific inductive effect is accomplished.

In the revised manuscript, at first, we describe the difference of neutrophil stimulation when using strains expressing and lacking candidalysin as compared to candidalysin stimulation alone. We added or modified the following phrases:

• Line (163) “As candidalysin-expressing C. albicans strains induced more NETs than candidalysin-deficient strains, we investigated the role of the toxin alone in stimulating neutrophil extracellular trap release”.
• Line (173) “In order to ensure consistency in NET/NLS quantification, NLS were quantified with the same criteria as previous described for NETs.”

• Line (182) “In summary, candidalysin alone triggers morphologically distinct NLS in a time- and dose-dependent manner, whereas candidalysin-producing C. albicans hyphae induce canonical NETs (Fig. 1a).” Next, we describe the different morphology of NLS triggered by candidalysin alone in comparison to canonical NETs triggered by C. albicans strains expressing candidalysin. We added or modified the following phrases:

• Line (198) “To investigate candidalysin-triggered NLS in further detail, we used scanning electron microscopy (SEM) that allows a more detailed view of the neutrophil-derived structures (Fig 3a).” Furthermore, to prevent switching between experiments using candidalysin alone and experiments with different Candida strains, we have moved the next paragraph “Candidalysin-expressing strains induce more NETs and higher citrullination levels than candidalysin-deficient strains” to the end of the result section (old Fig. 4 is now new Fig. 8). In doing so, we focus on the direct morphological, signalling and functional effects of candidalysin alone on neutrophils and towards the end, we analyse how the strains are affected by different neutrophil killing mechanisms (phagocytosis and NETs). Subsequently, we synthetize our findings by showing that candidalysin is the main driver of histone citrullination by quantifying this histone modification in the context with NET induction comparing candidalysin-expressing and -deficient strains. We conclude that citrullination-induced chromatin decondensation in combination with candidalysin-induced ROS production are most probably the main contributors of increased NET formation stimulated by C. albicans hyphae expressing candidalysin. This is the also a good conclusion of the manuscript showing that candidalysin alone is not enough but together with growing C. albicans cells it contributes to NET induction and increased NETs in turn inhibit growth and limit spreading of C. albicans.

• We modified and added the following sentences to the discussion section: (line 457) “The data suggests that candidalysin is the key driver of histone citrullination in neutrophils infected with C. albicans and that addition of evenly distributed, external candidalysin in high concentration (15 µM) drives neutrophils towards NLS despite the presence of C. albicans cells. We conclude that, during infection, candidalysin-triggered Ca2+ influx and histone hypercitrullination amplify processes in neutrophils which are induced by C. albicans hyphae. These amplified processes culminate in a strongly increased release of NETs that in turn are formidable weapons to control hyphal filaments.”

  *2. If candidalysin on its own only induces NLS - what is the relevance of this for disease? A lot of work has been provided on the pathway driving NLS formation, but it wasn't clear to me why this is important. More in discussion needed or evidence of disease relevance. *

Responses:

Thank you for giving us the opportunity to clarify this issue. Candidalysin expression strongly increases with and is restricted to hyphal growth, which is the adhesive growth form of C. albicans. Given that epithelial cells expunge candidalysin for their own protection while hyphae remain attached, it could be possible that neutrophils get exposed to candidalysin before they encounter C. albicans cells. Therefore, it is relevant to understand how candidalysin per se shapes neutrophil responses. We have added the following sentences to the discussion section: (line 527) ”As epithelial cells are able to expunge candidalysin for protection while C. albicans hyphae remain adeherent 46 recruited neutrophils may encounter candidalysin before direct contact with hyphae.”

With regard to relevance for candidiasis the observation that candidalysin-deficient strains are poor inducers of NETs is most important. Since candidalysin expression is entirely restricted to hyphal growth. this finding gives crucial, new insight into the previous observation that hyphae are better NET inducers than yeast from C. albicans. In this context, we wanted to make it very clear to the reader that this effect only works when C. albicans cells and candidalysin are combined and that candidalysin alone does not lead to full-blown NET formation. Therefore, we have included a thorough investigation of the effects of candidalysin on neutrophils to be able to better contextualize our findings comparing candidalysin-expressing and candidalysin-deficient strains.

To make this point clearer. we added the following sentence to the summary at the end of the discussion section: (line 565) “Neutrophils encountering candidalysin-expressing hyphae are able to adequately respond by releasing increased amounts of NETs whereas secretion of candidalysin does not allow hyphae to evade from neutrophil attack.”

In addition, we are convinced that the move of former Fig.4 to the end of the result section (now Fig. 8) additionally helps the reader to better understand the importance to first delineate the effect of candidalysin on neutrophils alone and then to conclude the manuscript with experiments using different C. albicans strains to put the findings into context.

To add more substance to our conclusions we wrap up the new version of the manuscript with data comparing wild-type and candidalysin-deficient strains in neutrophil antimicrobial assays and quantification of histone citrullination. With the newly added antimicrobial assays we demonstrate that candidalysin expression does not affect phagocytic killing (Fig. 7d and 7e) as assessed by plating assays and that candidalysin does not affect inhibition by PMA-induced NETs (Fig. 7f and 7g). Thus, as stated above, during the interaction of hyphae and neutrophils candidalysin promotes the release of more NETs, but does otherwise not affect anti-Candida activity by neutrophils. Increased NETs in turn, however, inhibit growth and limit spreading of C. albicans. The manuscript now ends with the data on differences in histone citrullination when using wild-type and candidalysin-deficient strains indicating that citrullination-induced chromatin decondensation in combination with C. albicans cells ultimately leads to increased NET release.

We added the following text to the manuscript: (Line 416) “To corroborate, whether candidalysin deficiency affects C. albicans’ susceptibility to neutrophil attack we performed two antimicrobial assays. In the first assay we determined NET-mediated anti-Candida activity by preformed NETs comparing wild-type and candidalysin-deficient strains. We used the same imaged-based analysis with calcofluor white staining. To be able to better observe differences in susceptibility of the different strains we used a slightly higher MOI than for the previous NET inhibition assays which explains higher survival percentage (Fig. 7c, black bars on the right side). As expected, candidalysin did not affect the inhibitory effect on C. albicans imposed by NETs (Fig. 7c). In the second assay, we determined short-term anti-Candida activity of intact neutrophils, which is predominantly phagocytic elimination, by serial dilution and plating for colony counts. Candidalysin-deficient and wild-type strains are killed similarly over the time of 1 to 4 h, both at MOI 1 and 3 (Fig. 7d and 7e). This indicates that candidalysin expression does not enable evasion from neutrophil phagocytic attack and this result agrees well with our previous finding that wild-type C. albicans engulfed by human neutrophils are unable to escape by hyphal outgrowth 16. In conclusion, while candidalysin strongly increases the NET-inductive capacity of C. albicans hyphae, the toxin does neither affect the anti-Candida effect of intact neutrophils nor of NETs.”

Notably, it is not informative to use C. albicans as inducer of NETs and as target of anti-Candida activity by NETs in the same assay, since both induction and anti-Candida activity are dependent on the amount of C. albicans cells. We therefore chose to show two separate assays where we (i) quantify short-term killing by plating (mainly phagocytosis) and (ii) quantify growth inhibition of C. albicans by pre-stimulated NETs.

*3. In Figure 2, it would be helpful to include images of ionomycin-stim neutrophils for comparison of the NLS structures across different stim conditions. *

Response:

This is a very good point. We supply a structural comparison between NLS and NETs induced by PMA, ionomycin and candidalysin in Figure 3. Additionally, the time-dependent changes for ionomycin are now included in the supplementary Figure S1.

4. Few places where reference manager has failed (see bottom on page 10, line 190 for example)

Response:

We have fixed this issue, thank you for pointing it out.

*5. Lines 191-198 - I was confused here by the text. I thought the point was that candidalysin induced NLS similar to ionomycin, but here the point is being made that the two are different? This led me to being confused as to the point of all the comparisons made between ionomycin NLS and candidalysin NLS... this could be made clearer. *

Responses:

Thank you for highlighting this. According to previous literature ionomycin, a bacterial peptide toxin, was the most prominent example for induction of leukotoxic hypercitrullination. Therefore, we used ionomycin to put our findings with candidalysin, a fungal peptide toxin, into context. We find that candidalysin share similarities but also some striking differences to ionomycin. While we could not investigate the nature of these differences in more detail, this could be the basis of a follow-up study, we think it is important to give the reader the comparison in order to better understand how candidalysin shapes neutrophil responses. One clear difference which we show in the manuscript is that candidalysin induces some ROS whereas ionomycin does not at all (Fig. 4).

We changed the text in the result section accordingly to make our point clearer: (line 203) “PMA exposure generated widespread chromatin fibers in the extracellular space (Fig. 3a, left panels) whereas ionomycin exposure resulted in more compact, patchy areas occasionally dispersed with long, thin chromatin fibres (Fig. 3b, middle panels). With regard to morphological changes, candidalysin treatment resulted in compact, fibrous structures resembling those stemming from ionomycin treatment, however long, thread-like structures were absent in candidalysin-treated neutrophil samples (Fig. 3a right panels, for 7 h treatment see Fig. S1c).”

And did so as well in the discussion section: (Line 513) ”While ionomycin- and candidalysin-induced NLS shared similar key features, such as increased histone citrullination, our study revealed striking differences between the two toxins. In contrast to ionomycin, candidalysin stimulation led to ROS production in neutrophils.”

*6. Could the authors include some unstim neutrophil control images in Fig 3 for the SEM? Can the SEM sample processing affect neutrophil structure in anyway? Feels like an important control although I don't have much experience with SEM personally *

Response:

This is of course a relevant control image. We have included an image showing unstimulated neutrophils from similar time points, but without exposure to candidalysin (Fig. 3). The unstimulated neutrophils are spherical and morphologically distinctly different from candidalysin-treated neutrophils.

*7. I was very intrigued by the experiments where the authors added candidalysin in to neutrophils infected with ece1-null strain. Those experiments showed that candidalysin addition still drove NLS instead of NETosis. Can the authors investigate why this is? Is membrane intercalation different when candidalysin is delivered by hyphae vs added on its own? Could that explain some of the differences they have seen? *

Responses:

Thank you for this comment. Yes, there is a clear difference, since we add candidalysin to the medium such that the peptide is evenly distributed and reaches membranes rather evenly from the extracellular space. When released from growing C. albicans hyphae candidalysin is then predominantly released on hyphal tips as demonstrated in the referenced article (doi.org/10.1111/cmi.13378). Hyphal tips in turn are readily attacked by human neutrophils (doi.org/10.1189/jlb.0213063). Hence, we can safely assume according to these previous publications that there will be a more uneven distribution of candidalysin concentrations over neutrophil membranes, when the sole source of the toxin stems from growing hyphae interacting with neutrophils. It would of course be very interesting to know how the toxin exactly intercalates into membranes and which morphologies potential pores may have. These questions are currently under investigation in the laboratories of Profs Hube and Naglik. To include these findings here would certainly be far beyond the scope of this study.

We include and modify the following sentences to the discussion of this manuscript to clarify the issue: (Line 541). ”One of the main goals of the study was to delineate contribution of candidalysin to neutrophil responses either as factor released by C. albicans hyphae or as singular peptide toxin. Our data demonstrates that candidalysin is the main driver of histone citrullination in neutrophils infected with C. albicans (Fig. 8). Lack of candidalysin production in C. albicans results in significantly reduced histone citrullination, accompanied with decreased NET formation. However, citrullination is not required for NET release, but rather governs the formation of NLS, which is dominant when candidalysin is added exogenously with even distribution throughout the cell suspension. With regard to C. albicans hyphae secreting candidalysin, local concentrations of the toxin are likely to vary to a large degree, particularly when the candidalysin-secreting hypha is engulfed by a neutrophil. Therefore, it may be difficult to discriminate NLS form NETs during the interaction of neutrophils and C. albicans, as both structures may be induced concurrently 10. It seems logical that the pore-forming activity of candidalysin augments the release of NET fibres during C. albicans infection, where PRRs will additionally be triggered on neutrophils, resulting in combinatorial activation of downstream pathways. In line with this notion, candidalysin drives histone citrullination, which contributes to chromatin decondensation.”

*8. Is phagocytosis needed for NETosis induction by candidalysin? What happens if you add beads or beta-glucan particles with candidalysin stimulation? Do you get NLS or NETs? *

Responses:

This is an interesting question. Physical contact is required for the induction of NET formation (10.1111/j.1462-5822.2005.00659.x, 10.1371/journal.ppat.1000639) and physical contact leads to pattern recognition unequivocally followed by phagocytic events in neutrophils. Hence, at the least indirectly, phagocytosis and NET formation are connected, but may not be so causally.

While glucan-covered particles have been shown to induce NETs (10.1159/000365249), we show that C. albicans cells devoid of candidalysin induce NETs, but to a much lesser extent than wild-type C. albicans. In addition, the experiment shown in Fig. 8 shows exactly that. Instead of glucan-covered beats we used C. albicans cells (Fig. 8f) which by virtue are glucan covered.

*9. Please confirm what the n numbers refer to in the figure legends - are these biological or technical replicates? How many experiments are the representative images representing? *

Response:

Thank you very much for pointing this out. We adapted our figure legends accordingly and added the number of biological and technical replicates (n=x(y), x=biological replicates, y=technical replicates). Each experiment has been performed with at least three biological replicates which includes the use of different neutrophil donors.

---

*Reviewer #2 (Significance (Required)):

*

*The advantage of this work is the presentation of the mechanism associated with NLS formation in contact with candidalysin, where activation of NADPH oxidase and calcium influx have been documented to be important. This toxin can trigger ROS production and activate downstream signaling that is important for morphological changes and NLS formation. The important finding is also that NLS are resistant to nuclease treatment and increase the ability of neutrophils to control C. albicans hyphae formation and fungal cell growth. These findings provide a better understanding of the role of neutrophils in the treatment of infections caused by these microorganisms. Below I present are minor suggestions that, in my opinion, will improve the text and correct the presentation of the results, making this set of results a valuable source for explaining such a complex problem.

*

Response:

Thank you for this assessment. In cases which we have identified as crucial for our message we have decided to include additional experiments to better convey our message (Fig. 6e-f and Fig. 7d-g). We also included a time course for ionomycin stimulation of neutrophils in Fig. S1. We appreciate that the overall assessment was that no additional experiments were required.

1/ The authors should decide what thesis about NLS they want to prove: 100 NLS are less fibrous and ....... than canonical NETs and are triggered in an NADPH oxidase-independent fashion.

* 121 NLS were dependent on NADPH oxidase-mediated reactive oxygen species (ROS) production

*

Response:

This was indeed imprecisely formulated from our side. NLS were previously described as NADPH-independent processes stimulated by toxins (see ionomycin). Candidalysin seems to trigger NADPH-dependent and NADPH-independent pathways. However, the main differentiation criteria were described through the hypercitrullination which we could observe for candidalysin. To clarify, we have modified the following sentence: (line 121) ”In contrast to previously described stimuli of NLS, candidalysin induced NLS in partial dependence on NADPH oxidase-mediated reactive oxygen species (ROS) production, wheras PAD4-mediated histone citrullination could be observed as well. Notably, candidalysin alone failed to induce NETs as indicated by a lack of cell cycle activation determined via lamin A/C phosphorylation assays.”

*2/ for the experiment described in the line below, MOI 2 was chosen; did the authors conduct an analysis of the response/eventual change in it, depending on the MOI?

*

Response:

Yes, from our experience in in vitro experiments with human neutrophils MOI3 C. albicans overgrows too quickly. This is why an MOI 1-3 is the best option to analyse NET induction capacities.

131 we infected neutrophils with wild-type C. albicans, ECE1-deficient (ece1ΔΔ), and corresponding revertant (ece1ΔΔ*+ECE1) strains,

3/ Has the effect of deletion of ECE1 on other aspects of virulence, such as adhesion, virulence factor production, or biofilm formation, been analyzed? *

Response:

Yes indeed, the effect of candidalysin on other aspects has been studied. Candidalysin has no effect on adhesion and is expressed during biofilm formation. It has a broad effect on virulence in general and promotes neutrophil recruitment indirectly by a robust induction of damages responses. To clarify the amount of studies investigating these other aspects and to pinpoint the knowledge gap for direct interaction of neutrophils and candidalysin we include the following sentence: (line 132) “C. albicans hyphae release candidalysin and while the effects of the toxin for instance on virulence in general and on adhesion to host cells have been widely studied 17,18,23,28,30, the direct impact of candidalysin on the neutrophil immune response towards C. albicans, remains poorly understood. To investigate the role of candidalysin, we infected neutrophils with wild-type C. albicans,…”

*137 the ECE1- and candidalysin-deficient strains triggered reduced levels

4/ Fig.1 - How were C. albicans cells stained? Does 100%NET mean the number of cells netting after PMA treatment? This information should be given.

*

Response:

Thank you for pointing this out. We were a bit unclear here. We added details in the respective figure legend and method section. C. albicans cells were visualised with anti-Candida antibody (1 µg/mL, ProSci, Cat#35-645). Furthermore, C. albicans nuclei are stained by DAPI, too. 100% NETs would mean that every single neutrophil (an image event which stains for neutrophil markers) in the analysed microscopic picture shows NET or NET-like morphology. We did not normalize to PMA treated cells.

5/ 168 dependent effect with increased NLS formation from 3 μM to 15 μ*M. However, the reduced NLS

How was determined the limiting concentration value of the toxin, for which an increase in NLS was observed? Was a wide range of concentrations used in the analysis or was the determination made only for these three selected values? A complete concentration analysis should be performed. *

Response:

This is of course a valid point. We showed data on these concentrations as established from previous studies of our collaborators (10.1111/cmi.13378; 10.1038/nature17625; 10.1038/s41467-019-09915-2). Under 3 µM we did not observe much measurable results and therefore omitted these. Concentrations above 70 µM did not change the outcome anymore than at 70 µM, so higher concentrations were omitted. We, thus, show 3µM at which we see mild effects, show 15 µM (a 5-fold increase compared to 3µM) at which we see profound effects and show 70 µM (again approximately a 5-fold increase compared to 15 µM) at which we see an overwhelming effect. Additional concentrations in between the applied concentration values would not add much new information.

6/ 169 formation was observed at 70 μ*M (Fig. 2b), which can be explained by neutrophil cell death induced by the toxin as determined by a DNA Sytox Green assay (Fig. S1a).

Was another viability test conducted? AnnexinV? Caspase 3/7? Sytox is not a specific staining in this regard. Furthermore, in Fig. S1a you state the kinetics of cell death, also after PMA treatment. On the one hand, you say that the production of candidalysin of NLS above 70 uM is reduced due to cell death, but at the same time you define as cell death the changes under PMA, which induce netosis. Please explain this reasoning better. *

Responses:

Thank you for pointing this out. We have no indication that candidalysin stimulates apoptosis in neutrophils. Therefore, no AnnexinV/Caspase 3/7 stain was performed. What we wanted to emphasize is that at 70 µM candidalysin the cytotoxic character of candidalysin is overwhelming leading to rather quick cell death, as assessed by the Sytox assay. Sytox is specific in the regard that it determines whether the plasma membrane is permeable and gives the stain access to the nuclear DNA to result in a positive signal. We use this assay to quantify NET formation, since it is a quantitative assay and less laborious than microscopy. However, we always back up NET assays with microscopic, image-based analyses and do not use the Sytox assay as standalone experiment for NET quantification, since the Sytox assay is not specifically staining netting cells, but it also stains other types of cell death.

We clarify this in the text as follows: (line 659) “Neutrophil cell death or the presence of extracellular DNA was quantified using a Sytox Green-based (Invitrogen) fluorescence assay similar to previous descriptions 2,35. To ultimately quantify NETs or NLS we always used image-based assys, the cell death assay was only used as complementation.”

*7/ 175 mixing of granular and nuclear components at ~120 min after stimulation (Fig. 2d and Fig. S2).

Figure S2 does not show mixing with the content of the granules. You are not labeling any granule component, only histones. You cannot draw that conclusion from these results. *

Response:

We respectfully disagree. As indicated in the figure legend for Figure 2d we were labelling for neutrophil elastase (red) which is located in azurophilic granules and thereby presents a marker for granular content. Since we wrongfully referred to Figure S2 here, we removed this from the text. The latter reference probably remained erroneously from a previous version.

*8/ Fig. 2. What concentration of PMA was used? What does 100% NLS mean? How is it different from 100% NET, since you are using PMA in both cases. Please explain. *

Responses:

We have now defined PMA concentration in the respective figure legend (100nM). The criteria for image-based assessment of NLS and NET quantification are the same for reason of comparison. PMA is included in each of the experiments as a positive control to show that the used neutrophils react upon stimulation. To clarify, we now specify at the y-axis %NETs or NLS. As stated above, 100% NLS means that each cell event in the image has increased in diameter such that it is considered as a NET or NLS. Hence, we use a common coordinate system to quantify extracellular events (NETs and NLS) based on size.

We have adjusted the figure legend as follows: (line 186) “Fig 2. Candidalysin induces ____NLS ____in human neutrophils. Candidalysin, but not scrambled candidalysin or pep2, another Ece1p-derived peptide (all 15 µM), induce (a) DNA decondensation in human neutrophils after 4 h (n = 4(10-14)) in a (b) dose-dependent manner (n = 3(10-14)). To allow comparability, NLS were quantified with the same criteria as previously described for NETs. Data shown as mean ± SEM. Confocal images (c) of immunostained cells display morphological changes involving nuclear and granular proteins after 4 h compared to unstimulated cells or 100 nM PMA, or cells exposed to scrambled candidalysin and pep2. The morphological changes evoked by PMA considerably deviate from morphological changes evoked by candidalysin and, hence, are defined as NETs (for PMA) and NLS (for candidalysin). Time-dependent progression of morphological changes (d) in neutrophils induced by candidalysin over the course of 5 h (all images are with 60X magnification).”

*9/ 181 NLS were quantified with the same criteria as previous described for NETs.

The criterion for NETs was an area above 100um2, so what is the criterion for NLS? If we assume that this is the same as for NETs, then what is the difference between NLS and NETs? The criteria adopted do not differentiate between the two forms and appear to be subjective. *

Responses:

As stated above, for us it was very important to find a common coordinate system to quantify NETs and NLS, since we wanted to deliver comparable and solid quantitative data. Hence, the quantification method does not discriminate between NETs and NLS. The notable morphological differences of NETs and NLS are thoroughly described with Figure 2 and Figure 3 and defined by differences in their structure. In addition, we present differences and similarities of induced pathways leading to canonical NETs or candidalysin-induced NLS in Figure 6 and Figure 7. We are convinced that, since NETs and NLS vary in size (DNA area covered), it will not be accurate for quantification purposes to include an additional size cut-off in the attempt to discriminate NLS and NETs. Instead we have established that candidalysin alone induces morphologically distinct NLS, whereas Candida albicans hyphae induce morphologically distinct NETs. By combination of quantitative data and image-based assessment, both structures can be discriminated from each other. In addition, we have established that during neutrophil and C. albicans interaction, citrullination of histone mainly stems from candidalysin. We show here and others have shown previously (10.3389/fimmu.2018.01573) that citrullination of histone occurs during but is not required for NET formation. But histone citrullination is promoted mainly by candidalysin and is also required for formation of NLS. Thus, histone citrullination constitutes another important discriminatory factor between NETs and NLS.

We added modified and added text to the respective figure legend: (line 188) ”To allow comparability, NLS were quantified with the same criteria as previously described for NETs. Data shown as mean ± SEM. Confocal images (c) of immunostained cells display morphological changes involving nuclear and granular proteins after 4 h compared to unstimulated cells or 100 nM PMA, or cells exposed to scrambled candidalysin and pep2. The morphological changes evoked by PMA considerably deviate from morphological changes evoked by candidalysin and, hence, are defined as NETs (for PMA) and NLS (for candidalysin).”

*10/ 190 allows a more detailed view of the neutrophil-derived structures (Error! Reference source not Please, eliminate this error. *

Response:

Thank you for pointing this out to us. We have fixed this error.

*11/ 193 Ionomycin has been previously reported to induce NLS, also... 194 Both, PMA and ionomycin generated widespread chromatin fibers in the extracellular space 197 In addition, C. albicans hyphae induced NETs with observable fibers and 198 threads similar to PMA- and ionomycin-stimulated neutrophils (Fig. 3b). 199 Image-based quantification of NLS events (candidalysin and ionomycin)

In a sentence earlier (193) you mentioned that the action of PMA leads to classical netosis and ionomycin leads to NLS. You pointed out earlier that NLS are poorly developed NETs (line 100), and here you write that PMA and ionomycin generate the same developed structures. You again differentiate between these structures depending on the stimulating factors. Pointing out the differences between the two forms, you should be more precise and consistent in your descriptions. This comment applies to the entire manuscript. *

Responses:

Thank you, we agree that consistency and clarity is required to describe the observed phenomena. We therefore modified or included the following sentences to the manuscript:

• (line 203) ”PMA exposure generated widespread chromatin fibres in the extracellular space (Fig. 3a, left panels) whereas ionomycin exposure resulted in more compact, patchy areas occasionally dispersed with long, thin chromatin fibres (Fig. 3b, middle panels). With regard to morphological changes, candidalysin treatment resulted in compact, fibrous structures resembling those stemming from ionomycin treatment, however long, thread-like structures were absent in candidalysin-treated neutrophil samples (Fig. 3a right panels, for 7 h treatment see Fig. S1c)”
• (Line 513) ”While ionomycin- and candidalysin-induced NLS shared similar key features, such as increased histone citrullination, our study revealed striking differences between the two toxins. In contrast to ionomycin, candidalysin stimulation led to ROS production in neutrophils.”

  12/ 203 NLS after 3 h and 5 h, respectively, and led to overall fewer NLS events. This was confirmed by observation. 204 area-based analysis of the events (Fig. 3d). The average area per event that exceeded 100 μ*m2 was 205 determined using the images from the DNA stain. What is the accepted criterion for distinguishing between NLS and NETs? *

Response:

The main criteria distinguishing canonical NETs from NLS is a higher compactness for NLS and an increased citrullination of histones, the latter being absent in canonical NETs (10.3389/fimmu.2016.00461; 10.1016/j.mib.2020.09.011). Please see our comment above (regarding reviewer comment 9). Comparing candidalysin and ionomycin as stimuli for NLS they share key similarities, such as increased citrullination of histone (Fig. 3) and more compact structures than NETs (Fig. 3) with an average size of 151 µm2 for candidalysin-induced and 149 µm2 for ionomycin-induced NLS compared to 262 µm2 for PMA-induced and 231 µm2 for C. albicans-induced NETs (for clarification these average sizes are stated in the text). However, the NLS triggered by candidalysin and ionomycin also show differences. Ionomycin occasionally results in extended chromatin threads, whereas candidalysin does not. Ionomycin induces no ROS at all, whereas candidalysin does to some extent. By consistent usage of the definitions for NETs and NLS and by pinpointing the differences between ionomycin and candidalysin in terms of NLS induction (which are previously unknown) we hope we have sufficiently addressed this comment.

*13/ line 218, 243 - reference error *

Response:

Thank you, we have fixed this error

14/ What form are we actually talking about? Are we focusing on the effect of a natural agent or a synthetic one in relation to NLS/NET? Perhaps it is more important to focus on the citrullination process.

• 247 synthetic candidalysin only induces NLS, we concluded that candidalysin augments NET release when the toxin is secreted by C. albicans hyphae. 256 This confirmed that candidalysin promotes C. albicans-triggered NET release. 262 Interestingly, the addition of synthetic candidalysin resulted in a shift to NLS, 274 External addition of synthetic candidalysin resulted in a shift to NLS structures rather than NETs as visualized by microscopy after 5 h incubation (20X).*

Response:

We used the adjective “synthetic” here to make clear that this is a synthetized peptide and not candidalysin isolated from growing C. albicans. Having said that, we fully agree that the synthetized peptide and the one released by C. albicans cells are essentially identical on the molecular level and thus it is irrelevant and confusing to state in this context here. Therefore, we removed the adjective “synthetic” throughout the study and refer the reader to the method section for information on the origin of candidalysin used in the study. At times, we state “candidalysin alone” when we want to emphasize that candidalysin was the sole trigger used for the respective assay.

15/ Has there been any method to track candidalysin production during contact of C. albicans with neutrophils?

Responses:

Thank you for this comment. Yes, there is a QVQ nanobody that can be used which is currently not to our disposal (doi.org/10.1111/cmi.13378). However, we already know from this publication that candidalysin concentrations vary when released naturally. The concentrations are particularly high in invasion pockets or dense biofilms. We also know that if we add candidalysin to the medium we have even distribution throughout and this is by definition different from concertation spikes at host cell-fungal interaction sites. As we have stated above, hyphal tips in turn are readily attacked by human neutrophils (doi.org/10.1189/jlb.0213063). Hence, we can safely assume, according to these previous publications, that there will be a more uneven distribution of candidalysin concentrations over neutrophil membranes, when the sole source of the toxin stems from growing hyphae interacting with neutrophils. It would of course be very interesting to know how the toxin exactly intercalates into membranes and which morphologies potential pores may have. These questions are currently under investigation in the laboratories of B. Hube and J. Naglik. To incorporate these findings here would certainly be far beyond the scope of this study.

We include and modify the following sentences to the discussion of this manuscript to clarify the issue: (Line 544). ” Lack of candidalysin production in C. albicans results in significantly reduced histone citrullination, accompanied with decreased NET formation. However, citrullination is not required for NET release, but rather governs the formation of NLS, which is dominant when candidalysin is added exogenously with even distribution throughout the cell suspension. With regard to C. albicans hyphae secreting candidalysin, local concentrations of the toxin are likely to vary to a large degree, particularly when the candidalysin-secreting hypha is engulfed by a neutrophil. Therefore, it may be difficult to discriminate NLS form NETs during the interaction of neutrophils and C. albicans, as both structures may be induced concurrently 10. It seems logical that the pore-forming activity of candidalysin augments the release of NET fibres during C. albicans infection, where PRRs will additionally be triggered on neutrophils, resulting in combinatorial activation of downstream pathways. In line with this notion, candidalysin drives histone citrullination, which contributes to chromatin decondensation.”

*16/ In Figure 4f-the given information indicates 1,2 hour incubation, in the caption of the figure there is information about 5 hour incubation - please clarify. The description of the stains used is lacking. *

Response:

Microscopic analysis performed after 5h incubation time, whereas candidalysin has been added to different time points indicated in the Figure (in the new version this is now Figure 8f). We clarified in the legend as follows: (line 472) “(f) Neutrophils were infected with C. albicans and 15 µM candidalysin was added 0 h, 1 h or 2 h after the infection. Addition of candidalysin at the different time points after C. albicans infection resulted in a shift to NLS structures rather than NETs as visualized by microscopy after 5 h total incubation (20X).” The description of the strains is depicted directly in the Figure, next to the microscopic images.

*17/ Fig. 5 - result for 15 uM MitoTEMPO - adds nothing to the results and introduces image information noise - should be removed. No information on the concentration of the peptide used. *

Responses:

We would like to keep the 15 µM MitoTEMPO concentration, since it is the more reasonable concentration at which we do not observe an effect. This argues that ROS is more-likely derived from NADPH oxidase and not mitochondrial ROS. We show TEMPOL effects at 15 µM and at 100 µM to document the dose dependency and for the sake of comparability, we would like to keep both concentrations also for MitoTEMPO.

The indicated peptide concentration was added to the figure legend. Thank you for pointing this out.

*18 / Fig. 5, line 309: and cell-permeable Sytox Green DNA dye (250310 nM) to determine the total number of cells".

Please correct the information on the use of both dyes, according to the manufacturer's description: "SYTOX® Green nucleic acid stain is an excellent green-fluorescent nuclear and chromosome counterstain that is impermeant to live cells, making it a useful indicator of dead cells within a population." *

Response:

Thank you for highlighting this error. Indeed, we used Syto Green for this particular staining, a dye which stains both live and dead cells since the dye is cell-permeable. We corrected the error at this section of the text.

*19/ 324 At later time points, BAPTA-AM led to an increase in NLS, probably due to toxic effects as indicated by higher background levels of NLS formation in non-stimulated, BAPTA-AM-treated neutrophils (Fig. 6d).

If such an assumption is made, the toxic effect should also be observed for the control. *

Response:

The toxic effect was observed while conducting the experiments, but cannot be seen in the size-base quantification which is the read out for this particular experiment. We have performed a cytotoxicity assay using flow cytometry and PI staining to confirm the effect. The results are added as supplemental Figure (Fig. S3b).

*20/Fig. 6C PAD inhibitor should affect PMA-induced netosis, but the figure presents NLS existence - how was this change found? *

Responses:

We are grateful for the opportunity to explain this more thoroughly. PMA does not trigger histone citrullination (10.3389/fimmu.2016.00461) and thereby there is no effect of the PAD inhibitor on PMA-induced NETs. Notably, some level of histone citrullination can also be observed in unstimulated neutrophils (see Fig. 3, 5 and 8), since histone modification is not exclusively dependent on stimulation. However, upon PMA stimulation we observe a decrease (Fig. S1b), not an increase, of histone citrullination consistent with previous reports.

We adjusted the text as follows: (line 235) “. Expectedly, citH3 levels upon PMA stimulation did not increase, but rather decreased which is consistent with previous reports 10 (Fig. 3d and Suppl. Fig. S1b). While citrullination levels in unstimulated neutrophils decreased over time, ionomycin stimulation sustained high levels over 5 h.

*21/ line320 "This indicates that candidalysin most probably causes Ca2+ influx via pore formation and not via direct receptor stimulation" And: line 358. As C.albicans hyphae bind to pathogen recognition receptors (PRRs), activate neutrophils and ultimately promote the release of NETs, we aimed to elucidate whether candidalysin alone leads to the activation of similar pathways in neutrophils. Hence, we stimulated neutrophils with candidalysin in the presence or absence of specific inhibitors for SYK, PI3K, and Akt.

Lack of consistency in conclusion. *

Response:

Thank you for pointing this out. We adjusted the paragraph (line 331) as follows: “As C. albicans hyphae bind to pathogen recognition receptors (PRRs), activate neutrophils and ultimately promote the release of NETs, we aimed to elucidate whether candidalysin alone can trigger similar pathways in neutrophils via signalling cross talk induced by Ca2+ influx. Hence, we stimulated neutrophils with candidalysin in the presence or absence of specific inhibitors for SYK, PI3K, and Akt (Fig. 6b).”

*22/ Fig. 7 It would be good to verify these results with experiments using mutants. Figures 7b, 7c, and 7d can be combined to make the whole drawing clearer. *

Response:

We thought this is very relevant and included additional experiments showing that the mutant strains also induce phosphorylation of lamin A/C independent of the expression of candidalysin (new Fig. 6e and 6f).

*23/ line 603 'The percentage of dead cells was calculated using TritonX-100 lysed neutrophils as 100% control' - maybe use " treated or permeabilized" *

Response:

Thank you, we changed the phrasing accordingly.

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

The study presented the role of cytolic fungal toxin - candidalysin, secreted by the hyphal form of Candida albicans, in the formation of neutrophil extracellular traps during C. albicans contact with neutrophils, which serve as the first line of the responses. The key conclusions are convincing. The authors considered the whole mechanism of NLT formation, which explained previous observations made by others. Some of the additional proposed experiments are not necessary to perform. They would only complement the results already presented by the authors. A few missing citations in the text need to be filled in. The others have been used appropriately and present earlier work on the subject. The reviewer indicated minor corrections to the drawings in the detailed comments. This paper is very interesting and is crucial to understanding some unusual observations made earlier about NET production in fungal infections. On the other hand, the text requires minor corrections to understand better the occurrence of both forms of extracellular neutrophil traps. The authors of the paper have experience in the subject matter presented, and the lead authors are among the leading researchers who analyze this problem. The presented research is a valuable addition to the previous work.

Significance

The advantage of this work is the presentation of the mechanism associated with NLS formation in contact with candidalysin, where activation of NADPH oxidase and calcium influx have been documented to be important. This toxin can trigger ROS production and activate downstream signaling that is important for morphological changes and NLS formation. The important finding is also that NLS are resistant to nuclease treatment and increase the ability of neutrophils to control C. albicans hyphae formation and fungal cell growth. These findings provide a better understanding of the role of neutrophils in the treatment of infections caused by these microorganisms.

Below I present are minor suggestions that, in my opinion, will improve the text and correct the presentation of the results, making this set of results a valuable source for explaining such a complex problem.

1. The authors should decide what thesis about NLS they want to prove: 100 NLS are less fibrous and ....... than canonical NETs and are triggered in an NADPH oxidase-independent fashion. 121 NLS were dependent on NADPH oxidase-mediated reactive oxygen species (ROS) production
2. for the experiment described in the line below, MOI 2 was chosen; did the authors conduct an analysis of the response/eventual change in it, depending on the MOI?

131 we infected neutrophils with wild-type C. albicans, ECE1-deficient (ece1ΔΔ), and corresponding revertant (ece1ΔΔ+ECE1) strains, 3. Has the effect of deletion of ECE1 on other aspects of virulence, such as adhesion, virulence factor production, or biofilm formation, been analyzed?

137 the ECE1- and candidalysin-deficient strains triggered reduced levels 4. Fig.1 - How were C. albicans cells stained? Does 100%NET mean the number of cells netting after PMA treatment? This information should be given. 5. 168 dependent effect with increased NLS formation from 3 μM to 15 μM. However, the reduced NLS

How was determined the limiting concentration value of the toxin, for which an increase in NLS was observed? Was a wide range of concentrations used in the analysis or was the determination made only for these three selected values? A complete concentration analysis should be performed. 6. 169 formation was observed at 70 μM (Fig. 2b), which can be explained by neutrophil cell death induced by the toxin as determined by a DNA Sytox Green assay (Fig. S1a).

Was another viability test conducted? AnnexinV? Caspase 3/7? Sytox is not a specific staining in this regard. Furthermore, in Fig. S1a you state the kinetics of cell death, also after PMA treatment. On the one hand, you say that the production of candidalysin of NLS above 70 uM is reduced due to cell death, but at the same time you define as cell death the changes under PMA, which induce netosis. Please explain this reasoning better. 7. 175 mixing of granular and nuclear components at ~120 min after stimulation (Fig. 2d and Fig. S2).

Figure S2 does not show mixing with the content of the granules. You are not labeling any granule component, only histones. You cannot draw that conclusion from these results. 8. Fig. 2. What concentration of PMA was used? What does 100% NLS mean? How is it different from 100% NET, since you are using PMA in both cases. Please explain. 9. 181 NLS were quantified with the same criteria as previous described for NETs.

The criterion for NETs was an area above 100um2, so what is the criterion for NLS? If we assume that this is the same as for NETs, then what is the difference between NLS and NETs? The criteria adopted do not differentiate between the two forms and appear to be subjective. 10. 190 allows a more detailed view of the neutrophil-derived structures (Error! Reference source not Please, eliminate this error. 11. 193 Ionomycin has been previously reported to induce NLS, also... 194 Both, PMA and ionomycin generated widespread chromatin fibers in the extracellular space 197 In addition, C. albicans hyphae induced NETs with observable fibers and 198 threads similar to PMA- and ionomycin-stimulated neutrophils (Fig. 3b). 199 Image-based quantification of NLS events (candidalysin and ionomycin)

In a sentence earlier (193) you mentioned that the action of PMA leads to classical netosis and ionomycin leads to NLS. You pointed out earlier that NLS are poorly developed NETs (line 100), and here you write that PMA and ionomycin generate the same developed structures. You again differentiate between these structures depending on the stimulating factors. Pointing out the differences between the two forms, you should be more precise and consistent in your descriptions. This comment applies to the entire manuscript. 12. 203 NLS after 3 h and 5 h, respectively, and led to overall fewer NLS events. This was confirmed by observation. 204 area-based analysis of the events (Fig. 3d). The average area per event that exceeded 100 μm2 was 205 determined using the images from the DNA stain. What is the accepted criterion for distinguishing between NLS and NETs? 13. line 218, 243 - reference error 14. What form are we actually talking about? Are we focusing on the effect of a natural agent or a synthetic one in relation to NLS/NET? Perhaps it is more important to focus on the citrullination process:

247 synthetic candidalysin only induces NLS, we concluded that candidalysin augments NET release when the toxin is secreted by C. albicans hyphae. 256 This confirmed that candidalysin promotes C. albicans-triggered NET release. 262 Interestingly, the addition of synthetic candidalysin resulted in a shift to NLS, 274 External addition of synthetic candidalysin resulted in a shift to NLS structures rather than NETs as visualized by microscopy after 5 h incubation (20X). 15. Has there been any method to track candidalysin production during contact of C. albicans with neutrophils? 16. In Figure 4f-the given information indicates 1,2 hour incubation, in the caption of the figure there is information about 5 hour incubation - please clarify. The description of the stains used is lacking. 17. Fig. 5 - result for 15 uM MitoTEMPO - adds nothing to the results and introduces image information noise - should be removed. No information on the concentration of the peptide used. 18. Fig. 5, line 309: and cell-permeable Sytox Green DNA dye (250310 nM) to determine the total number of cells".

Please correct the information on the use of both dyes, according to the manufacturer's description: "SYTOX® Green nucleic acid stain is an excellent green-fluorescent nuclear and chromosome counterstain that is impermeant to live cells, making it a useful indicator of dead cells within a population." 19. 324 At later time points, BAPTA-AM led to an increase in NLS, probably due to toxic effects as indicated by higher background levels of NLS formation in non-stimulated, BAPTA-AM-treated neutrophils (Fig. 6d).

If such an assumption is made, the toxic effect should also be observed for the control. 20. Fig. 6C PAD inhibitor should affect PMA-induced netosis, but the figure presents NLS existence - how was this change found? 21. line320 "This indicates that candidalysin most probably causes Ca2+ influx via pore formation and not via direct receptor stimulation" And: line 358. As C.albicans hyphae bind to pathogen recognition receptors (PRRs), activate neutrophils and ultimately promote the release of NETs, we aimed to elucidate whether candidalysin alone leads to the activation of similar pathways in neutrophils. Hence, we stimulated neutrophils with candidalysin in the presence or absence of specific inhibitors for SYK, PI3K, and Akt.

Lack of consistency in conclusion. 22. Fig. 7 It would be good to verify these results with experiments using mutants. Figures 7b, 7c, and 7d can be combined to make the whole drawing clearer. 23. line 603 'The percentage of dead cells was calculated using TritonX-100 lysed neutrophils as 100% control' - maybe use " treated or permeabilized"

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

This study evaluates the effect of fungal toxin candidalysin on neutrophils. The authors show that candidalysin induces NETosis when secreted by hyphae, but when candidalysin is added on its own, NLS are formed instead which are distinct from NETs. The authors have done lots of carefully controlled experiments, and delineated key components of the pathway inducing NLS, including the role of ROS and histone modifications. The data provided is high quality and well presented in figures.

Significance

Strengths are the depth of analysis - many different aspects of NETosis is assessed and robustly tested.

Comments:

1. I was a bit confused by what should be the main message of the paper - is it that candidalysin on its own doesn't induce NETosis but only NLS? The answer to this question wasn't well addressed in my opinion, but the paper switches between using live fungi and purified candidalysin so it became confusing at times.
2. If candidalysin on its own only induces NLS - what is the relevance of this for disease? A lot of work has been provided on the pathway driving NLS formation, but it wasn't clear to me why this is important. More in discussion needed or evidence of disease relevance.
3. In Figure 2, it would be helpful to include images of ionomycin-stim neutrophils for comparison of the NLS structures across different stim conditions.
4. Few places where reference manager has failed (see bottom on page 10, line 190 for example)
5. Lines 191-198 - I was confused here by the text. I thought the point was that candidalysin induced NLS similar to ionomycin, but here the point is being made that the two are different? This led me to being confused as to the point of all the comparisons made between ionomycin NLS and candidalysin NLS... this could be made clearer.
6. Could the authors include some unstim neutrophil control images in Fig 3 for the SEM? Can the SEM sample processing affect neutrophil structure in anyway? Feels like an important control although I don't have much experience with SEM personally.
7. I was very intrigued by the experiments where the authors added candidalysin in to neutrophils infected with ece1-null strain. Those experiments showed that candidalysin addition still drove NLS instead of NETosis. Can the authors investigate why this is? Is membrane intercalation different when candidalysin is delivered by hyphae vs added on its own? Could that explain some of the differences they have seen?
8. Is phagocytosis needed for NETosis induction by candidalysin? What happens if you add beads or beta-glucan particles with candidalysin stimulation? Do you get NLS or NETs?
9. Please confirm what the n numbers refer to in the figure legends - are these biological or technical replicates? How many experiments are the representative images representing?

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

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

1. General Statements

We would like to thank the reviewers for their professional comments and constructive suggestions. Our current plan is to revise the manuscript and supplemental materials in response to the reviewers’ requests and suggestions. Toward this goal we began experiments to obtain new data requested by the reviewers and anticipate the outlined experiments can be completed within the next three months.

2. Description of the planned revisions:

__Reviewer #1: __

The results of experiments where Arp2/3 is blocked (Fig.2) should be confirmed by Arp2/3 knock-down and with an independent Arp2/3 inhibitor. Several are available (CK-869, Benproperine, Pimozide). For Fig.3 and 4, that would not be necessary, but to establish the specificity of the effect in fig.2 this is absolutely required.

__Response: __As requested, we will include new data with CK-869, the indicated Arp2/3 complex inhibitor. We purchased the inhibitor and are currently confirming its efficacy before testing whether it inhibits transition to the hESC naïve state. However, we respectfully disagree with generating a Arp2/3 knock-down hESC line. Arp2/3 complex genes are known to be essential genes in both mouse and human embryonic stem cells (PMID: 29662178 and PMID: 31649057). Furthermore, reports on successful knockout of complex subunits indicate that additional genetic manipulations are needed to maintain cell survival, including knockout of INK4A/ARF to bypass apoptosis associated with Arp2 shRNA knockdown (PMID: 22385962) and genetic manipulations in mouse models (PMID: 22492726. Thus, knock-down of Arp2/3 complex members in our cells is beyond the scope of this manuscript.

Yilmaz A, Peretz M, Aharony A, Sagi I, Benvenisty N. Defining essential genes for human pluripotent stem cells by CRISPR-Cas9 screening in haploid cells. Nat Cell Biol. 2018 May;20(5):610-619. doi: 10.1038/s41556-018-0088-1. Epub 2018 Apr 16. PMID: 29662178.

Shohat S, Shifman S. Genes essential for embryonic stem cells are associated with neurodevelopmental disorders. Genome Res. 2019 Nov;29(11):1910-1918. doi: 10.1101/gr.250019.119. Epub 2019 Oct 24. PMID: 31649057; PMCID: PMC6836742.

Wu C, Asokan SB, Berginski ME, Haynes EM, Sharpless NE, Griffith JD, Gomez SM, Bear JE. Arp2/3 is critical for lamellipodia and response to extracellular matrix cues but is dispensable for chemotaxis. Cell. 2012 Mar 2;148(5):973-87. doi: 10.1016/j.cell.2011.12.034. PMID: 22385962; PMCID: PMC3707508.

I believe that the status of the actin cytoskeleton in both states is not well enough characterized. This is especially obvious for branched actin networks themselves that depend on the Arp2/3. To this end, the authors may localize Arp2/3 or cortactin, a useful surrogate that often gives a better staining. This point is particularly important since contractile fibers are not made of branched actin. Myosin cannot walk or pull along branched actin networks because of steric hindrance. It might well be that branched actin networks are debranched after Arp2/3 polymerization. I suggest staining tropomyosins that would indicate where the transition between branched and unbranched actin would be. Along this line, phosphoERMs should be localized and revealed by Western blots (we expect an increase from primed to naive state) because they cannot perform the proposed function of linker between the membrane and actin filaments if they are not phosphorylated.

Response: As requested we will include new data with cortactin immunolabeling, which we already completed. These new data, shown below, confirm that cortactin, which binds to branched actin filaments, co-localizes with the F-actin fence around hESC naïve colonies, suggesting that the fence includes branched F-actin. Also as requested, we are currently immunoblotting for phosphorylated ERMs to more thoroughly assess if they may serve as a linker between the membrane and actin filaments.

Branched actin is required for cell cycle progression and cell proliferation in normal cells. This requirement is lost in most cancer cells (Wu et al., Cell 2012; Molinie et al., Cell Res 2019). This would be really important to know whether ESCs stop proliferating upon CK-666 treatment. In other words, do they behave like normal cells or transformed cells. Proliferation is a major function that depends on the YAP pathway. Cell counts and EdU incorporation can easily provide answers to this important question.

__Response: __As requested, we will include new data on proliferation. We anticipate that these new data will complement data we already have showing that CK-666 does not impair proliferation compared with hESC controls. We also note that the role of the actin cytoskeleton in proliferation is well established and an increase in proliferation is a hallmark of acquisition of the naïve state of pluripotency (PMID: 35005567).

Chen C, Zhang X, Wang Y, Chen X, Chen W, Dan S, She S, Hu W, Dai J, Hu J, Cao Q, Liu Q, Huang Y, Qin B, Kang B, Wang YJ. Translational and post-translational control of human naïve versus primed pluripotency. iScience. 2021 Dec 17;25(1):103645. doi: 10.1016/j.isci.2021.103645. PMID: 35005567; PMCID: PMC8718978.

Minor Comments:

What about the rescue of cell morphology? Does active YAP restore the intercellular contractile bundle?

__Response: __As requested, we obtained these data, as shown below. Expression of the YAP-S127A mutant does rescue the formation of the actin ring architecture in the presence of CK666. We are currently performing additional dedifferentiation assays to immunolabel for pMLC and address the question of if expression of YAP-S127A restores the contractile bundle.

__Reviewer #2: __

The authors found that a ring of actin filaments at the colony periphery was characteristic of the naive hESCs. However, because all the data are presented as an image of a single confocal section, the 3D organization of the actin filaments is not clear. Although the authors drew a scheme for this actin ring being located in the apical domain of polarized cells, such data have not been provided in the manuscript. Since naive hESCs form dome-like colonies, it is important to show the 3D organization of actin filaments in the colony. 3D reconstruction of confocal microscopy images of the naive hESC colonies is required to show the relationship between actin filaments, adherens junctions, and the nuclei (as a reference for the Z axis). If 3D reconstruction is not technically possible, confocal images at different Z levels and maximum projection images should be obtained and provided.

__Response: __As requested, we are currently generating 3D images of the actin fence by using Imaris software, which we previously used to show 3D images of mitochondrial morphology (PMID: 34038242)

Manoli SS, Kisor K, Webb BA, Barber DL. Ethyl isopropyl amiloride decreases oxidative phosphorylation and increases mitochondrial fusion in clonal untransformed and cancer cells. Am J Physiol Cell Physiol. 2021 Jul 1;321(1):C147-C157. doi: 10.1152/ajpcell.00001.2021. Epub 2021 May 26. PMID: 34038242; PMCID: PMC8321791.

Some of the statistical analyses were inappropriate. The authors have used Student's t-test for all analyses; however, one-way ANOVA and post-hoc analysis must be used to compare three or more groups (Figs. 2B, D, E, 3G, 4B, D, E).

__Response: __As requested, we will re-evaluate our statistical analysis. We note that our submission reports comparisons between two groups, and hence, Student’s t-test is appropriate. For example, we compared primed and naïve to demonstrate successful acquisition of naïve pluripotency, and then we compared the naïve condition to the CK666-treated conditions to demonstrate the impact of CK666-treatment. As Reviewer 2 suggests we will reanalyze all quantifications using one-way ANOVA with post-hoc analysis in the full revision and we will also discuss with Stuart Gansky, a statistician at UCSF whom we previously consulted for most appropriate statistical analysis of our studies.

Minor Comments:

Page 9, second paragraph. In the discussion section, authors have written that "Cells within the ICM of mouse blastocysts exclude YAP from the nucleus whereas cells within the ICM of human blastocysts maintain nuclear YAP." However, a recent study has reported that the ICM/epiblast of mouse late blastocysts also express nuclear YAP. Epiblast Formation by TEAD-YAP-Dependent Expression of Pluripotency Factors and Competitive Elimination of Unspecified Cells. Hashimoto M, Sasaki H. Dev Cell. 2019, 50:139-154.e5. doi: 10.1016/j.devcel.2019.05.024.

__Response: __As requested, we will revise our Discussion section to include findings from the indicated new publication.

Reviewer #3:

Many of their conclusions seem to be based on the qualitative analysis of a single image (e.g. Figures 1D-G, Fig 2G, Supplementary Figure 2). The authors should provide quantitative information regarding these analyses and indicate the number of cells/replicas collected for each experiment.

__Response: __As requested, our revision will have added quantitative data when feasible. We note that in the field, traction force microscopy isn’t commonly quantified beyond including scale bars, which our original manuscript shows. Moreover, pluripotency is standardly not quantified because it is a binary switch - cells are either double positive or they are not. We show 100% double positive, and rtPCR data with known stage-specific markers.

The actin ring surrounding hESCs colonies was previously described by Närvä et al. Although the authors cited this previous work, they do not discuss in deep the differences and similarities with their observations.

__Response: __As requested, our revised manuscript with include additional detail comparing our results with those from Närvä et al. In brief, we observe the formation of this actin ring only in the naïve state of pluripotency, whereas Närvä et al. observe an actin architecture in the primed state. One possible source of difference between their study and ours are the cells used for analysis. Närvä et al. utilize induced pluripotent stem cells, long since proposed to be closer to naïve pluripotency than primed stem cells as conventionally isolated and maintained (see PMID: 27424783 and PMID: 19497275). Additionally, we observe that the contractile actin ring in naïve pluripotent stem cells is in a higher z-plane than reported by Närvä et al., although a direct comparison is difficult to make.

Theunissen TW, Friedli M, He Y, Planet E, O'Neil RC, Markoulaki S, Pontis J, Wang H, Iouranova A, Imbeault M, Duc J, Cohen MA, Wert KJ, Castanon R, Zhang Z, Huang Y, Nery JR, Drotar J, Lungjangwa T, Trono D, Ecker JR, Jaenisch R. Molecular Criteria for Defining the Naive Human Pluripotent State. Cell Stem Cell. 2016 Oct 6;19(4):502-515. doi: 10.1016/j.stem.2016.06.011. Epub 2016 Jul 14. PMID: 27424783; PMCID: PMC5065525.

Nichols J, Smith A. Naive and primed pluripotent states. Cell Stem Cell. 2009 Jun 5;4(6):487-92. doi: 10.1016/j.stem.2009.05.015. PMID: 19497275.

The qualitative observation of Figure 3F suggests a lower overall YAP levels in primed and +CK666 cells in comparison to naive cells. Could the authors check if this is correct and, if this is the case, explain the observation?

__Response: __As requested, our revision will include new data on YAP protein expression by immunoblotting.

The authors should discuss deeper the rationale of the pan-ERM immunostaining experiments (since they used the individual antibodies afterwards) and provide a brief discussion of their results and, in particular, the colocalization with moesin but not with ezrin or radixin.

__Response: __As requested, our revised manuscript will include a more detailed discussion of our results with ERM immunolabeling.

2. Description of the revisions that have already been incorporated in the transferred manuscript:

__Reviewer #1: __

Minor Comments:

Fig2F: non-representative pictures or wrong quantification of the CK666 condition.

__Response: __We thank the review for alerting us to this error. The CK666 Primed and Naïve condition images were swapped. We have edited the figure to correct this.

Fig3A: Y-axis? What is it? How is it adjusted? -Log P?

__Response: __Please see the methods section. Differential expression analysis was performed using DESeq2 R package. The resulting P values were adjusted (padj) using the Benjamini and Hochberg’s approach for controlling the False Discovery Rate (FDR). Genes with a padj

Colors of dots not really visible (in reference to Figure 3A).

__Response: __We thank the reviewer for this comment and have updated the figure to use more standard, colorblind-friendly color choices (see the above figure). Additionally, we fixed a drawing error in the figures when creating the volcano plots.

Typos: Apr2/3 in the abstract, Hoeschst in Fig.S1B.

__Response: __We thank the review for alerting us to these errors. We have edited the manuscript to correct them.

__Reviewer #3: __

There are many experimental details missing that are extremely relevant to fully understand the experiments and evaluate the robustness of the analyses (e.g., microscopy setup, fluorescent probes used for immunostaining, incubation conditions with the inhibitors SMIFH2 and CK666).

__Response: __As requested we have updated the Materials and Methods section with more detailed information on procedures and reagents.

Minor Comments:

The Introduction makes the reader think that actin is the only cytoskeletal network involved in embryo development and stem cell properties. They should also include a brief discussion on the relevance of the other cytoskeletal networks in mechanotransduction and cell fate decisions.

__Response: __As requested, we will revise our Introduction. We note, however, that in the field additional cytoskeleton components, including intermediate filaments and microtubules have mostly been shown for interacting with the nucleus with limited evidence for roles in differentiation.

Many of the images seem to require a flat-field correction. Could the authors check that the illumination is homogeneous? This artifact could affect the data analysis.

__Response: __As we indicate in the Methods section, the spinning disc confocal microscopes used in our study are equipped with a Borealis to mitigate uneven illumination across the field of view. Additionally, quantification in Figures 2C-E, Figures 3F-G, and Figures 4A-D are comparing measurements to a local background (i.e. cytoplasm nearby) in order to normalize for any uneven illumination effects.

There are many abbreviations that are not defined in the text and are extremely specific to the field.

__Response: __As requested, we have expanded the definition of many abbreviations in the text and any additional abbreviations changes will be clearly defined in our revised manuscript.

Could the authors explain the selection of the pluripotency markers studied by qPCR? Specifically, why they studied DNMT3L, DPPA3, KLF2, and KLF4 (Fig. 1B) and the different set PECAM1, ESRRB, KLF4, and DNMT3L in Fig. 2B.

__Response: __Defining the exact molecular and cell behavioral characteristics of naïve pluripotency remains an evolving point of development within the field. The pluripotency markers used in both original panels are known and established markers of naïve pluripotency. The original panel of DNMT3L, DPPA3, KLF2, and KLF4 was established based upon RNAseq datasets publicly available, whereas the secondary panel of PECAM1, ESRRB, KLF4, and DNMT3L was a more targeted analysis of genes found in the literature which have been interrogated in more detail for potential roles in naïve pluripotency. To facilitate clarity within the manuscript, we have updated Fig. 1B to match Fig. 2B for the purposes of defining a transcriptional hallmark of naïve pluripotency for the purposes of this manuscript.

Figures 1G and 2G, please include the images of the colonies.

__Response: __As requested, our revised manuscript will include phase contrast images, which we already have, as shown below. These images will be included Supplemental Figure 1 and Supplemental Figure 2 for the colonies used to show representative tractions in Figure 1G and 2G, respectively.

3. Description of analyses that authors prefer not to carry out

Reviewer #1:

The results of experiments where Arp2/3 is blocked (Fig.2) should be confirmed by Arp2/3 knock-down and with an independent Arp2/3 inhibitor. Several are available (CK-869, Benproperine, Pimozide). For Fig.3 and 4, that would not be necessary, but to establish the specificity of the effect in fig.2 this is absolutely required.

__Response: __As requested, we will include new data with CK-869, the indicated Arp2/3 complex inhibitor. We purchased the inhibitor and are currently confirming its efficacy before testing whether it inhibits transition to the hESC naïve state. However, we respectfully disagree with generating a Arp2/3 knock-down hESC line. Arp2/3 complex genes are known to be essential genes in both mouse and human embryonic stem cells (PMID: 29662178 and PMID: 31649057). Furthermore, reports on successful knockout of complex subunits indicate that additional genetic manipulations are needed to maintain cell survival, including knockout of INK4A/ARF to bypass apoptosis associated with Arp2 shRNA knockdown (PMID: 22385962) and genetic manipulations in mouse models (PMID: 22492726. Thus, knock-down of Arp2/3 complex members in our cells is beyond the scope of this manuscript.

Yilmaz A, Peretz M, Aharony A, Sagi I, Benvenisty N. Defining essential genes for human pluripotent stem cells by CRISPR-Cas9 screening in haploid cells. Nat Cell Biol. 2018 May;20(5):610-619. doi: 10.1038/s41556-018-0088-1. Epub 2018 Apr 16. PMID: 29662178.

Shohat S, Shifman S. Genes essential for embryonic stem cells are associated with neurodevelopmental disorders. Genome Res. 2019 Nov;29(11):1910-1918. doi: 10.1101/gr.250019.119. Epub 2019 Oct 24. PMID: 31649057; PMCID: PMC6836742.

Wu C, Asokan SB, Berginski ME, Haynes EM, Sharpless NE, Griffith JD, Gomez SM, Bear JE. Arp2/3 is critical for lamellipodia and response to extracellular matrix cues but is dispensable for chemotaxis. Cell. 2012 Mar 2;148(5):973-87. doi: 10.1016/j.cell.2011.12.034. PMID: 22385962; PMCID: PMC3707508.

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 #3

Evidence, reproducibility and clarity

The manuscript entitled "Arp2/3 Complex Activity Enables Nuclear YAP for Naïve Pluripotency of Human Embryonic Stem Cells" by Nathaniel Meyer, Tania Singh, Matthew Kutys, Todd Nystul, and Diane Barber analyzes the formation of the actin ring that surrounds naive but not primed colonies of human embryonic stem cells (hESCs). The authors claim that the formation of this actin ring requires Arp2/3 which also modulates YAP localization. Despite the overall topic is relevant to understand key aspect of stem cells and embryo development, I have found several flaws in the manuscript (stated below) that, in my opinion, prevent its publication in its current form:

1. Many of their conclusions seem to be based on the qualitative analysis of a single image (e.g. Figures 1D-G, Fig 2G, Supplementary Figure 2). The authors should provide quantitative information regarding these analyses and indicate the number of cells/replicas collected for each experiment.
2. Many of the images seem to require a flat-field correction. Could the authors check that the illumination is homogeneous? This artifact could affect the data analysis.
3. The actin ring surrounding hESCs colonies was previously described by Närvä et al. Although the authors cited this previous work, they do not discuss in deep the differences and similarities with their observations.
4. There are many experimental details missing that are extremely relevant to fully understand the experiments and evaluate the robustness of the analyses (e.g., microscopy setup, fluorescent probes used for immunostaining, incubation conditions with the inhibitors SMIFH2 and CK666).
5. The qualitative observation of Figure 3F suggests a lower overall YAP levels in primed and +CK666 cells in comparison to naive cells. Could the authors check if this is correct and, if this is the case, explain the observation?
6. The authors should discuss deeper the rationale of the pan-ERM immunostaining experiments (since they used the individual antibodies afterwards) and provide a brief discussion of their results and, in particular, the colocalization with moesin but not with ezrin or radixin.

Minor observations:

1. The Introduction makes the reader think that actin is the only cytoskeletal network involved in embryo development and stem cell properties. They should also include a brief discussion on the relevance of the other cytoskeletal networks in mechanotransduction and cell fate decisions.
2. There are many abbreviations that are not defined in the text and are extremely specific to the field.
3. Could the authors explain the selection of the pluripotency markers studied by qPCR? Specifically, why they studied DNMT3L, DPPA3, KLF2, and KLF4 (Fig. 1B) and the different set PECAM1, ESRRB, KLF4, and DNMT3L in Fig. 2B.
4. Figures 1G and 2G, please include the images of the colonies.

Significance

The manuscript analyzes the formation of the actin ring that surrounds naive but not primed colonies of human embryonic stem cells (hESCs). The authors claim that the formation of this actin ring requires Arp2/3 which also modulates YAP localization. Despite the overall topic is relevant to understand key aspect of stem cells and embryo development, I have found several flaws in the manuscript (stated below) that, in my opinion, prevent its publication in its current form (see Evidence, reproducibility and clarity)

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

Summary:

This paper describes the involvement of the actin regulator Arp2/3 in the dedifferentiation of primed human embryonic stem cells (hESCs) into naive pluripotency. The authors initially demonstrated a reorganization of the actin cytoskeleton during the transition to naive pluripotency, which included the formation of a contractile actin ring at the colony periphery. Actin reorganization was also associated with a reduction in cell-substrate traction forces and an increase in cell-cell junction traction forces. The authors showed that the activity of the Arp2/3 complex was required for actin reorganization and acquisition of a naive pluripotent state. RNA-seq analysis revealed that the Arp2/3 complex regulates Hippo signaling. Furthermore, inhibition of Arp2/3 suppressed the nuclear localization of YAP, and expression of nuclear-localized YAP restored naive dedifferentiation of the Arp2/3 inhibited hESCs. Based on these results, the authors have proposed a model in which naive pluripotency is characterized by Arp2/3 complex-dependent remodeling of the actin cytoskeleton and colony mechanics. Additionally, it has been suggested that Arp2/3 activity facilitates naive dedifferentiation by promoting the nuclear translocation of YAP.

Major comments:

The experiments were of high quality, and the paper was clearly written with these major conclusions being supported by the experiments. However, the three-dimensional (3D) organization of F-actin, including the actin ring surrounding naive colonies, is unclear.

1. The authors found that a ring of actin filaments at the colony periphery was characteristic of the naive hESCs. However, because all the data are presented as an image of a single confocal section, the 3D organization of the actin filaments is not clear. Although the authors drew a scheme for this actin ring being located in the apical domain of polarized cells, such data have not been provided in the manuscript. Since naive hESCs form dome-like colonies, it is important to show the 3D organization of actin filaments in the colony. 3D reconstruction of confocal microscopy images of the naive hESC colonies is required to show the relationship between actin filaments, adherens junctions, and the nuclei (as a reference for the Z axis). If 3D reconstruction is not technically possible, confocal images at different Z levels and maximum projection images should be obtained and provided.
2. Some of the statistical analyses were inappropriate. The authors have used Student's t-test for all analyses,; however, one-way ANOVA and post-hoc analysis must be used to compare three or more groups (Figs. 2B, D, E, 3G, 4B, D, E).

Minor comments:

1. Page 9, second paragraph. In the discussion section, authors have written that "Cells within the ICM of mouse blastocysts exclude YAP from the nucleus whereas cells within the ICM of human blastocysts maintain nuclear YAP." However, a recent study has reported that the ICM/epiblast of mouse late blastocysts also express nuclear YAP.

Epiblast Formation by TEAD-YAP-Dependent Expression of Pluripotency Factors and Competitive Elimination of Unspecified Cells. Hashimoto M, Sasaki H. Dev Cell. 2019, 50:139-154.e5. doi: 10.1016/j.devcel.2019.05.024.

Significance

The importance of actin dynamics and cell mechanics as regulators of cell fate transition has been demonstrated in several systems, including the differentiation of hESCs (ref 54). The importance of YAP in the generation of naive hESCs has been reported previously (24). This study further extends this knowledge by showing the importance of actin dynamics and cell mechanics during the naive dedifferentiation process of hESCs. Although the advancement is not significant, the identification of the Arp2/3 complex as an essential upstream regulator of actin dynamics, cell mechanics, and YAP provides novel and important information for the field of stem cell biology, specifically for researchers working on hESC reprogramming and regenerative medicine.

The field of expertise of this reviewer is mouse preimplantation development and Hippo signaling.

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

This manuscript by Meyer et al., shows that the transition from primed to naïve human Embryonic Stem cells is associated with changes with the organization of the actin cytoskeleton, mechanics exerted on the substratum and YAP activity. These changes require Arp2/3 activity and if these changes are blocked with an Arp2/3 inhibitor, the phenotype can be rescued by the expression of a constitutively active YAP form.

This brief manuscript is overall well written and presented. The results are quite original, since branched actin polymerized by Arp2/3 is generally associated with membrane protrusions, and not with contractile actin fibers, as described here. Similarly, YAP activation has been shown to be regulated by RhoA-mediated contractility and here seems to depend on branched actin networks. I have nothing against these provocative conclusions, but I believe that to make their point stronger, more than just the use of an Arp2/3 inhibitor is required.

Major Comments

1. The results of experiments where Arp2/3 is blocked (Fig.2) should be confirmed by Arp2/3 knock-down and with an independent Arp2/3 inhibitor. Several are available (CK-869, Benproperine, Pimozide). For Fig.3 and 4, that would not be necessary, but to establish the specificity of the effect in fig.2 this is absolutely required.
2. I believe that the status of the actin cytoskeleton in both states is not well enough characterized. This is especially obvious for branched actin networks themselves that depend on the Arp2/3. To this end, the authors may localize Arp2/3 or cortactin, a useful surrogate that often gives a better staining. This point is particularly important since contractile fibers are not made of branched actin. Myosin cannot walk or pull along branched actin networks because of steric hindrance. It might well be that branched actin networks are debranched after Arp2/3 polymerization. I suggest staining tropomyosins that would indicate where the transition between branched and unbranched actin would be. Along this line, phosphoERMs should be localized and revealed by Western blots (we expect an increase from primed to naive state) because they cannot perform the proposed function of linker between the membrane and actin filaments if they are not phosphorylated.
3. Branched actin is required for cell cycle progression and cell proliferation in normal cells. This requirement is lost in most cancer cells (Wu et al., Cell 2012; Molinie et al., Cell Res 2019). This would be really important to know whether ESCs stop proliferating upon CK-666 treatment. In other words, do they behave like normal cells or transformed cells. Proliferation is a major function that depends on the YAP pathway. Cell counts and EdU incorporation can easily provide answers to this important question.

Minor comments

1. Fig2F: non-representative pictures or wrong quantification of the CK666 condition.
2. Fig3A: Y-axis ? What is it ? How is it adjusted ? -Log P ?
3. Colors of dots not really visible.
4. What about the rescue of cell morphology ? Does active YAP restore the intercellular contractile bundle ?
5. Typos: Apr2/3 in the abstract, Hoeschst in Fig.S1B.

Significance

This brief report would be a strong report if the major points are addressed. The conclusions are original, with a role of branched actin in inducing an intercellular contractile bundle and activating YAP, and important for a cell system of importance, human ESC. It would interest a wide variety of readers with either an interest in the actin cytoskeleton or in stem cells.

I believe that the time required to address these 3 points is reasonable in the order of 3 months only.

My expertise is the actin cytoskeleton.

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

1. General Statements

We thank the reviewers for their constructive feedback, which has helped us improve the manuscript considerably (no comment on whether the improvements are “significant”). Below are our point-by-point responses. We have also highlighted all changes in the manuscript.

2. Point-by-point description of the revisions

Reviewer 1

Summary

In this study, Obodo et al. present a new iteration of their popular rhythm analysis tool LimoRhyde. The conceptual advancement in this new iteration is the focus on effect sizes (in the form of point estimates of amplitude and their prediction intervals) rather than the p-values, which has been the predominant form of statistical testing for rhythm analysis. Therefore, compared to a well-established non-parametric method for rhythm testing, LimoRhyde2 selects genomic features with larger amplitudes (effect-sizes) as it is designed to do.

Major Comments

1. (LimoRhyde2 algorithm, Page 2-) It is unclear what exactly the contributions/advancements of the authors are? Is it a novel statistical method, the combination of well-established tools in a novel workflow, or is it a novel application to a new field (rhythms)? I am afraid the sentence "LimoRhyde2 builds on previous work by our group and others to rigorously analyze data from genomic experiments [9,16,17], capture non-sinusoidal rhythms [18], and accurately estimate effect sizes [14,19]." is rather ambiguous.

We have revised this sentence in the last paragraph of the Introduction to clarify LimoRhyde2’s contributions.

1. (Moderate model coefficients, Page 3-) The authors implement empirical Bayes shrinkage on the coefficients. But the state-of-the-art methods used in LimoRhyde2 for linear model fitting, such as DESeq2/limma-voom/limma-trend, already implement shrinkage for the coefficients. Does algorithm implement a second round of Bayes shrinkage on the rhythm effect-sizes? How or why is this a statistically valid procedure? If not, how does Limorhyde2 add to shrinkage already implemented in DESeq2/limma-voom/limma-trend? Please elaborate.

To our understanding, the two shrinkage procedures work at different levels and serve different purposes. Limma applies shrinkage on residual variances to account for any technical variation and to give a higher power to detect effects for data with smaller sample sizes within each condition; it does not shrink coefficients. In practice, limma’s shrinkage has little effect given the relatively large sample sizes of most circadian experiments. LimoRhyde2, on the other hand, uses mashr to apply shrinkage to the coefficients themselves to account for shared patterns of effects and variation across both features and conditions. We see no reason this approach is invalid, and in our conversations with Matthew Stephens, the author of ashr and mashr, he felt the same. We elaborate on each method’s contributions in the Discussion (paragraph 2).

1. I think the goal to move to effect-sizes which lead to more reproducible results and better biological significance is sound and highly appreciated. However, to make the community switch to a completely different way of viewing their genomic analysis requires more convincing examples(s)/use-cases on why they should abandon the old method that they are used to. Now, results section merely shows that this algorithm performs as designed (to find large amplitude rhythms).

We appreciate the comment and acknowledge that some readers may be particularly attached to p-values and our current analysis may not wholly convince them of the value of effect sizes. We believe the manuscript stands on its own, however, and are using LimoRhyde2 to guide experiments whose conclusions we hope to describe in future work. Nonetheless, we have revised the Discussion (paragraph 4) to clarify that some known relevant genes highly ranked by LimoRhyde2 were underappreciated by BooteJTK.

1. Related to point 3, others have previously proposed using amplitude (effect-size) thresholds in addition to the p-value cutoffs (Lück & Westermark, 2016, Pelikan et al, 2022), how would the results of Limorhyde2 compare in a fairer contrast where both p-value and amplitude thresholds are implemented? Does the proposed sound method outperform the two-step approach. The authors may perform this analysis on their chosen datasets as well.

Thank you for raising this point. Indeed, one way to view LimoRhyde2 is as a data-driven balancing of raw effect size and p-value. However, the approach of considering both raw amplitude and p-value is uncommon and requires yet another arbitrary cutoff, which complicates any genewise ranking and side-by-side comparison with other methods. Thus, we have decided to not perform this analysis, and instead mention what we see as the advantage of LimoRhyde2 in the Discussion (paragraph 2).

1. I am also not completely convinced of the author's approach to compare their tool against BooteJTK. P-values only show ordering when the alternative hypothesis is true. P-values under the null hypothesis are uniformly distributed in [0,1] so would be meaningless for the purpose of ordering. Without knowing the ground-truth, ordering by p-values is rather risky. I understand the authors' difficulty. But maybe point 4 above yields a better evaluation strategy for LimoRhyde2.

If one accepts that these datasets have a non-zero number of “true” rhythmic genes, which to us seems more than reasonable, then we don’t see this is a large issue. Ranking by (adjusted) p-value is also the standard in differential expression analyses.

1. (OPTIONAL) LimoRhyde2 orders results by the point estimates of the effect-sizes (amplitudes). Is this biologically the most meaningful? Should the effect-size CIs be ordered at all? Maybe we only care about whether the lower limit of the CI is greater than a chosen threshold without any ordering. A discussion of this would be valuable to a user.

We discussed this issue amongst ourselves as well, and ultimately elected for simplicity in ranking by only the point estimate and not the credible interval. We have now mentioned this issue in the penultimate paragraph of the Discussion.

1. (OPTIONAL) If indeed the authors want to move away from p-values, one could argue that most of the insights from p-value analysis are or could be biased. So why compare against ordering by p-values at all in the results?

We are not arguing that results from p-value-based analyses are biased. We seek to show the differences on real data between an analysis based on p-values, the dominant approach in the field, and one based on estimated effect sizes. We believe this has greater potential to promote thoughtful progress than does outright rejection of p-values based on a purely theoretical argument.

Minor Comments

1. In page 3, it is unclear why averaging the three fits is the best thing to do? How bad would the performance be if m = 1 was chosen compared to m=3.

We have elaborated the relevant section of the Methods. For most genes in most datasets, the difference between m=1 and m=3 wasn’t much. However, m=1 tended to go noticeably sideways for some of the most rhythmic genes, depending on the relative locations of timepoints and spline knots, whereas m=3 did not.

1. In page 4, "To account for this uncertainty, LimoRhyde2 constructs..." was difficult to understand and sounded arbitrary. Please explain further.

We have revised this sentence.

1. Lachmann et al. (2021) also use bootstrap confidence intervals rather than p-values to quantify rhythmicity that ought to be mentioned.

We have now cited this paper in the Introduction.

Significance Comments

1. General assessment: The authors present an exciting new way of viewing results of high-throughput data analysis in the context of biological rhythms using a Bayesian-like approach. Previously work has revealed the flaws in focusing on p-values and how focusing of effect-sizes (in this context amplitudes) can yield more robust, reproducible results. Although this promises to also yield more biological meaningful results, it is unclear from this study how this might be.

See reply to Major Comment 3 above.

1. Advance: This study presents the first tool in the context of the rhythm analysis to provide prediction intervals for different rhythm parameters to facilitate a move away from the hypothesis testing framework of p-values. This is a technical advance in the field of rhythm analysis, but it is unclear what insights this could yield.

See reply to Major Comment 6 above.

Reviewer 2

Major Comments

1. The manuscript introduces a new tool to select rhythmic genes and to quantify amplitudes and phases. The authors combine splines, linear regression, Bayes sampling, and Mash. They focus on amplitudes instead p-values as in other packages. The performance and independence of JTK methods are illustrated using selected circadian expression profiles from different mammalian tissues. The paper is clearly written and provides a valuable extension of existing tools. I miss, however, an intuitive explanation of Mash.

Thank you.

1. I agree with their claim that amplitudes are quite important for physiological regulations. However, p-values are also helpful to explore, e.g., transcription factor binding sites. Moreover, amplitudes are taken into account in many studies (see e.g. papers of Naef, Korencic, Westermark, Ananthasubramaniam...). Since JTK or RAIN are non-parametric methods amplitudes are not in focus. The authors should discuss the biological relevance of amplitudes more clearly.

Thanks for raising this point. We are careful to limit our claims to bulk transcriptome data, and have tried to cite the relevant prior work. We have revised the Discussion to clarify what we see as the potential value of amplitudes, as illustrated by our analysis.

1. The selection of the 3 data sets and of specific genes seems reasonable since a range of technologies (microarrays versus RNS-seq), of durations (1 day versus 2 days), and of gene amplitudes are represented. Still the authors should comments their selections of data sets and genes.

We have added justification for our choices.

1. I find also the tissue-dependent phase distributions of clock-controlled genes of interest. However, a comparison with other studies (Zhang, GTEx from Talamanca et al.) and a discussion how amplitude thresholds such as 10%, 25%, 50% affect the phase distributions would be valuable.

Thank you for the suggestion. We initially explored several values of the amplitude threshold for those histograms (Figure S4C) before selecting the top 25%, all led to the same conclusion. We consider this a minor issue and tangential to the main point of the paper, so we have left the figure as is. We invite any interested reader to explore the publicly available results.

Reviewer 3

Summary

The authors developed LimoRhyde2, a method for quantifying rhythmicity in genomic data, and applied it to mouse transcriptome data from liver, lung, and suprachiasmatic nucleus (SCN) tissues. The method uses periodic spline-based linear models and an Empirical Bayes procedure (Mash) to produce posterior fits and rhythm statistics. LimoRhyde2 prioritizes high-amplitude rhythms of various shapes rather than monotonic rhythms with high signal-to-noise ratios, which contrasts with previous methods like BooteJTK. The authors demonstrated the value of LimoRhyde2 in quantifying rhythmicity and highlighted some of its advantages over traditional methods. However, they also acknowledged limitations, such as the inability to compare rhythmicity between conditions and the assumption of fixed rhythms.

Major Comments

1. The key conclusions are convincing, as the authors demonstrated LimoRhyde2's ability to fit non-sinusoidal rhythms and prioritize high-amplitude rhythms over monotonic rhythms with high signal-to-noise ratios. This is shown by the comparison with BooteJTK, a popular method in the field, and by the analysis of real circadian transcriptome data from mouse tissues. However, the authors acknowledged some limitations that could impact the method's broader applicability.

Thank you.

1. Data and methods are presented in a reproducible manner, with detailed descriptions of the periodic spline-based linear models, the use of Mash for moderating raw fits, and the calculation of rhythm statistics. This information is sufficient for other researchers to replicate the study and apply the LimoRhyde2 method to their own datasets. The code is available already.

Thank you.

1. Adequate replication and statistical analysis are provided, with the authors analyzing the same datasets using both LimoRhyde2 and BooteJTK to compare their performance. The use of Spearman correlation to assess the relationship between the adjusted p-values from BooteJTK and the amplitudes from LimoRhyde2 further supports the statistical rigor of the study.

Thank you.

Minor Comments

1. Addressing LimoRhyde2's limitations would help improve the study.

We have extensively addressed the method’s limitations to the best of our knowledge in Discussion paragraphs 6 and 7.

1. Authors could provide more details on how LimoRhyde2 could be applied to single-cell RNA-seq data to improve the presentation. Single-cell quantification over time would be a challenging task, so some insight into this would be appreciated, rather than a brief comment at the end of the paper.

Thank you for your interest in this topic. To do it justice, however, requires its own project and paper, so scRNA-seq is beyond the scope of the current paper.

Significance Comments

1. This study represents a technical advance in the field of genomic analysis of biological rhythms by introducing LimoRhyde2, a method that prioritizes high-amplitude rhythms and directly estimates biological rhythms and their uncertainty. The method's ability to capture non-monotonic rhythms and account for uncertainty makes it a valuable tool for researchers interested in understanding circadian systems and their physiological impact.

1. The work is placed in the context of existing literature, as the authors compare LimoRhyde2 with BooteJTK, a refinement of the popular JTK_CYCLE method. The comparison highlights the differences in output, prioritization, and runtime, demonstrating LimoRhyde2's potential advantages over traditional methods in the field.

2. However, BooteJTK is relatively underused compared to many other methods, partly because of the difficulty and time required to run the analysis. The paper would be improved by comparing LimoRhyde2 to JTK_Cycle itself, as well as RAIN and ARSER. The latter are the most commonly used methods for rhythm detection, and thus the value of the paper's findings would be far greater by comparing to these methods. Like LimoRhyde2, they are also not resource-intensive to run.

Thanks for your feedback on this point, which is one we discussed at length amongst ourselves. In the end, we decided on BooteJTK because it seems to be the best performing version of the most common method. ARSER and RAIN are simply not the standard, and based on our interpretation of the evidence, not generally superior to JTK. If we had selected the vanilla JTK_Cycle, we felt a reviewer could discard our results by saying "well, they're comparing their method to a version of a method known to be flawed". Given our objective to highlight the differences between prioritization based on estimated effect size and prioritization based on p-value, we do not see the value of including additional methods in the analysis.

1. LimoRhyde2's ability to efficiently prioritize large effects with functional significance in the circadian system can provide valuable insights for these researchers and advance the understanding of biological rhythms. The LimoRhyde2 approach is different to conventional reliance on arbitrary p- or q-values, which are taken as almost sacrosanct in the field as a measure of a dataset's worth. LimoRhyde2 could thus help to change this false perception of how to rate a circadian rhythm, which has particularly been ushered in by a reliance on JTK_Cycle p- and q-values as the method of choice for assigning meaningfulness to rhythms. Unfortunately, JTK_Cycle is very conservative and is limited to detecting sinusoidal-type rhythms. LimoRhyde2 could overcome these limitations (as RAIN does too) if widely adopted. However, to do this, it must be compared to things like JTK_Cycle directly.

See reply to Significance Comment 3 above.

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 #3

Evidence, reproducibility and clarity

Summary

The authors developed LimoRhyde2, a method for quantifying rhythmicity in genomic data, and applied it to mouse transcriptome data from liver, lung, and suprachiasmatic nucleus (SCN) tissues. The method uses periodic spline-based linear models and an Empirical Bayes procedure (Mash) to produce posterior fits and rhythm statistics. LimoRhyde2 prioritizes high-amplitude rhythms of various shapes rather than monotonic rhythms with high signal-to-noise ratios, which contrasts with previous methods like BooteJTK. The authors demonstrated the value of LimoRhyde2 in quantifying rhythmicity and highlighted some of its advantages over traditional methods. However, they also acknowledged limitations, such as the inability to compare rhythmicity between conditions and the assumption of fixed rhythms.

Major comments:

1. The key conclusions are convincing, as the authors demonstrated LimoRhyde2's ability to fit non-sinusoidal rhythms and prioritize high-amplitude rhythms over monotonic rhythms with high signal-to-noise ratios. This is shown by the comparison with BooteJTK, a popular method in the field, and by the analysis of real circadian transcriptome data from mouse tissues. However, the authors acknowledged some limitations that could impact the method's broader applicability.
2. Data and methods are presented in a reproducible manner, with detailed descriptions of the periodic spline-based linear models, the use of Mash for moderating raw fits, and the calculation of rhythm statistics. This information is sufficient for other researchers to replicate the study and apply the LimoRhyde2 method to their own datasets. The code is available already.
3. Adequate replication and statistical analysis are provided, with the authors analyzing the same datasets using both LimoRhyde2 and BooteJTK to compare their performance. The use of Spearman correlation to assess the relationship between the adjusted p-values from BooteJTK and the amplitudes from LimoRhyde2 further supports the statistical rigor of the study.

Minor comments:

1. Addressing LimoRhyde2's limitations would help improve the study.
2. Authors could provide more details on how LimoRhyde2 could be applied to single-cell RNA-seq data to improve the presentation. Single-cell quantification over time would be a challenging task, so some insight into this would be appreciated, rather than a brief comment at the end of the paper.

Significance

1. This study represents a technical advance in the field of genomic analysis of biological rhythms by introducing LimoRhyde2, a method that prioritizes high-amplitude rhythms and directly estimates biological rhythms and their uncertainty. The method's ability to capture non-monotonic rhythms and account for uncertainty makes it a valuable tool for researchers interested in understanding circadian systems and their physiological impact.
2. The work is placed in the context of existing literature, as the authors compare LimoRhyde2 with BooteJTK, a refinement of the popular JTK_CYCLE method. The comparison highlights the differences in output, prioritization, and runtime, demonstrating LimoRhyde2's potential advantages over traditional methods in the field.
3. However, BooteJTK is relatively underused compared to many other methods, partly because of the difficulty and time required to run the analysis. The paper would be improved by comparing LimoRhyde2 to JTK_Cycle itself, as well as RAIN and ARSER. The latter are the most commonly used methods for rhythm detection, and thus the value of the paper's findings would be far greater by comparing to these methods. Like LimoRhyde2, they are also not resource-intensive to run.
4. LimoRhyde2's ability to efficiently prioritize large effects with functional significance in the circadian system can provide valuable insights for these researchers and advance the understanding of biological rhythms. The LimoRhyde2 approach is different to conventional reliance on arbitrary p- or q-values, which are taken as almost sacrosanct in the field as a measure of a dataset's worth. LimoRhyde2 could thus help to change this false perception of how to rate a circadian rhythm, which has particularly been ushered in by a reliance on JTK_Cycle p- and q-values as the method of choice for assigning meaningfulness to rhythms. Unfortunately, JTK_Cycle is very conservative and is limited to detecting sinusoidal-type rhythms. LimoRhyde2 could overcome these limitations (as RAIN does too) if widely adopted. However, to do this, it must be compared to things like JTK_Cycle directly.

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

The manuscript introduces a new tool to select rhythmic genes and to quantify amplitudes and phases. The authors combine splines, linear regression, Bayes sampling, and Mash. They focus on amplitudes instead p-values as in other packages. The performance and independence of JTK methods are illustrated using selected circadian expression profiles from different mammalian tissues. The paper is clearly written and provides a valuable extension of existing tools. I miss, however, an intuitive explanation of Mash.

Significance

I agree with their claim that amplitudes are quite important for physiological regulations. However, p-values are also helpful to explore, e.g., transcription factor binding sites. Moreover, amplitudes are taken into account in many studies (see e.g. papers of Naef, Korencic, Westermark, Ananthasubramaniam...). Since JTK or RAIN are non-parametric methods amplitudes are not in focus. The authors should discuss the biological relevance of amplitudes more clearly.

The selection of the 3 data sets and of specific genes seems reasonable since a range of technologies (microarrays versus RNS-seq), of durations (1 day versus 2 days), and of gene amplitudes are represented. Still the authors should comments their selections of data sets and genes.

I find also the tissue-dependent phase distributions of clock-controlled genes of interest. However, a comparison with other studies (Zhang, GTEx from Talamanca et al.) and a discussion how amplitude thresholds such as 10%, 25%, 50% affect the phase distributions would be valuable.

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

Summary:

In this study, Obodo et al. present a new iteration of their popular rhythm analysis tool LimoRhyde. The conceptual advancement in this new iteration is the focus on effect sizes (in the form of point estimates of amplitude and their prediction intervals) rather than the p-values, which has been the predominant form of statistical testing for rhythm analysis. Therefore, compared to a well-established non-parametric method for rhythm testing, LimoRhyde2 selects genomic features with larger amplitudes (effect-sizes) as it is designed to do.

Major Comments:

1. (LimoRhyde2 algorithm, Page 2-) It is unclear what exactly the contributions/advancements of the authors are? Is it a novel statistical method, the combination of well-established tools in a novel workflow, or is it a novel application to a new field (rhythms)? I am afraid the sentence "LimoRhyde2 builds on previous work by our group and others to rigorously analyze data from genomic experiments [9,16,17], capture non-sinusoidal rhythms [18], and accurately estimate effect sizes [14,19]." is rather ambiguous.
2. (Moderate model coefficients, Page 3-) The authors implement empirical Bayes shrinkage on the coefficients. But the state-of-the-art methods used in LimoRhyde2 for linear model fitting, such as DESeq2/limma-voom/limma-trend, already implement shrinkage for the coefficients. Does algorithm implement a second round of Bayes shrinkage on the rhythm effect-sizes? How or why is this a statistically valid procedure? If not, how does Limorhyde2 add to shrinkage already implemented in DESeq2/limma-voom/limma-trend? Please elaborate.
3. I think the goal to move to effect-sizes which lead to more reproducible results and better biological significance is sound and highly appreciated. However, to make the community switch to a completely different way of viewing their genomic analysis requires more convincing examples(s)/use-cases on why they should abandon the old method that they are used to. Now, results section merely shows that this algorithm performs as designed (to find large amplitude rhythms).
4. Related to point 3, others have previously proposed using amplitude (effect-size) thresholds in addition to the p-value cutoffs (Lück & Westermark, 2016, Pelikan et al, 2022), how would the results of Limorhyde2 compare in a fairer contrast where both p-value and amplitude thresholds are implemented? Does the proposed sound method outperform the two-step approach. The authors may perform this analysis on their chosen datasets as well.
5. I am also not completely convinced of the author's approach to compare their tool against BooteJTK. P-values only show ordering when the alternative hypothesis is true. P-values under the null hypothesis are uniformly distributed in [0,1] so would be meaningless for the purpose of ordering. Without knowing the ground-truth, ordering by p-values is rather risky. I understand the authors' difficulty. But maybe point 4 above yields a better evaluation strategy for LimoRhyde2.
6. (OPTIONAL) LimoRhyde2 orders results by the point estimates of the effect-sizes (amplitudes). Is this biologically the most meaningful? Should the effect-size CIs be ordered at all? Maybe we only care about what whether the lower limit of the CI is greater than a chosen threshold without any ordering. A discussion of this would be valuable to a user.
7. (OPTIONAL) If indeed the authors want to move away from p-values, one could argue that most of the insights from p-value analysis are or could be biased. So why compare against ordering by p-values at all in the results?

Minor Comments:

1. In page 3, it is unclear why averaging the three fits is the best thing to do? How bad would the performance be if m = 1 was chosen compared to m=3.
2. In page 4, "To account for this uncertainty, LimoRhyde2 constructs..." was difficult to understand and sounded arbitrary. Please explain further.
3. Lachmann et al. (2021) also use bootstrap confidence intervals rather than p-values to quantify rhythmicity that ought to be mentioned.

Significance

General assessment:

The authors present an exciting new way of viewing results of high-throughput data analysis in the context of biological rhythms using a Bayesian-like approach. Previously work has revealed the flaws in focusing on p-values and how focusing of effect-sizes (in this context amplitudes) can yield more robust, reproducible results. Although this promises to also yield more biological meaningful results, it is unclear from this study how this might be.

Advance:

This study presents the first tool in the context of the rhythm analysis to provide prediction intervals for different rhythm parameters to facilitate a move away from the hypothesis testing framework of p-values. This is a technical advance in the field of rhythm analysis, but it is unclear what insights this could yield.

Audience:

This will be useful to all chronobiologists (clinical and basic research) who use high-throughput genomic assays. Since this is an open R-package, I suspect most of those who want to will be able to easily use it. My expertise is in chronobiology, data science and systems biology.

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

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 report, Bilgic and colleagues study the diversity of progenitor cell types in the developing ferret cerebral cortex, a valuable in vivo model to understand cortex expansion and folding, as in primates including human. Using a single-cell transcriptomics approach, they describe a diversity of progenitor cell types and their interrelation by transcriptomic trajectories, which are conserved but biased as development progresses. Most interestingly, they identify in ferret a type of cell only identified in human before, tRG, which they then characterize throughoutly by transcriptomics. They also identify these cells in histological sections, and via time-lapse videomicroscopy they characterize their cell type of origin. They also provide indirect evidence that tRG may be the source of ependymal cells in the ventricle of the mature cerebral cortex, as well as astroglial progenitor cells. Finally, they extend their analyses to identify oRG in ferret based on previous human single cell data, concluding that they have in ferret a quite different transcriptomic profile than in human.

Major issues:

The authors must provide evidence that the cortical area they are examining will give rise to Somatosensory cortex. Their sampling area appears more like Cingulate cortex, while somatosensory may be a bit more lateral. The cingulate cortex is a very unique region, with some unique characteristics including lamination and connectivity. It would be important to provide some justification as to why they chose this particular part of the cerebral cortex, and keep this into consideration when discussing the general value of their findings.

It seems that the single cell datasets were collected from only 1 replica at each developmental stage. Current best practice sets the inclusion of several biological replicates. Whereas this represents multiplying the workload (and costs) and re-doing many of the analyses, it is currently highly valued. On the other hand, the authors already have their analysis pipelines defined, and so the time involved should be much shorter than before.

Single cell QC methods are incomplete as described in Methods. It is key to consider the relative abundance of mitochondrial RNAs when assessing the integrity and validity of cells, and thus a key criterion to select the cells for clustering analysis. The criteria for the selected choice of clustering resolution is also missing.

When first describing tRGs (line 171), orthogonal views of the image z-stacks must be shown to demonstrate the full morphology of these cells. The basal process might have been cut during tissue sectioning. The same applies to images in Fig. 2C, 2D, S3A.

In Figure 3, the authors perform time-lapse imaging to visualize and characterize the cells and lineage that give rise to tRGs. While very nice and a technical challenge that must be properly acknowledged, they unfortunately only obtained a total of three examples, which is clearly insufficient to reach any meaningful conclusion on this respect. These conclusions, while fascinating, are based only on 3 cell divisions. If this is to be taken as a strong argument for the conclusions of the study, the authors must obtain. If the authors want to make a solid statement out of this experimental approach, they must obtain a sufficient amount of additional data, which will depend on the variability of the results they find.

Still regarding the time-lapse results presented in Figure 3, it is unclear why after first division the authors identify the blue cell as IPC, when it has the exact features of tRG: apical process anchored in VZ surface + short basal process. This is applicable to all three examples shown. For example, the authors describe: "the mother IPC of tRG also possessed both an apical endfoot and a short basal fiber (Fig. 3D)". Why is this identified as IPC, when it looks exactly like vRG, NOT as an IPC? The interpretation of IPCs being the mother cells to tRGs must be changed, to those being vRGs. Or else, more convincing data must be provided. In fact, their analyses in Fig 4A contradict their interpretation on tRG mother cells, showing that the transcriptomic trajectory leading to tRGs does not inlcude Eomes+ cells, accumulated in the neurogenic state 2. At the end of this section, the authors indicate: "our data suggest that tRG cells are formed by apical asymmetric division(s) from unique apical IPC with a short basal fiber (Tsunekawa et al, in preparation).". Being as important as this point is, if there is solid supporting data the authors must include it in this study.

Alternative interpretation of time-lapse images (lines 196-197): maybe a tRG can generate one tRG CRYAB+, and one IPC CRYAB-.

Arrows in Fig 5E are shifted between the top and bottom panels. There is no obvious evidence of mitosis visible. This should be unequivocally labeled with anti-PH3 antibodies.

Line 277: "Transcriptomic trajectories were homologous across the two species". What does this refer to? What are these trajectories? Pseudotime? Is this statistically tested?

When comparing tRG cells between ferret and human, the authors indicate a remarkable similarity between the two species as represented by CRYAB, EGR1, and CYR61 expression. As shown in Fig 6E, EGR1 and CYR61 are not expressed selectively in human tRG as they clearly are in ferret tRG. Hence, this argument is not valid.

In the last part, the authors try to identify oRG-like cells in ferret by comparison with their transcriptomes identified in human. For this, they decide to call ferret oRG-like cells those that are near human oRGs in the integrated UMAP, as identified in a previous human study. What was the criterion for this? How much near is "near"? The fact that the selected cells have higher oRG scores is expected and obvious, as these cells were selected precisely based on their proximity in the UMAP. Even more importantly, the identification of oRGs in the human study is not unambiguous. Therefore, the correlate in ferret cells is also non-conclusive as to the identity of such cells.

Discussion is surprisingly short, given the emphasis that the authors place on the importance of their findings. I would suggest extending it for a better coverage of those findings that have the greatest relevance and interest to a wider readership.

In Discussion, the authors state that "ferret (and presumably also human) tRG cells differentiate into ependymal cells and astrogenic cells." Again, this conclusion is purely based on transcriptomic trajectories, which must not be confused with cell lineage. This sentence must be rephrased and toned down accordingly.

In Discussion: "our cross-species analysis highlights the notable role of tRG as progenitors contributing to the formation of the ependyma and white matter". As mentioned above, this is only based on transcriptomic trajectories, it is not demonstrated in this study. In vivo analyses of cell fate are needed to support this conclusion, and a more extensive videomicroscopy analysis is needed to confirm the cell lineage progression suggested by transcriptomes.

Minor issues:

In line 59, the authors state: "cerebral carcinogenesis independently evolved to gain an additional germinal layer (outer SVZ (OSVZ);". Assuming that they mean "cerebral neurogenesis", what is the evidence for this independent evolution? Original publications demonstrating this must be cited.

Lines 60-61, the third key publication reporting the existence of bRG must be cited together with Hansen 2010 and Fietz 2010: Reillo et al., 2011, Cerebral Cortex.

When introducing ferret as an interesting or important animal model, suitable original studies should be cited.

In Figure 2F-H, layer borders should be labeled. The density of CRYAB+ cells in VZ (?) at P5 seems much greater in Fig 2E,F than in Fig. 2B. Clarifying this discrepancy is important to validate the quantification of Fig 2D.

Co-expression of CRYAB and FOXJ1. In Fig 5B this must be demonstrated with merged channels.

Line 247: "near which nuclear line aggregates are observed more frequently (Fig. 2B)". It is very much unclear what the authors refer to. Please, define nuclear line aggregates.

There are a number of typos along the main text and figures, which must be fully checked and corrected. For example, line 59 "cerebral carcinogenesis"; also in Figure S4, Figure 5E. Labeling of graphs in Fig 5C is wrong. The plots present the fraction of CRYAB+ cells that express FOXJ1 (FOXJ1+/CRYAB+ cells), not the reverse.

Significance

This manuscript is of interest for being the first ferret single-cell study, and for identifying and characterizing to a great extent a unique population of cortical progenitor cells that so far had only been observed in human. The study is presented as a resource for studies of ferret cortex development, which as such is clearly of interest to a very limited audience. A more appealing perspective might be if this study in ferret is of interest or of use to the more general community studying cortex development, or even maybe cortex evolution.

Advance: it is, so far, the first study of single cell profiling of the ferret cerebral cortex, a well established and highly valued model of gyrencephalic mammals, and a suitable best-alternative to work in primates. In addition to the technical advance, providing a new resource for work in ferret, it shows for the first time the existence of truncated Radial Glia (tRG) in a non-human cortex, and even more importantly in this model, strengthening even more its value.

This study as is presented will be of most interest to a specialized audience, those directly working with ferret. Nevertheless, it will also be of conceptual interest to the community of cortex development and evolution for the concepts that one can extract on cell type conservation.

My expertise: cerebral cortex development, brain evolution, ferret, cortex folding, neurogenesis, progenitor cell lineage, transcriptomics of developing brain

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

Summary: Provide a short summary of the findings and key conclusions (including methodology and model system(s) where appropriate).

In this manuscript, the authors conduct a series of single-cell transcriptomic analyses and imaging assays in the developing ferret cortex suggesting that (1) ferrets harbor a radial glia (RG) subtype similar to the truncated radial glia (tRG) described previously in humans that may have the potential to (2) produce ependymal and astrogenic lineages which (3) can also be found in the developing human cortex. These findings appear to be an important step in the validation and development of the ferret model towards a tool that can be used to study tRG cell biology, a feat currently difficult due to the inaccessibility of a genetically tractable source of tRG for molecular and cell biology experiments.

Major comments:

• Are the key conclusions convincing?

I found the key conclusions described above and in the authors' abstract convincing. I found the identification of a distinct, tRG-like cell type from the authors' single-cell transcriptomic analysis of the ferret cortex compelling, particularly because (1) the expression of the previously utilized tRG marker gene CRYAB is specific to the tRG-like cluster and (2) the tRG-like cluster marker genes (including CRYAB) are relatively unique to the tRG-like cluster. I found this strengthened by their morphological analyses showing the tRG-characteristic apical endfoot and short basal process in these CRYAB+ cells in the ferret cortex. I found the combination of imaging and bioinformatic analyses showing the increase in FOXJ1 co-expression in CRYAB+ cells to compellingly suggest that CRYAB+ cells can produce FOXJ1+ ependymal cells, and similarly with the authors' analyses to suggest that tRG-like cells can also contribute to SPARCL1+ astrocyte cells. I found that the cluster score analyses compelling suggest that the tRG-like cells in the ferret dataset correlate with the tRG cells annotated in a separate, human developing cortical dataset. I also appreciated the comparison of astroglial, ependymal, and uncommited ferret tRG sub populations from the pseudo time analysis with the clusters generated from the integrated ferret-human dataset in Fig. 7.<br /> - Should the authors qualify some of their claims as preliminary or speculative, or remove them altogether? - Would additional experiments be essential to support the claims of the paper? Request additional experiments only where necessary for the paper as it is, and do not ask authors to open new lines of experimentation. - Are the suggested experiments realistic in terms of time and resources? It would help if you could add an estimated cost and time investment for substantial experiments.

The weakest claim in the paper is lines 202: "...tRG cells are formed by apical asymmetric division(s) from unique apical IPC". From my understanding, the main evidence that the tRG parent cells shown in Fig. 3 are not tRGs are the data from Fig. 2E-G showing the low amounts of CRYAB+ cells co-expressing KI67, TBR2, or OLIG2 in P5 and P10. Especially given that these timepoints are after those used in Fig. 3, I believe further evidence is needed to confirm the cell type identity of tRG parent cells in Fig. 3. Such experiments (isolating IPCs from ferret cortex and growing in vitro to determine progeny cells) may be outside of the scope of this paper, in which case I believe the text can be strengthened with either (1) presenting the data from the cited Tsunekawa et al, in preparation that would suggest this claim or (2) rephrasing these claims to omit the mention of IPCs.

I also believe the claim in Line 365-366 is overstated: "We found that ferret (and presumably also human) tRG cells differentiate into ependymal cells and astrogenic cells." While I believe the transcriptomic comparisons suggest the presence of uncommitted tRG in both the ferret and human datasets, I would appreciate further analyses to confirm the prevalence of astroglial and ependymal tRG in the humans and/or functional analyses before claiming that human tRG cells make ependymal and astrogenic cells. I appreciate the authors' note that "GW25 is...the latest stages experimentally available" (line 376-377), but their comparative approaches could be applied to existing datasets of the human cortex (Herring et al., 2022, PMID: 36318921) that span later developmental ages. Identifying the presence of astroglial and ependymal tRGs in this and/or similar datasets would provide more convincing evidence of the tRGs' developmental potential. If this computational analysis is outside the scope of the paper, I believe paring the certainty of these claims (especially lines 379 - 383) and recognizing the need for further functional analyses would negate the need for deeper mechanistic validation.

I believe the most significant advance for this paper is the potential to use ferret tRG cells to model those of the human brain. However to support this claim (see Lines 83-84), I believe a comparison of the ferret tRG cells with existing cortical organoid datasets (Bhaduri et al., 2020, PMID: 31996853) would be helpful. If cortical organoids currently lack the presence of tRG cell types, that would strengthen the importance of the ferret model and the findings of this paper - otherwise, I feel that the use of the ferret model needs to be justified in light of the greater accesibility and genetic tractability of the cortical organoid system. - Are the data and the methods presented in such a way that they can be reproduced? Yes - Are the experiments adequately replicated and statistical analysis adequate?

I found the total number of tRG-like cells in the ferret dataset quite small (162), but I understand the difficulty with isolating and sequencing rare cell types from primary tissue sources. I believe most of the transcriptomic analyses were conducted with this low n in consideration, but this caveat is even more reason to pare down the wording for the weaker claims mentioned above. - Are prior studies referenced appropriately?

I found it interesting that tRGs persist and even expand in number in postnatal timepoints (Fig. 2C). I'd be interested to know if this is in line with what is known in human developing cortex. If so, it would strengthen the conclusion that ferret tRGs can model that of humans - and if not, this would either be an important finding regarding tRG function or an important caveat in the use of ferret tRGs to model the cell type in humans. - Are the text and figures clear and accurate? Yes - Do you have suggestions that would help the authors improve the presentation of their data and conclusions?

For Fig. 2A, I would find it helpful to compare the morphology of GFP+/CRYAB+ cells vs GFP+/CRYAB- cells, with the hypothesis that GFP+/CRYAB- cells will have elongated basal processes. I believe this could be done by finding GFP+/CRYAB- cells in the raw images obtained to generate Fig. 2A (or similar), and showing those cells in an adjacent panel. This side-by-side comparison could provide more support that the CRYAB+ cells from the single-cell analyses are indeed specifically linked to tRG-like morphology.

Significance

This paper primarily presents a technical advance in the field, showing that tRG cells that can model those found in the developing human cortex are found in the developing ferret cortex.

• Place the work in the context of the existing literature (provide references, where appropriate).
• State what audience might be interested in and influenced by the reported findings. Several studies in the human and macaque brain have identified the presence of tRGs (deAzevedo et al., 2003; Nowakowski et al., 2016), but understanding the molecular functions and development of these cells - and many human-specific cell types in the brain - is difficult due to the lack of tractable models of human neurodevelopment. Ferrets, given their layered cortices, may be a potential model system for these cell types, but further analyses to determine their transcriptomic similarity to the developing human cortex and their ability to recapitulate human cell types are required in order to evaluate their use as a model system. By generating a useful resource in the ferret single-cell transcriptomic atlas, this study provides evidence that - at least for the tRG subtypes - ferrets may be useful in dissecting the generation and functional importance of tRG cells. With the caveat that a direct comparison with the use of cortical organoids to study tRG is lacking in this paper (see above), I believe this work can provide useful insight into the field's current search for model systems to functionally interrogate human-specific aspects of cortical development.
• Define your field of expertise with a few keywords to help the authors contextualize your point of view. Indicate if there are any parts of the paper that you do not have sufficient expertise to evaluate.

Single-cell transcriptomic profiling of primary developing human cortex and cortical organoids

Did not have sufficient expertise in: - Ferret biology

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

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

Manuscript number: RC-2023-01910

Corresponding author(s): Michael W. Sereda

1. General Statements

Reviewer #1:

In this paper the authors report a direct correlation between PMP22 and PTEN expression levels in the nerve of CMT mutants. In CMT1A Pmp22tg rat nerves, PTEN levels are increased, whereas in Pmp22+/- mutants, a model of the HNPP neuropathy, PTEN levels decrease. Consistent with this, Pmp22tg nerves display lower Akt phosphorylation and, vice versa, Pmp22+/- nerves have higher Akt phosphorylation. The authors lowered PTEN in the transgenic and inhibited mTOR using Rapamycin in the Pmp22+/- to support the functional relevance of the PMP22-PTEN correlation. ... In conclusion, the correlation between PMP22 and PTEN is a potential interesting observation. However, in my opinion, experiments as shown don't support the conclusion that PMP22 controls PTEN expression level and activity, which is suggested at the basis of the pathogenesis of PMP22 dosage-related neuropathies.

We thank Reviewer #1 for this detailed feedback. We appreciate the Reviewer’s assessment that our observation that PMP22 and PTEN are correlated in CMT1A and HNPP is of potential interest. In the revised manuscript we addressed this key point by adding additional quantifications (Figure 1a, d; Figure 5d) and novel Western Blot analyses (Figure 1a, d). Regarding the pathophysiological significance of the correlation, we point out that both the original as well as the partially revised manuscript contain multiple pieces of evidence demonstrating that altered PTEN activity is critical for both PMP22 gene-dosage related neuropathies:

1. The inhibition of the PI3K/PTEN/AKT/mTOR axis upstream (LY294002) or downstream (Rapamycin) of decreased PTEN ameliorates myelin defects in an in vitro HNPP model (Figure 2b, c).
2. Downstream of PTEN, Rapamycin treatment ameliorates myelin defects, motor behavior and electrophysiology in the HNPP mouse model in vivo (Figure 3c, d, e,____ g, i)
3. Targeting of increased PTEN directly by inhibiting its activity pharmacologically (VO-OHpic) in a CMT1A rat model or by depleting it genetically in a CMT1A model leads to ameliorated myelination in vitro (Figure 4b, c; Figure 5f, g).
4. The genetic depletion of PTEN in a CMT1A mouse model increases myelination in vivo, albeit not in the long term (Figure 6a, b, c, d). We therefore feel that any additional evidence to show that "PMP22 controls PTEN activity" is not vital for supporting the major claims of the manuscript, i.e. that the observed correlation of PTEN levels with PMP22 gene dosage has relevance for the etiology of PMP22 dosage diseases and and that targeting the PI3K-PTEN-AKT-mTOR axis downstream of PTEN provides a potential pharmacological therapy of HNPP (while directly targeting PTEN ultimately fails to rescue CMT1A). However, we agree that the activity of PTEN on the molecular level is interesting, and such evidence would further strengthen our conclusions. Therefore, in the final revised version, we plan to add further Western Blots and explore possible downstream effects of altered PTEN levels.

Reviewer #2:

This study investigates the modulation, both genetically and pharmacologically, of the PI3K/Akt/mTOR signaling in preclinical animal models for the inherited peripheral neuropathies HNPP and CMT1A. These conditions result from a gene dosage abnormality of the peripheral myelin protein gene PMP22. The exact biological molecular mechanisms remain enigmatic despite it having been over 30 years since the major genetic lesions, the CMT1A duplication and HNPP deletion, were described. With respect to myelin biology one observes focally slowed nerve conduction at pressure palsies and local/segmental hypermyelination in HNPP whereas hypomyelination occurs in CMT1A. The study is nicely conducted, data illustrations very informative, and writing clear and concise. This paper will likely be of great interest to your readers. The authors provide convincing evidence that the HNPP pathobiology is ameliorated by PI3K/Akt/mTOR inhibitors. Interestingly they found radial myelin growth was most affected by this approach and suggest an interesting transdermal approach in injured nerves in the acute prevention of pressure palsies.

We thank Reviewer #2 for this positive evaluation.

Reviewer #3:

*In this paper Sareda and co-workers demonstrate that the PTEN/mTOR pathway is indirectly involved in regulating myelin thickness and wrapping in models of altered PMP22 gene dosage both in vitro and in vivo. Inhibition of this pathway decreases myelin thickness in models of HNPP, while increasing myelin thickness in models of CMT1A. The evidence for these conclusions is complex but reasonably presented, and the conclusions mainly supported by the data. The abstract for this paper, however, presents a somewhat oversimplified conclusion that the PTEN pathway mainly modifies models of HNPP, where the paper clearly demonstrates that models of CMT1A are also affected by this same pathway. This should be clarified. *

We thank Reviewer #3 for the feedback on the manuscript. We agree with the Reviewer that the same pathway (PI3K/Akt/mTOR) also affects CMT1A, but it is of importance for us to highlight that the disease mechanisms are -at least partly- different between HNPP and CMT1A. This is supported by our observation that PTEN reduction in CMT1A only transiently improves myelination in vivo (Figure 6) and the persistent alteration of differentiation markers despite PTEN reduction, which is not observed in HNPP (Figure 7).

2. Description of the planned revisions

Reviewer #1

Regarding the activity of PTEN

Figure 1

• Additional experiments are needed to support the conclusion of Figure 1 that, in the two mutants, Pten levels reversely correlate with PI3K-Akt-mTOR pathway activation, which represents the rationale of all further experiments. For example, it should be shown systematically in both mutants both Akt and ERK phosphorylation levels (Akt at both T308 and S473), and mTOR activity read outs. In the previously published paper (Fledrich et al.) only increased Akt phosphorylation in Pmp22+/- nerves was reported, whereas Pmp22tg analysis was focused on the interdependence between Akt and ERK without exploring mTOR activation, which is relevant here. 2) (Figure 4) A different model, the C61 mouse a Pmp22tg overexpressing PMP22 is used here (rather than the CMT1A rat). This should be explained in the results. Is also this model characterized by increased Pten levels in the nerve? And low Akt-mTOR activation for instance? 3) (Figure 5) How is Akt-mTOR signaling in the double mutant as compared to Pmp22tg? Is that increased at P18? * Response: We fully agree with the Reviewer that further exploration of PTEN downstream effects will add value to the manuscript. We already justified the usage of the C61 mouse model more clearly, added P-S6 staining of wildtype in addition to an improved representation in Figure 5e, and performed extra Western Blot analysis of PTEN expression (described in the next section “Incorporated *revisions”). Moreover, we will further evaluate the downstream signaling components of PTEN and will perform additional Western Blot analyses of peripheral nerves of HNPP mice, CMT1A rats as well as C61 and C61xPTENhKO mice.

Figure S1

• *Figure S1, page 4: what does it mean "in line with this finding we were unable to detect protein-protein...". May be the authors meant: since there is a direct correlation between Pmp22 and Pten expression levels in the mutants, the authors explored the possibility of an interaction between the two. Regarding the co-IPs, in panel a, the co-IP at the endogenous level, the immunoprecipitation efficiency of PMP22 is very low. May be a pull-down experiment using either exogenous purified PMP22 or PTEN and nerve lysates can help to rule out the possibility of an interaction. The experiments in b, c are performed in overexpression in a heterologous system (293 cells). * Response: We agree with the Reviewer that we might have missed a possible interaction between PMP22 and PTEN in the experiments performed so far. Indeed, pull-down experiments may prove helpful to rule out / reveal protein-protein interaction. Therefore, we will use purified PMP22 and perform pull-down experiments using nerve lysates of wildtype and CMT1A rats.

Figure 5

• *Pten Fl/+ Dhh-Cre cultures seem to have axonal fasciculation. * Response____: We thank the Reviewer for this observation. We will systematically inspected all recorded images for features of fasciculation. We will also assess whether fasciculation is a representative feature in cultures derived from any of the genotypes, and if so, whether the genotypes differ in this regard.

3. Description of the revisions that have already been incorporated in the transferred manuscript

Changes in the text are highlighted in green in the revised manuscript

Reviewer #1:

Figure 1

• *Panel a: the decrease of Pten expression should be quantified with at least n=3 taking into account the variability among different samples at the different time points indicated (the same applies in panel b, even if here the increase of Pten expression level in Pmp22tg nerves is more evident). * Response: We agree with the Reviewer that the timeline is not sufficient to demonstrate alteration in PTEN expression in PMP22 gene dosage diseases CMT1A and HNPP. Therefore, we performed new Western Blot experiments evaluating PTEN expression in (i) HNPP mice, (ii) CMT1A rat (iii) C61 mice and (iv) C61xPTENhKO mice with minimum n = 3 biological replicates and performed the respective quantification which is shown in Figure1 (i, ii) and Figure 5 (iii, iv). The results of the Western Blot analysis and quantification show an increase in PTEN abundance in CMT1A rat (Figure 1d) and C61 mice (Figure 5d) while a decrease is observed in HNPP mice (Figure 1a) and PTENhKOxC61 mice (Figure 5d) when compared to wildtype controls.

• *Panel a and b: the statement that Pten is more expressed at P18 at the peak of myelination in wildtype nerves is not supported by the blots as shown. * Response: We agree that this observation is only partly supported by the Western Blot analysis, as seen in the HNPP mouse model, and deleted this part in the results section.

• Figure S1, page 4: what does it mean "in line with this finding we were unable to detect protein-protein...". May be the authors meant: since there is a direct correlation between Pmp22 and Pten expression levels in the mutants, the authors explored the possibility of an interaction between the two. Response: We thank the Reviewer for pointing out the lack of clarity here. We changed the respective sentence accordingly:

“Since there is a direct correlation between PMP22 and PTEN expression levels in the mutants, we explored the possibility of an interaction between the proteins. By immunoprecipitation experiments we were unable to detect protein-protein interaction between PMP22 and PTEN (Figure S1).” (Page 4)

• *Page 4: "Taken together, Pmp22 dosage inversely correlates with the abundance of PTEN...": please revise this statement * Response: We thank the reviewer for spotting this mistake. We changed the sentence accordingly, which now reads:

“Taken together, Pmp22 dosage directly correlates with the abundance of PTEN and presumably the activation level of the PI3K/Akt/mTOR pathway in myelinating Schwann cells (Figure 1i)." (Page 4, Line 23)

Figure 2:

• The aberrant myelin figures displayed are similar to myelin ovoids preceding degeneration rather than myelin outfoldings. It is also strange that these alterations are in the wildtype cultures treated with RAPA, that instead, in this system, has been reported to increase myelination as it improves protein homeostasis (autophagy, quality control, etc). Response: We thank the Reviewer for pointing this out. Indeed, in the way the images have been presented the aberrant myelin profiles can be mistaken for ovoids. However, a close inspection of the TUJ1 channel images revealed continuity of the axons below the aberrant myelin, thereby excluding ovoid formation. In the partially revised manuscript, we now also show the TUJ1 channel individually (Figure 2), so that it can be appreciated that the defects are confined to the myelin. Concerning the incidence of the myelin defects in RAPA treated wildtype cultures, our analysis can have missed a potential amelioration due to the rather high variability in the data.

Figure 3

*Panel c-e: aberrant fibers should be normalized on total number of fibers and on the area, particularly because RAPA is used. *

Response: We agree with the Reviewer that number of tomacula and recurrent loops should be normalized to the total number of fibers on the area. We have quantified the total number of fibers in the whole sciatic nerve and normalized the tomacula and recurrent loops number accordingly. Results show a decrease in both tomacula and recurrent loops after Rapamycin treatment in the HNPP mice (Figure 3c, d, e, f).

Figure 4

The improvement in the number of myelin segments following PTEN inhibition in Pmp22tg co-cultures is very weak. The 500 nM has instead a consistent effect in reducing myelin segments in the wildtype and I think that these results overall don't support the conclusion that myelination is ameliorated by reducing PTEN activity in Pmp22tg co-cultures.

Response: We thank the Reviewer for this important point. We like to emphasize that we treated whole cultures with the PTEN inhibitor and we cannot rule out a (probably) negative effect on axonal PTEN, resulting in only weak improvement of myelination in PMP22tg cultures and strong effects also on the wildtype co-cultures. Therefore, we decided against a treatment of CMT1A models in vivo and further explored the effects of PTEN reduction specifically in Schwann cells using the genetic model as described Figure 5. The Reviewer made clear to us that this is inappropriately explained in the results section and we therefore adapted this in the manuscript on page 6:

“Similarly, the prolonged inhibition of PTEN with VO-OHpic (for 14 days) caused a dosage-dependent reduction in myelinated segments in wildtype co-cultures (Figure 4c, Figure S2). The mechanism is currently unexplained but cannot rule out a negative effect of PTEN inhibition on DRG neurons and myelination.”

Figure 5:

• *A different model, the C61 mouse a Pmp22tg overexpressing PMP22 is used here (rather than the CMT1A rat). This should be explained in the results. Is also this model characterized by increased Pten levels in the nerve? And low Akt-mTOR activation for instance? * Response: We agree with the Reviewer that it has not been clear in the text why we changed here to the C61 mouse model. We clarified this in the Results section which now reads on page 6:

“To reduce Pten function in CMT1A models also in vivo, we applied a genetic approach (Figure 5a). As the genetic tools to specifically target Schwann cells were only available in the mouse and not the rat, we used the C61 mouse model of CMT1A. We reduced PTEN by about 50% selectively in CMT1A Schwann cells by crossbreeding Pmp22 transgenic mice with floxed Pten and Dhh-cre mice, yielding PTENfl/+Dhhcre/+PMP22tg experimental mutants (Figure 5b). Western blot analyses of sciatic nerve lysates confirmed the increase of PTEN in PMP22tg mice and the reduction of PTEN in the double mutants (Figure 5c, d).”

Moreover, regarding the PTEN expression we added Western Blot analysis and quantification in Figure 5c, d showing increased PTEN expression in the C61 mouse model of CMT1A and decreased PTEN in the PTENhKOxC61 double mutants. Further analysis of the downstream signaling is planned (see “planned revision”).

• *PTEN, Akt-mTOR expression/activation levels should be checked biochemically also in this model. And quantified (panel c). * Response: We added an explanation for the use of the C61 mouse model (see point Figure 5.1 above). Moreover, we quantified the Western Blot analysis and added it in Figure 5d. The expression of PTEN was included in the Western Blot analysis (Figure 5c) showing increased PTEN expression also in the C61 mouse model. Further biochemical analysis of the C61 mouse model is planned (see “planned revision”).

• *In panel d overactivation of mTOR (PS6 staining) in Schwann cells is not evident. * Response: We agree with the Reviewer that the way the image was displayed is not sufficient to show P-S6 activation in the double mutants. We have now split the image (Figure 5e) to better visualize the P-S6 staining alone compared to the co-staining with P0 (marker for compact myelin) and DAPI (nuclei). Further, we added staining of wildtype nerve. We hope this way the differences in P-S6 activation can be easier appreciated.

Figure 6:

*G-ratio analysis: which are the mean values (numbers) with SEM in the three groups analyzed wildtype, Pmp22tg and Pmp22tg; Pten fl/+; Dhh-Cre? *

Response: We thank the Reviewer for pointing this out. We added the quantification of the mean g-ratios in Figure 6d, f.

Figure 7:

• *If more fibers are committed to myelinate in the double mutant as compared to the single Pmp22tg at P18 ,particularly, it is unclear why there is no difference in differentiation marker expression in Figure 7 (Oct6 and Hmgcr). * Response: We thank the reviewer for this comment. We do not necessarily expect to see a strong difference in the expression of differentiation markers given the mild increase in myelination in the double mutants. Similarly, we do not observe alterations in the expression of differentiation markers in HNPP mice, while these fibers produce more myelin. Therefore, we concluded that alterations in PTEN-PI3K/Akt/mTOR signaling do not influence differentiation in the mouse models while in the PMP22 overexpressing situation of CMT1A other mechanisms alter differentiation of the Schwann cells. We also note that experiments were performed at postnatal day 18 and we cannot rule out possible alterations in differentiation marker expression at earlier time points in development in the double mutants.

• In conclusion, the correlation between PMP22 and PTEN is a potential interesting observation. However, in my opinion, experiments as shown don't support the conclusion that PMP22 controls PTEN expression level and activity, which is suggested at the basis of the pathogenesis of PMP22 dosage-related neuropathies. Response: Please also see section 1. In order to avoid any overstatement that "PMP22 controls PTEN expression level and activity", in our revised version we have clarified this point and changed the wording in the main text:

"The mechanisms that link the abundance of PMP22 to that of PTEN are still unclear and we here neither show direct nor indirect control of PTEN expression by PMP22." (Page 8)

Reviewer #2:

1. Regarding in the Introduction: "...the molecular mechanisms causative for the abnormal myelination remain largely unknown and still no therapy is available." Suggest consider modifying to perhaps: '...no small molecule or pharmacological therapeutic intervention exist.' To say "no therapy" exist is 'myopic' and untrue.

*Suggest adding question mark to end of sentence or changing ‘asked’ to “investigated” for following thought: “Here, we asked whether PI3K/Akt/mTOR signaling provides therefore a therapeutic target to treat the consequences of altered Pmp22 gene-dosage.” *

Rather than attempt to establish PRIORITY perhaps ‘softening’ the INTRODUCTION concluding statement “Our results thus identify a potential pharmacological target for this inherited neuropathy.

[This makes thePI3K/Akt/mTOR pathway a promising target for a preventive treatment of affected nerves also in human patients.] *Does this belong in RESULTS? Or rather DISCUSSION? *

Response: We thank the Reviewer for the suggestions. We changed the sentences accordingly in the manuscript (1.: Page 3, Line 23; 2.: Page 3, Line 26; highlighted in green). Regarding point 3, we are convinced that identifying pharmacological targets for peripheral neuropathies should be given priority. Indeed, the aspect concerning point 4 is already highlighted in the discussion therefore we removed the sentence from the result section.

Reviewer #3:

*The abstract for this paper, however, presents a somewhat oversimplified conclusion that the PTEN pathway mainly modifies models of HNPP, where the paper clearly demonstrates that models of CMT1A are also affected by this same pathway. This should be clarified. *

We agree with the Reviewer that the same pathway (PI3K/Akt/mTOR) also affects CMT1A, but it is of importance for us to highlight that the disease mechanisms are -at least partly- different between HNPP and CMT1A. This is supported by our observation that PTEN reduction in CMT1A only transiently improves myelination in vivo (Figure 6) and the persistent alteration of differentiation markers despite PTEN reduction, which is not observed in HNPP (Figure 7). For clarification we have altered the wording in the abstract which now reads: "In contrast, we found that CMT1A pathogenesis was only transiently ameliorated by altered PI3K/Akt/mTOR signaling, which drives radial but not longitudinal growth of peripheral myelin sheaths".

3. Description of analyses that authors prefer not to carry out

Reviewer #1:

Figure 1:

*Figure 1, Panel e: may be with this experiment the authors aim to suggest that Pten and Pmp22 are unlikely to interact directly or indirectly since Pten is cytosolic and Pmp22 myelin-membrane enriched. However, this myelin purification shows that Pmp22 as P0 expression levels are also abundant in the cytosol, may be also because P18 has been chosen as time point. What about a different type of membrane-cytosol fractionation experiment and/or another time point? *

Response: We want to clarify that in this experiment not myelin and cytosol fractions were separated but myelin and whole sciatic nerve lysate (which is the input before isolation of the myelin fraction, called “lysate”). Therefore, the analysis aimed at showing an enrichment of PMP22 and P0 in the myelin fraction while PTEN and TUJ (as a control) are not, which makes it more unlikely for PTEN and PMP22 to interact directly. This experiment, together with the immunohistochemical analysis in Figure 1h should highlight the location of PMP22 and PTEN in the Schwann cell. Together with the newly suggested experiments of the Reviewer for Figure S1 (see planned Revision point 1) we do not see the need for extra membrane-cytosol fractionations and/ or another timepoint as the more detailed as the improved experiment on protein-protein interaction using nerve lysate (not only cell culture) is the experiment of choice to clarify whether we have a direct interaction or not.

Regarding in vitro Schwann cell- DRG co-culture experiments:

(Figure 2, Figure 4 and Figure 5e)

1. *(Figure 2) For this experiment, pulse treatment may be beneficial rather than in continuous. Is Akt-mTOR phosphorylation-signaling increased also in Pmp22+/- co-cultures as in mutant nerves? Is the treatment reducing the overactivation? *
2. *(Figure 4) Similarly to Figure 2, is PTEN level increased in Pmp22tg cultures along with Akt-mTOR downregulation? *
3. *(Figure 5) Panel e: co-cultures are established using ex vivo Dhh-Cre recombination. The downregulation of Pten in the cultures should be documented. * Response: We agree with Reviewer #1 that a deeper analysis of the co-culture system regarding the downstream signaling of PTEN would increase the value of the experiments. Unfortunately, the experiments were designed in a very small scale with the intention of only evaluating myelin alterations on a histological level and we did have enough tissue to collect cells for deeper protein expression analysis. Moreover, we tried to use the co-culture system as a proof-of-principle experiment in parallel to our in vivo studies which we value more important due to the still quite artificial co-culture setup. We hope that the Reviewer can understand our approach with the focus we set on the in vivo work.

Figure 3:

1. *The RAPA treatment seems to increase Pten level in the mutant even above wildtype levels (panel b), which can result in decreased myelin thickness due to downregulation of Akt-mTOR. A different method to normalize expression levels should be used. * Response: Comparing the mean, relative expression levels resulting from our quantification as plotted in the graph (panel b) revealed no increase above wildtype level after Rapamycin treatment in the HNPP mouse. Further, we decided for whole protein staining as the superior approach to loading control because we have observed alterations in the expression of other frequently used “housekeepers” such as GAPDH, Actin and Vinculin in the CMT1A rodent models.

*Panel c-e: Can these data also be reproduced in quadriceps nerves as tomacula are more prominent in these Pmp22+/- nerves showing less variability due to the prevalence of large caliber axons? *

Response: Unfortunately, quadriceps nerves were not collected for histology in the experiment and therefore we cannot redo the quantification. Nevertheless, we agree that the quadriceps nerves have less variability than the sciatic nerve and will definitely include the tissue in our future experiments.

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 #3

Evidence, reproducibility and clarity

In this paper Sareda and co-workers demonstrate that the PTEN/mTOR pathway is indirectly involved in regulating myelin thickness and wrapping in models of altered PMP22 gene dosage both in vitro and in vivo. Inhibition of this pathway decreases myelin thickness in models of HNPP, while increasing myelin thickness in models of CMT1A. The evidence for these conclusions is complex but reasonably presented, and the conclusions mainly supported by the data. The abstract for this paper, however, presents a somewhat oversimplified conclusion that the PTEN pathway mainly modifies models of HNPP, where the paper clearly demonstrates that models of CMT1A are also affected by this same pathway. This should be clarified.

Significance

These data are significant, since they would provide new targets for treating inherited neuropathy associated with altered PLP22 gene dosage.

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

This study investigates the modulation, both genetically and pharmacologically, of the PI3K/Akt/mTOR signaling in preclinical animal models for the inherited peripheral neuropathies HNPP and CMT1A. These conditions result from a gene dosage abnormality of the peripheral myelin protein gene PMP22. The exact biological molecular mechanisms remain enigmatic despite it having been over 30 years since the major genetic lesions, the CMT1A duplication and HNPP deletion, were described. With respect to myelin biology one observes focally slowed nerve conduction at pressure palsies and local/segmental hypermyelination in HNPP whereas hypomyelination occurs in CMT1A.

The study is nicely conducted, data illustrations very informative, and writing clear and concise. This paper will likely be of great interest to your readers. A few things the authors may want to consider:

1. Regarding in the Introduction: "...the molecular mechanisms causative for the abnormal myelination remain largely unknown and still no therapy is available." Suggest consider modifying to perhaps: '...no small molecule or pharmacological therapeutic intervention exist.' To say "no therapy" exist is 'myopic' and untrue.
2. Suggest adding question mark to end of sentence or changing 'asked' to "investigated" for following thought: "Here, we asked whether PI3K/Akt/mTOR signaling provides therefore a therapeutic target to treat the consequences of altered Pmp22 gene-dosage."
3. Rather than attempt to establish PRIORITY perhaps 'softening' the INTRODUCTION concluding statement "Our results thus identify a potential pharmacological target for this inherited neuropathy.
4. [This makes thePI3K/Akt/mTOR pathway a promising target for a preventive treatment of affected nerves also in human patients.] Does this belong in RESULTS? Or rather DISCUSSION?

Significance

The authors provide convincing evidence that the HNPP pathobiology is ameliorated by PI3K/Akt/mTOR inhibitors. Interestingly they found radial myelin growth was most affected by this approach and suggest an interesting transdermal approach in injured nerves in the acute prevention of pressure palsies.

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 paper the authors report a direct correlation between PMP22 and PTEN expression levels in the nerve of CMT mutants. In CMT1A Pmp22tg rat nerves, PTEN levels are increased, whereas in Pmp22+/- mutants, a model of the HNPP neuropathy, PTEN levels decrease. Consistent with this, Pmp22tg nerves display lower Akt phosphorylation and, viceversa, Pmp22+/- nerves have higher Akt phosphorylation. The authors lowered PTEN in the transgenic and inhibited mTOR using Rapamycin in the Pmp22+/- to support the functional relevance of the PMP22-PTEN correlation.

I have major concerns on the data as shown, which, in my opinion, don't support the main conclusion of this paper. In more detail:

Figure 1 Panel a: the decrease of Pten expression should be quantified with at least n=3 taking into account the variability among different samples at the different time points indicated (the same applies in panel b, even if here the increase of Pten expression level in Pmp22tg nerves is more evident) Panel a and b: the statement that Pten is more expressed at P18 at the peak of myelination in wildtype nerves is not supported by the blots as shown

Figure S1, page 4: what does it mean "in line with this finding we were unable to detect protein-protein...". May be the authors meant: since there is a direct correlation between Pmp22 and Pten expression levels in the mutants, the authors explored the possibility of an interaction between the two. Regarding the co-IPs, in panel a, the co-IP at the endogenous level, the immunoprecipitation efficiency of PMP22 is very low. May be a pull-down experiment using either exogenous purified PMP22 or PTEN and nerve lysates can help to rule out the possibility of an interaction. The experiments in b, c are performed in overexpression in a heterologous system (293 cells).

Panel e: may be with this experiment the authors aim to suggest that Pten and Pmp22 are unlikely to interact directly or indirectly since Pten is cytosolic and Pmp22 myelin-membrane enriched. However, this myelin purification shows that Pmp22 as P0 expression levels are also abundant in the cytosol, may be also because P18 has been chosen as time point. What about a different type of membrane-cytosol fractionation experiment and/or another time point?

Page 4: "Taken together, Pmp22 dosage inversely correlates with the abundance of PTEN...": please revise this statement

Additional experiments are needed to support the conclusion of Figure 1 that, in the two mutants, Pten levels reversely correlate with PI3K-Akt-mTOR pathway activation, which represents the rationale of all further experiments. For example, it should be shown systematically in both mutants both Akt and ERK phosphorylation levels (Akt at both T308 and S473), and mTOR activity read outs. In the previously published paper (Fledrich et al.) only increased Akt phosphorylation in Pmp22+/- nerves was reported, whereas Pmp22tg analysis was focused on the interdependence between Akt and ERK without exploring mTOR activation, which is relevant here.

Figure 2 The aberrant myelin figures displayed are similar to myelin ovoids preceding degeneration rather than myelin outfoldings. It is also strange that these alterations are in the wildtype cultures treated with RAPA, that instead, in this system, has been reported to increase myelination as it improves protein homeostasis (autophagy, quality control, etc). Also for this experiment, pulse treatment may be beneficial rather than in continuous. Is Akt-mTOR phosphorylation-signaling increased also in Pmp22+/- co-cultures as in mutant nerves? Is the treatment reducing the overactivation?

Figure 3 The RAPA treatment seems to increase Pten level in the mutant even above wildtype levels (panel b), which can result in decreased myelin thickness due to downregulation of Akt-mTOR. A different method to normalize expression levels should be used. Panel c-e: aberrant fibers should be normalized on total number of fibers and on the area, particularly because RAPA is used. Can these data also be reproduced in quadriceps nerves as tomacula are more prominent in these Pmp22+/- nerves showing less variability due to the prevalence of large caliber axons?

Figure 4 A different model, the C61 mouse a Pmp22tg overexpressing PMP22 is used here (rather than the CMT1A rat). This should be explained in the results. Is also this model characterized by increased Pten levels in the nerve? And low Akt-mTOR activation for instance?

The improvement in the number of myelin segments following PTEN inhibition in Pmp22tg co-cultures is very weak.. The 500 nM has instead a consistent effect in reducing myelin segments in the wildtype and I think that these results overall don't support the conclusion that myelination is ameliorated by reducing PTEN activity in Pmp22tg co-cultures. Similarly to Figure 2, is PTEN level increased in Pmp22tg cultures along with Akt-mTOR downregulation?

Figure 5 As for Figure 4, the use of the mouse transgenic instead of the CMT1A rat should be specified and PTEN, Akt-mTOR expression/activation levels should be checked biochemically also in this model. And quantified (panel c). In panel d overactivation of mTOR (PS6 staining) in Schwann cells is not evident. Panel e: co-cultures are established using ex vivo Dhh-Cre recombination. The downregulation of Pten in the cultures should be documented. Pten Fl/+ Dhh-Cre cultures seem to have axonal fasciculation.

Figure 6 G-ratio analysis: which are the mean values (numbers) with SEM in the three groups analyzed wildtype, Pmp22tg and Pmp22tg; Pten fl/+; Dhh-Cre? How is Akt-mTOR signaling in the double mutant as compared to Pmp22tg? Is that increased at P18? If more fibers are committed to myelinate in the double mutant as compared to the single Pmp22tg at P18 ,particularly, it is unclear why there is no difference in differentiation marker expression in Figure 7 (Oct6 and Hmgcr).

Significance

In conclusion, the correlation between PMP22 and PTEN is a potential interesting observation. However, in my opinion, experiments as shown don't support the conclusion that PMP22 controls PTEN expression level and activity, which is suggested at the basis of the pathogenesis of PMP22 dosage-related neuropathies.

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

Reviewer 1:

• *

I am not an expert in algal ultrastructure or TEM, and I principally found this manuscript informative and well-written. I can only make minor recommendations, although I urge the Editors to supplement this review with others from specialists within the field.

We thank the reviewer for their positive comment on the manuscript as well as for the interesting following discussion points.

To date, to my knowledge, the Pentapharsodinium chloroplast has not been characterised at a molecular level. The placement of the nuclear lineage within the Peridiniales as a close relative or Durinskia spp. and other dinoflagellates with diatom endosymbionts raises the question of whether Pentapharsodinium possesses a peridinin-containing chloroplast, per most other dinoflagellates, or possesses a chloroplast of an alternative endosymbiotic derivation, although I suppose the single chloroplast would be typical for peridinin-containing chloroplasts from the Peridiniales. __Can the authors make any inference on this from their data? __

In the course of this study, we did not perform molecular characterization of the chloroplast. Therefore, formal answer to this question would not be possible based on our data.

However, we have evidence that suggest that the chloroplast from P. tyrrhenica does not come here from a diatom endosymbiont.

1/ In the suggested cases of diatom endosymbiont origin of the chloroplast, eg in the genus Durinskia, Yamada et al., 2019 show that Durinskia capensis or Durinskia kwazulunatalensis can present diatom organelles such as chloroplast, mitochondria, nucleus, or their remnants for a certain amount of time after uptake. The detailed analysis of our vEM data does not show these diatom organelles, thus ruling out the hypothesis.

2/ Additionally, the eyespot of dinoflagellate species possessing diatom association seems to be characteristic to these cells (Horiguchi., 2007, Hoppenrath et al., 2017). In our case, the eyespot of P. tyrrhenica corresponds to a distinct type (type IA, thought to be in peridinin dino-chloroplast according to Hoppenrath., 2017), that varies from the one observed in Durinksia baltica, or other dinoflagellates species possessing these diatom associations (Horiguchi., 2007, Hoppenrath et al., 2017).

Therefore, the chloroplast of our cell does not seem to be acquired from a diatom. However, as we don’t have molecular evidence at this stage, and are not able to perform such analyses, we prefer not to address this question in the manuscript.

While I appreciate that this is a study of a single cell only, I would prefer some more extensive evidence that the partial chromosome unfolding identified correlates to transcriptional activity. The nucleolus is surrounded by a layer of heterochromatin and perhaps the filamentous structures involved are transcriptionally quiescent. Were the authors able to take any preliminary images of cells harvested mid-day or exposed to higher light intensities, and do they see greater chromatin unfolding in this case? Similarly I would be curious if cells visualised later in the day possess multiple rather than single chloroplasts.

Our description is here based on one microorganism of very low abundance, and we do not have data for it across conditions. Therefore, we would not be comfortable inferring too many details in this paper and will thus modify the text in order be more careful about the potential role of these fibres and their putative association to transcription. We are planning to address this interesting point in the future, in a more biologically focused paper in preparation. Nevertheless, similar structures have been described protruding in the nucleoplasm in other species (Soyer., 1981, Bhaud et al., 2000, Decelle et al., 2021) and have been suggested to be associated to RNA transcription (Sigee., 1983). We will make sure to cite and discuss more literature on this point.

Line 39: should be "dinoflagellate biology" <br /> Line 131: "a far red signal" <br /> Line 196: should be "number of chloroplasts"

We thank the reviewer, the errors will be corrected in the revised version.

Line 200: by curiosity, how does the measured chloroplast volume compare to those computed in vEM studies of Symbiodinium (c.f., Uwizeye 2021)

The measured chloroplast volume in our cell differs to those computed in Symbiodinium in Uwizeye et al., 2021. Indeed, in Uwizeye et al., 2021, the chloroplast represents 30% of the cell volume. This is more than in our cell for which the chloroplast represents only 9.5% of the cell volume. However, these two cells are different species and come from different growth conditions (culture VS environment). These factors potentially contribute to morphometric variations as the environment possesses different types and amounts of nutrients and light resources.

Line 221: how does the number of observed chromosomes compare to estimated chromosome numbers in dinoflagellates from karyotyping or whole-genome sequencing?

To our knowledge, there is a wide variability in terms of number of chromosomes observed between dinoflagellate species. Bhaud et al (2000) reports the presence of between 4 and 200 chromosomes depending on the species. Unfortunately, we did not find information concerning the number of chromosomes for this species to compare with our analysis. However, we are planning in the revised version to discuss how our workflow is complementary to other methods such as genomics, transcriptomics and light microscopy towards better understanding of marine organism.

Line 240: does the eyespot show any proximity to the mitochondria, as per the hybrid chloroplast/ mitochondrial-derived eyespots found in Warnowiacean dinoflagellates?

In our study the eyespot is composed of pigment globules, organized as a sheet, located inside the chloroplast and facing towards the theca. This type of arrangement seems to be distinct to the chloroplast/mitochondria-derived eyespot described in Warnowiaceae (Colley and Nilson., 2016). Additionally, for the revision of this paper we will further segment the mitochondria to investigate any potential proximity to the eyespot. We will also put the eyespot of P. tyrrhenica in context with other types described in the literature.

Reviewer #1 (Significance (Required)):

The application of vEM to environmental algal samples has to my knowledge not been attempted previously. If these approaches could be scaled up to a multi-cell approach and is not completely destructive to the cells, it could provide a fascinating way to connect algal morphology in the wild to other culture-free methods to understand algal biology (e.g., meta-barcoding).

We would like to address the interesting comment concerning the possibility to combine vEM along with other molecular tools to study organisms from a culture free method. While FIB-SEM is a destructive method, it could be used in combination with other methods and thus synergize towards the characterization of the environmental samples. We will add a few sentences in the discussion, as a forward look, as it represents indeed an important axis of our future research, where molecular taxonomy (e.g. metabarcoding) will be correlated to morphological characterisations. We believe such approaches will be useful beyond the study of micro-algae, and will make this point clear in the discussion.

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

This is avery important and interesting contribution to the ultrastructural analysis of one scientific species among many others living together in a rich sample as a marine microplankton. In addition the work also shows the possibility to obtain fundamental volumetric informations on the various structures and organelles. The work was well planned and executed and certainly represents a tremendous effort of the members of the group. It is clearly written, explaining each detail of the methodology necessary to the understanding of the whole work.

Reviewer #2 (Significance (Required)):

This a phantatis piece of scientific work where the authros were able to use the moderns three dimensional reconstruction technique possible using high resolution scanning electron to reconstruct one specific cell in a large population of heterogenous cells. The identification of one specific cell based on fluorescence images detected in a light microscopy was very important to present one new methodology to observed such types of cells. In addition to a detailed description of the strucutres and organelles found they were able to determne the area/volume occupies by each of them in cells. <br /> Therefore I strong recommend the acceptance of the manuscript as it is.

We thank the reviewer for their appreciation of our work.

Reviewer #3 (Evidence, reproducibility and clarity (Required)):

# Summary --

The authors present a workflow to characterize a microorganism, dinoflagellate cell, from environmental samples by vEM. The workflow enables the identification of specific taxa of interest in a heterogenous environmental sample and allows correlative fluorescence and FIB-SEM acquisition. I see no major flaws with this study. The authors present a proof of principle pipeline for utilizing CLEM to explore the ultrastructure of microorganisms.

# Major comments --

What is not clear, is how highly reproducible this workflow is. For example, what is the time required for each sample? How manual vs automated is each step? While, the workflow is important and sound, it would be helpful for the reader to understand a little more about the variability and throughput. It doesn't need to be exhaustive, but further characterization of the workflow would significantly improve the impact of the manuscript and should be included. As a reader, it would be tremendously helpful to clearly state what is part of the preparation/imaging workflow and what is an application example of the workflow. If it is intended that the post-processing is also part of the workflow, there should be significantly more details provided into the segmentation and analyses processes.

We thank the reviewers for pointing this out. This workflow has been applied to several blocks and repeated imaging of a subset of species could be achieved with 100% success. We will show this in a follow up paper. Moreover, the same method has been applied for targeted FIB-SEM acquisition of samples expressing fluorescent proteins (Ronchi et al., 2021). Therefore, we believe our workflow is highly reproducible and robust. We are planning to address the concern on technical variability in our revised version, together with a better explanation of time scales and automated VS manual aspect of the steps that are used for this workflow. We will also add information concerning the segmentation.

# Minor comments --

As this study describes a workflow that was developed for identification and imaging of microorganisms, it would be highly beneficial to the reader to have a figure that shows each of the steps, end to end.

We thank the reviewer for this suggestion. We plan to add a supplementary figure that describes the workflow step by step.

It would also be helpful to add annotation labels and a scale bar to supplemental video 2.

We thank the reviewer for spotting this missing information. We plan to add annotation labels as well as a scale bar to the video S2.

Do the nonphotosynthetic species also have nuclear autofluorescence or is this just a trait of a subset of the photosynthetic species?

We thank the reviewer for this question. From our analysis, we can’t conclude that it is a trait of some photosynthetic species compared to non-photosynthetic ones. However, heterogeneous autofluorescence profiles, even though unexplained, represent a valuable tool to discriminate between cell types. We would like to further address this point in the manuscript that such fluorescence profiles, even though unexplained, can contribute to the identification of specific cell types.

Reviewer #3 (Significance (Required)):

# General assessment --

The authors present an important, yet missing, workflow for characterizing microorganisms from environmental samples using vEM and CLEM. A strength of the study is that it enables identification and selection of taxa in heterogenous samples. Using correlative, confocal imaging of both auto-fluorescence and transmitted light, the authors show that specific taxa can be identified and selected for according to the organism's photosynthetic and morphological properties.

We thank the reviewer for their positive comments.

To improve upon this, it could be useful for the authors to provide a list or table of potential organisms that could be selected for in this manner to exemplify the use cases of this protocol.

We thank the reviewer for their suggestion. We plan to add a supplementary figure with a gallery of different cells and their identification.

# Advance --

This study, while a proof of concept, is also one of the first examples of using vEM on environmental studies. The author's also present the potential value add of using CLEM, not just for selection purposes, but also more comprehensive identification and mapping of subcellular structures. While the workflow itself is incremental (and important), the application is quite novel.

# Audience --

Because the authors' study combines a vEM workflow and microorganism characterization, the potential audience is broad reaching. The workflow itself could be adapted to many different systems - beyond microorganisms - making it of use to several biological fields.

We thank the reviewer for their positive comments.

# Expertise --

vEM, CLEM, FIB-SEM, membrane trafficking, cell biology, tissue cell biology, machine learning

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 #3

Evidence, reproducibility and clarity

Summary

The authors present a workflow to characterize a microorganism, dinoflagellate cell, from environmental samples by vEM. The workflow enables the identification of specific taxa of interest in a heterogenous environmental sample and allows correlative fluorescence and FIB-SEM acquisition. I see no major flaws with this study. The authors present a proof of principle pipeline for utilizing CLEM to explore the ultrastructure of microorganisms.

Major comments

What is not clear, is how highly reproducible this workflow is. For example, what is the time required for each sample? How manual vs automated is each step? While, the workflow is important and sound, it would be helpful for the reader to understand a little more about the variability and throughput. It doesn't need to be exhaustive, but further characterization of the workflow would significantly improve the impact of the manuscript and should be included.

As a reader, it would be tremendously helpful to clearly state what is part of the preparation/imaging workflow and what is an application example of the workflow. If it is intended that the post-processing is also part of the workflow, there should be significantly more details provided into the segmentation and analyses processes.

Minor comments

As this study describes a workflow that was developed for identification and imaging of microorganisms, it would be highly beneficial to the reader to have a figure that shows each of the steps, end to end.

It would also be helpful to add annotation labels and a scale bar to supplemental video 2.

Do the nonphotosynthetic species also have nuclear autofluorescence or is this just a trait of a subset of the photosynthetic species?

Significance

General assessment

The authors present an important, yet missing, workflow for characterizing microorganisms from environmental samples using vEM and CLEM. A strength of the study is that it enables identification and selection of taxa in heterogenous samples. Using correlative, confocal imaging of both auto-fluorescence and transmitted light, the authors show that specific taxa can be identified and selected for according to the organism's photosynthetic and morphological properties. To improve upon this, it could be useful for the authors to provide a list or table of potential organisms that could be selected for in this manner to exemplify the use cases of this protocol.

Advance

This study, while a proof of concept, is also one of the first examples of using vEM on environmental studies. The author's also present the potential value add of using CLEM, not just for selection purposes, but also more comprehensive identification and mapping of subcellular structures. While the workflow itself is incremental (and important), the application is quite novel.

Audience

Because the authors' study combines a vEM workflow and microorganism characterization, the potential audience is broad reaching. The workflow itself could be adapted to many different systems - beyond microorganisms - making it of use to several biological fields.

Expertise

vEM, CLEM, FIB-SEM, membrane trafficking, cell biology, tissue cell biology, machine learning

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

This is avery important and interesting contribution to the ultrastructural analysis of one determoned species among many others living together in a rich sample as a marine microplankton. In addition the work also shows the possibility to obtain fundamental volumetric informations on the various structures and organelles. The work was well planned and executed and certainly represents a tremendous effort of the members of the group. It is clearly written, explaining each detail of the methodology necessary to the understanding of the whole work.

Significance

This a phantatis piece of scientiic work where the authros were able to use the moderns three dimensional reconstruction technique possible using high resolution scanning electron to reconstruct one specific cell in a large population of heterogenous cells. The identification of one specific cell based on fluorescence images detected in a light microscopy was very important to present one new methodology to observed such types of cells. In addition to a detailed description of the strucutres and organelles found they were able to determne the area/volume occupies by each of them in cells. Therefore I strong recommend the acceptance of the manuscript as it is.

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

I am not an expert in algal ultrastructure or TEM, and I principally found this manuscript informative and well-written. I can only make minor recommendations, although I urge the Editors to supplement this review with others from specialists within the field.

To date, to my knowledge, the Pentapharsodinium chloroplast has not been characterised at a molecular level. The placement of the nuclear lineage within the Peridiniales as a close relative or Durinskia spp. and other dinoflagellates with diatom endosymbionts raises the question of whether Pentapharsodinium possesses a peridinin-containing chloroplast, per most other dinoflagellates, or possesses a chloroplast of an alternative endosymbiotic derivation, although I suppose the single chloroplast would be typical for peridinin-containing chloroplasts from the Peridiniales. Can the authors make any inference on this from their data?

While I appreciate that this is a study of a single cell only, I would prefer some more extensive evidence that the partial chromosome unfolding identified correlates to transcriptional activity. The nucleolus is surrounded by a layer of heterochromatin and perhaps the filamentous structures involved are transcriptionally quiescent. Were the authors able to take any preliminary images of cells harvested mid-day or exposed to higher light intensities, and do they see greater chromatin unfolding in this case? Similarly I would be curious if cells visualised later in the day possess multiple rather than single chloroplasts.

Finally, I have a few (very) small grammatical corrections:

Line 39: should be "dinoflagellate biology"

Line 131: "a far red signal"

Line 196: should be "number of chloroplasts"

Line 200: by curiosity, how does the measured chloroplast volume compare to those computed in vEM studies of Symbiodinium (c.f., Uwizeye 2021)

Line 221: how does the number of observed chromosomes compare to estimated chromosome numbers in dinoflagellates from karyotyping or whole-genome sequencing?

Line 240: does the eyespot show any proximity to the mitochondria, as per the hybrid chloroplast/ mitochondrial-derived eyespots found in Warnowiacean dinoflagellates?

Significance

The application of vEM to environmental algal samples has to my knowledge not been attempted previously. If these approaches could be scaled up to a multi-cell approach and is not completely destructive to the cells, it could provide a fascinating way to connect algal morphology in the wild to other culture-free methods to understand algal biology (e.g., meta-barcoding).

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

The authors do not wish to provide a response at this time.

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 #3

Evidence, reproducibility and clarity

The authors of this study utilize a novel nanobody-based technique to specify the location of the SPT complex to either the peripheral or nuclear membrane-associated endoplasmic reticulum membranes. Considering the potential importance of sub-ER compartmentalization on metabolic enzymes of the ER, this is a novel and useful approach. The studies are, with the minor exceptions noted below, comprehensive and very well executed and documented. The authors have combined genetic, proteomic, lipidomic, and flux experimental approaches to test whether sub-ER compartmentalization affects the function and regulation of the SPT complex. The results are, for the most part, negative, although there does seem to be some effect on the overall activity of the SPT complex as measured with flux analysis. Overall, while the authors do not detect dramatic effects on SPT complex localization, the technical advance using tethered nanobodies to direct complex localization, and the complementary approaches to testing SPT function and regulation, will be useful to workers in the sphingolipid field.

Minor points:

The results with YPK1-linker-CAAX are confusing. This construct does not result in Orm2 phosphorylation with heat shock, whereas endogenous YPK1 does. Yet it can support viability even without Orm deletion. In other words, this tethered construct appears functional in viability assays, but not in a biochemical assay.This discrepancy is not discussed by the authors. The manuscript would be improved by a discussion by the authors that addresses this issue. It is not clear why the figure legend to Figure 2 suggests that Ypk1 regulates Orms mainly in the peripheral ER. Considering that WT Ypk1 is more efficient than CAAX tethered YPK1, this statement does not seem supported. Perhaps the authors can elaborate on how they came to this conclusion.

The figures depicting Orm phosphorylation (Figure 1e, f Figure 2d,e, Figure 6 b,c) should be improved. The resolution of two forms is not sufficient in Figure1 and 2. The use of Phos-Tag might solve this issue. It would be helpful to the reader to include arrows that indicate the phosphorylated and unphosphorylated forms of Orm. Quantitation of these gels is essential.

Lines 318 and 319. Figure 6e and 6f are referred to. The correct assignment is 6f and 6g.

Referees cross-commenting

I agree with Reviewer #2's assessment that some of the conclusions are over stated. While Reviewer #2 is correct that the advances in this manuscript are modest, this is principally because expected differences in the function and regulation of the SPT in different ER sub-domains did not materialize. This may be disappointing, but is still important to document

Significance

This is a very well performed study, utilizing a variety of approaches to test whether localization of the SPT complex impacts on it activity and regulation. With very minor exceptions, it is well executed and documented.

The advances reported here are two-fold. First, the authors introduce a novel approach using nanobodies that are tethered to distinct regions of the yeast endoplasmic reticulum to localize intact and unmodified complexes to distinct locations. This could be a very useful tool in other contexts to examine the role of subcellular compartmentalization in the function of enzymes and signaling components. This targeting system is well characterized in this study. The second advance, utilizing this targeting system, is that localization of the SPT complex to distinct subcompartments of the ER has minimal effects on regulation, and observable, but relatively minor effects on SPT function in terms of sphingolipid production. While a positive result would have been more exciting, negative results can be equally informative.

This study will be of interest to workers in the signaling and metabolic fields that may utilize this unique targeting strategy. It will also be of interest to the sphingolipid community.

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

This manuscript uses a combination of immunoblotting, microscopy, and MS-based lipidomics to study sphingolipid synthesis in S. cerevisiae. The authors recently published a paper demonstrating that exogenous serine taken from the medium is preferentially used to generate LCBs. Given multiple levels of regulation of the SPT complex, the authors postulate that SPT in different sub-compartments of the ER could be differently regulated or at least have variable activities. Using a nanobody capture approach, they can restrict SPT activity to the peripheral ER and nuclear ER. Using this model, they investigate the role of Orm phosphorylation and SPT activity in response to 5-min heat shock. Phosphorylation does not seem to be a key element of the regulation. Ultimately, this is a collection of experiments without a clear story. In the end, the only take-home message I can find is that peripheral SPT is able to use exogenous heavy Serine as a substrate better than nuclear SPT.

Page 4. "We had previously demonstrated that increased de novo LCB biosynthesis is directly dependent on the uptake of exogenous serine through the general amino acid permease Gnp1. Consistent with our previous findings, deletion of GNP1 resulted in a blunted heat shock response, while deletion of the endogenous serine biosynthesis pathway (ser2) had no effect on LCB biosynthesis". This claim is too strong. If they feel this strongly, the authors should test a gnp1 agp1 double mutant. The alternative explanation is that to support the rapid increase in SPT activity, the Lcb1-Lcb2 enzyme use both a pre-existing cytoplasmic pool of serine and exogenous serine.

If cells are grown for 1 generation with heavy serine to label the cytoplasmic pool of serine and then cells as shifted to serine-free media and heat shock is induced is there any difference in LCB synthesis between allSPT, nSPT and pSPT?

It would be worth re-doing some of these experiments in rtn1 rtn2 yop1 yeast triple mutant (Stefan...Emr Cell 2011) or the delta super-tether mutant (Quon...Menon PLOS Biology 2016) both of which have substantially less cortical ER.

The authors make strong claims that are not supported by the data. Further, the presentation of the data is not optimal, and much of the data is qualitative, not quantitative. Too much of the data is presented as fold-change and i believe that the base-line LCB levels may change. The raw data should be included in a supplement excel file.

The blots of FLAG-tagged Orm1 and Orm2 are a critical part of this manuscript, but the data is not compelling in most figures. There is a lack of quantitation and replication. On the surface I agree with the authors that the phosphorylation is hard to align with the stimulation, but a more rigorous analysis is needed. Additionally, does the expression of an Orm2 mutant without the Ypk1 phosphorylation sites prevent the heat shock-induced increase in LCBs?

If this remains in the manuscript, the difference between ypk2 Ypk1-CaaX and ypk2 Ypk1-103aa-CaaX should be better highlighted in the main document. However, this whole course of experiments is problematic, and the authors' narrative changes to accommodate the findings; the arguments aren't internally consistent. Ypk is essential to regulate Orms. Ypk1-linker-CaaX can regulate Orms, Ypk1-linker can't increase Orm2 phosphorylation. Furthermore, a single prenyl group is not sufficient to restrict the localization of a protein. If there are other targeting motifs in addition to the last 4 amino acids this should be indicated. Ultimately, this is all negative data. Without a better interrogation of Orm phosphorylation it seems to have little value.

The "Orm proteins mediate SPT upregulation after heat shock" should be re-worded. The orm1 orm2 mutant already has a 25-fold increase in LCBs, and heat shock is unable to stimulate SPT activity any further. It suggests that heat shock is directly inhibiting Orm proteins which in turn removes the feedback inhibition on the SPT.

Figure 7b - the Y-axis scale needs to be corrected. A break should be inserted to alert the reviews of the scale's narrow range.

Is GFP-Orm1 and Orm2 functional? Both are more abundant in the nER than the peripheral.

Most papers I have read use "SL" as the abbreviation for sphingolipids, not SP. Consistency would be nice for readers.

Significance

The work appears technically sound but there are issues with the presentation of the data. Lack of raw data for the lipidomics, lack of quantitation.

The findings here are rather modest and much of the data is inconclusive or suggests certain pathways aren't involved rather than defining a clear mechanism.

This is an incremental finding that supports the 2020 PLOS genetics paper from the same authors. The target audience for this would be people interested in sphingolipid metabolism in yeast.

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

This is a nice paper showing that the intracellular location of the SPOTS complex affects SL de-novos synthesis. The experiments are well designed including a comprehensive set of controls. I just have a few minor points: It is hard for me to see the differences in ORM phosphorylation on the blots. Does phosphorylation refer to the more intense upper band (Fig 1e+f)? Given the general variability of WBs it might be better to demonstrate the differences in ORM phosphorylation with a more specific and quantitative method? e.g by a phosphoprotein stain or (better) by targeted (phospho)proteomics.

Significance

Unfortunately, the discussion focusses mostly on technical aspects of the method. However, the biological observations concerning the regulation of the SL metabolism itself are interesting and relevant too. They should be discussed in more detail, which might also be of interest to a more general readership. Concerning the factors involved in regulating SL de-novo synthesis, the authors did not mention CERT which also contributes to the regulation of SL de-novo formation (PMID: 36976648) and connects to SacI which is a component of the SPOTS complex.<br /> Why does yeast have two ORM isoforms and to which extend are they redundant? The authors see functional difference between the two ORM isoforms, which could be discussed in more depth. It might also be interesting to interlink the findings to the mammalian system, which is based on three ORMDL isoforms, and appear not to be regulated by phosphorylation. Another aspect that could be discussed in this contet is the observation that mammalian SPT appears preferentially located at MEM contact sites, which indicates a special role of SL de-novo synthesis at this location (PMID: 34785538). However, overall this is a well done paper.

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

Reply to Reviewers

We thank both Reviewers for their comments which we believe could greatly improve the manuscript by adding more functional data. Part of the revision process has been already carried out and part of it is ongoing as detailed in the following sections.

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

Comments to the Authors In this manuscript, Girolamo et al., describe the differences in molecular signature and biological features of MuSC populations between extraocular (EOM) and limb (Tibialis anterior, TA) muscles. Comprehensive approaches including scRNA-seq, bioinformatics, and live-cell imaging reveal that EOM MuSCs is more proliferative in vitro and express ECM components at high levels as well as non-myogenic markers such as Foxc1 and Pdgfrb. The transcription factor network described in this study will characterize the identity of EOM MuSCs, providing insights into stem cell-based therapies for muscle diseases.

Major comments 1. In Figure 1A, EOM-MuSCs appear to form myotubes as efficiently as TA-MuSCs. If so, why myotube formation is not affected in EOM-MuSCs even with lower Myog expression compared to TA-MuSCs? The authors should discuss about this point.

We thank the reviewer for raising this point. Actually, EOM myoblasts show a delay in activation of Myog that allows their sustained expansion, however Myog activation per se is not impaired. This is reflected by the presence of the EOM Differentiating cluster in our sc-RNAseq analysis which converges towards the TA Differentiating cluster and shows co-expression of Myog and canonical differentiation markers such as Myh3 and Troponins (Figure 2C, D). Accordingly, Stuelsatz et al. (Dev Biol 2015) showed efficient and robust differentiation of EOM myoblasts in vitro, while a higher fraction of self-renewed (Pax7+) cells was observed in long term cultures.

To formally address this point, we propose to:

• Isolate EOM and TA MuSCs from Tg:Pax7-nGFP;MyogtdTom mice, plate them at low density and quantify the tdTom+ cell proportions in short vs longer term cultures (D3 vs D5, D8).
• Re-isolate the tdTom+ mononucleated fraction at D4-5 upon activation to:
• plate EOM and TA myoblasts at high density and assess their fusion index. Here, we will evaluated fusion independent of proliferation history and Myog activation.
• perform qRT-PCR for Myomixer/Myomaker on the tdTom+ mononucleated fraction to assess the extent of fusion potential.

  1. In Figure 2, activated EOM-MuSCs express Myod1e at lower levels compared to TA-MuSCs. Are MyoD protein levels are also lower in EOM-MuSCs?

We have performed quantitation of the intensity of fluorescence for Myod at D3.5 in culture. In agreement with the sc-RNAseq in vitro activated dataset, we have noted lower Myod protein levels in EOM MuSCs. This data is now included as part of Suppl Fig 2A.

In Figure 7E-G: Does the Foxc1 knockdown change the myogenic potential, especially on myotube formation? In addition to FOXC1 and Myog, the mRNA levels of Pax7 and Myod1 would be better to be provided.

Previously, we used siRNA for short term silencing. As siRNA are diluted upon cell division, we now use lentivirus-mediated KD, which provides more consistent results. To address this point, we propose to use lentivirus carrying 3 different shRNAs for Foxc1 and address expression levels of Pax7, Myod, Myog together with EdU detection and assessment of myotube formation. Given that cell density influences myotube formation, we will pre-amplify the cells, replate them at high density, and silence Foxc1 concomitantly with induction of differentiation. These 3 shRNAs have been already validated in vitro and induce a massive reduction in EOM cell numbers. This data is now included as part of __Figure 7E-L. __

Please state a reasonable explanation for the physiological role of high amounts of ECM produced by EOM-MuSCs.

We regret that this explanation did not come across clearly on the manuscript. Previous studies have demonstrated a greater cellular output of cranial MuSCs in clonal assays in vitro (Ono Dev Biol 2010, Stuelsatz et al. Dev Biol 2015; Randolph et al. Stem Cells 2015) and better engraftment capacity in vivo (Stuelsatz et al. Dev Biol 2015). On the other hand, it has been shown that fetal MuSCs, which produce high amounts of extracellular matrix (ECM) cell autonomously, expand more efficiently and contribute more to muscle repair than the adult counterparts (Tierney et al. Cell Reports 2016). Finally, while in vitro expanded MuSCs show a reduced engraftment potential (Montarras et al. Science 2005, Ikemoto et al. Mol Therapy 2007), recapitulation of the endogenous niche ex vivo, allows maintenance of an undifferentiated proliferative state and their capacity to support regeneration in vivo (Ishii et al. Stem Cell Reports 2017). Therefore, we hypothesize that in vitro activated EOM MuSCs secrete high amounts of ECM to self-autonomously maintain stemness when removed from their niche. Alternatively, EOM MuSCs might contribute to connective tissue cells postnatally in vivo as we described previously in the embryo (Grimaldi et al. elife 2022). Secretion of ECM and expression of ECM-related regulons when activated in vivo might recapitulate this process.

To address these hypotheses, we propose to:

• Culture EOM and TA MuSCs, generate EOM and TA decellularized ECM (dECM) and test the proliferation/differentiation potentials of MuSCs on the dECM versus that on wells coated with Matrigel or Fibronectin alone. Shall this experiment not give conclusive results, we propose to assess the proliferation/differentiation potential of TA MuSCs on dishes coated with proteins present in the EOM ECM such as Sparc, Bgn, Mgp, Fbn1, Fn1, for which recombinant proteins exist.
• Assess the expression levels of EOM-specific ECM proteins and TF regulons identified in vitro expressed in EOM MuSCs activated in vivo. To do so we are trying to optimize an EOM injury protocol based on previous observations highlighting the sensitivity of EOMs to anesthetics (Carlson et al. Arch Opthal 1985).
• Use Pax7CreERT2:R26tdTom:PdgfraH2BGFP reporter mice to determine whether EOM MuSCs can give rise to muscle connective tissue postnatally. Minor comments :

On p7, the 3rd line: "marker" is redundant.

This has been corrected.

On p9: what is meant by "force-directed environment"? Is this the aggregation of regulons and targets based on interaction strength determined by the algorithmic arrangement used by pySCENIC?

A force-directed graph is a type of visualization technique where nodes are positioned based on the principles of physics that assign forces among the set of edges and the set of nodes. Spring like attractive forces are used to attract pairs of edges towards each other (connected nodes) while repulsive forces, like those of electrically charged particles, are used to separate all pairs of nodes. In the equilibrium state for this system, the edges tend to have uniform length (because of the spring forces), and nodes that are not connected by an edge tend to be drawn further apart (because of the electrical repulsion). This results in a layout that visually represents the relationships between the nodes, where each node (circle) is an active transcription factor and each edge (distance between nodes) is an inferred regulation between 2 transcription factors.

The text has been changed accordingly.

On p11: add space between (Sambasivan et al., 20009a)and.

Corrected

On p11: 1.7 fold increase → 2.7 fold?

Corrected

On p13: please describe the GRN abbreviation since it is the first used here and not clarified beforehand.

Corrected to gene regulatory network (GRN)

On p15: add space between (Vallejo et al., 2018)Other.

Corrected

On p16: add "a" after "Klf4 is" → Klf4 is a pioneer...

Corrected

In Fig1: it would be better to show the quantification of EdU incorporation, especially at D5 for highlighting the difference between EOM and TA

This has been done and now included Figure 1B.

In Fig1: MF20 staining seemingly describes larger myotubes in EOM compared with TA at D5. Is this most likely due to the higher number of cells to start with in EOM rather than having more fusion ability?

Indeed, as discussed in the Main Comments (question 1), the fusogenic ability of EOM an TA MuSCs will be addressed by plating the same number of cells at high density.

In Fig 2E,F: some font is very faint in color and hard to read in printed format e.g. "TP53 regulates transcription of". May want to change color. In addition, the bottom edge of the Figure is slightly cut off.

We have enhanced the color of some words so all terms are properly seen. The figure has been adjusted in order for the bottom edge to be seen.

In Fig 4D: COLVI should be changed to COLIV.

Corrected

In Fig 5B: Myod1 is redundant.

Actually Myod as regulon (Myod1_(+)) in Fig 5B is not redundant. It is written twice as it is a regulon of both EOM Diff and TA Diff.

In Fig 6D; please specify "Amp". Is Hey1 a myogenic marker?

Amp stands for Activated/Amplified cells. We have changed this to Act to keep consistency.

Hey1 is a bHLH transcription factor that is required in a cell-autonomous manner for maintenance of MuSCs (Noguchi et al. Development 2019). This information has now been added to the text.

In Fig 7F; there are Ctrl and control- please unify them.

Corrected

In Fig7C,D: When does Foxc1 start to be expressed in EOM progenitors in the embryo? If the authors tested, please mention it.

Foxc1 is a DEG and top regulon of EOM progenitors in the early embryo (E11.5, Grimaldi et al. elife 2022). We will formally address this point by immunostaining on tissue sections of E10.5, E12.5 and E14.5 embryos.

In Supp Fig 7A: graphs are lacking color-coded legends.

Corrected

In Supp Fig 7C: Font is very small and illegible in a paper format.

Corrected. For networks it is sometimes difficult to get bigger font size.

In Supp Fig 7F: what are the dynamics of PDGFβ+ cell populations in successive passages in culture? If the authors tested please mention it.

We have been that Pdgfrb expression increases upon passages (Fig 6D). However, this experiment does not tell us whether there is a higher fraction of PDGFRβ+ cells or a similar fraction compared to Day 3 but expressing higher levels of the protein. To distinguish these possibilities, we propose to assess the PDGFRβ+ fraction by FACS upon passages in culture.

In Supp Fig 8 A: Are the colors reversed for EOM? In Fig 8 E, EOM progenitor and differentiation stream are light blue and yellow respectively but in Supp 8A the colors are flipped and don't seem to match the Map laid out in Fig 8E.

The scvelo pipeline will be re-run to correct this error on the graphical output.

Some references are listed as redundant.

Corrected

Reviewer #1 (Significance (Required)):

Skeletal muscle stem cells (MuSCs) play an indispensable role in muscle regeneration in adults. MuSCs are distributed in all muscles throughout the body and their function and molecular properties are diverse among muscles. Extraocular muscles (EOMs) are known to be preferentially spared in muscular dystrophies and during aging. In addition, EOM-derived MuSCs are highly transplantable compared to those of limb muscles. However, intrinsic regulators of EOM MuSCs have not been fully characterized yet. This study by Girolamo et al. shows the differences in molecular signature and biological features of MuSC populations between EOM and limb (Tibialis anterior, TA) muscles. Comprehensive approaches including scRNA-seq, bioinformatics, and live-cell imaging revealed that a subset of the EOM-derived MuSC population is highly proliferative and expresses extracellular matrix components at high levels. The analysis also shows that EOM-derived MuSCs have non-myogenic signatures such as Foxc1 and Pdgfrb. A transcriptional factor Foxc1 is described as a pro-mitogenic factor in the cancer field and is known as a driver of endothelial/smooth muscle fate. In this study, the authors find that Foxc1 is expressed in EOM MuSCs but not TA MuSCs. A siRNA-mediated gene silencing study shows that Foxc1 is important for the population expansion of EOM MuSCs. Furthermore, the authors demonstrated that the EOM MuSCs contain a PDGFRβ+ve cell population that is more proliferative and less myogenic compared to a PDGFRβ-ve cell population. Altogether, this study provides new insights into the regional differences in MuSCs and will contribute to the development of stem cell-based therapies for muscle-wasting diseases including muscular dystrophies and age-related sarcopenia.

We appreciate the assessment of the Reviewer noting the new insights that our work provides on muscle stem cell heterogeneity.

Reviewer #2 (Evidence, reproducibility and clarity (Required)): Benavente-Diaz et al. address a question of why do progenitors isolated from extra ocular muscles have higher proliferative and regenerative capacity compared to progenitors from limb muscles. They perform transcriptomic profiling of the two progenitor populations and report which genes and functional groups are differentially expressed. They perform multiple bioinformatic analyses aimed at providing insights into which differentially expressed genes may be master regulators of these differences.

Major comments:

Many conclusions in this paper are based on bioinformatic analyses with little experimental data provided as a confirmation. Overall, to increase significance of the manuscript I would suggest expanding the experimental validation part of the study. By following up on the findings from the bioinformatic analyses and confirming them in vitro this manuscript could move from beyond preliminary and only potentially interesting.

The previous version of our manuscript relied on bioinformatics analysis of single cell transcriptomics and some experimental validations to investigate muscle stem cell heterogeneity. Now, we performed loss of function experiments in EOM cells and gain of function experiments in TA cells using lentiviruses to validate those points mechanistically (Figure 7E-L).

Authors suggest a number of TFs and ECM components as master regulators of progenitor identity. An experiment of a rather limited scope was provided in Figure 7E-F, where effects of silencing TF Foxc1 on EOM progenitor proliferation was assessed. It would be highly beneficial to expand these experiments to include other TFs found in their dataset. Importantly, by overexpressing the proposed master regulators in TA progenitors authors should investigate whether these TFs indeed confer higher proliferative and regenerative capacity. Otherwise, the authors should make it clear that their conclusions about TFs regulating or maintaining EOM progenitors are preliminary and based on bioinformatic analyses.

Indeed, these experiments are necessary to consolidate the bioinformatic analysis and provide further mechanistic insights on this point. We have obtained lentiviruses and optimized gain of function assays in TA cells. Gain of function experiments with Creb3l1 and Dmrta2, two other top regulon transcription factors have been planned besides the experiments already included with Foxc1.

Minor comments:

In Figure 2, authors describe single-cell RNAseq analysis of EOM and TA progenitors. They report a number of markers differentially expressed between these two populations. It would be good to perform higher-resolution subclustering of each these populations to understand whether markers are expressed in all EOM progenitors or whether there is a specific subpopulation that is characterized by Mgp/Bgn/... expression. A paper by Yartseva et al. (Cell Reports 2020) did describe a population of activated satellite cells that express extracellular matrix components. In addition, from the data presented it is not clear whether proposed EOM markers are uniquely expressed or only enriched in EOM progenitors.

Actually, In Figure 2 we are showing what the reviewer is requesting. The heatmap highlights the presence of subpopulations (Progenitor/Differentiating) in both EOM and TA. The 4-way heatmap in Figure 2D is showing the distinctly upregulated genes of each subcluster. Matrix related genes are indeed expressed by a subpopulation of EOM cells, that we identified as "Progenitors". Two of those, Bgn and Mgp, can be seen in Figure 3H. Moreover, in Figure 4B we show the Average Expression and Percentage of cells expressing certain ECM component.

Actinomycin D is commonly used in single cell preparations for RNAseq to mitigate stress gene activation due to isolation procedure (for example see Wu YE et al. Neuron 2017). It is possible that EOM progenitors are particularly responsive to the isolation procedure, which would imply that stress genes are not involved in EOM progenitor maintenance, but that their expression is an artifact of isolation. These experiments should be repeated with actinomycin D to exclude potential artifacts.

While this is a possibility, the likelihood of this is relatively low as activated cells were obtained by quick trypsinization of cells. To formally exclude the problem mentioned by the reviewer we will perform qRT-PCR for EOM markers on activated cells at Day 3 that were fixed in PFA or treated with Actinomycin D prior to trypsinization. Moreover, in Suppl Figure 5G-E, we already used the stress index calculation defined by Machado et al. 2020 and this does not seem to affect our activated dataset.

It is unclear why authors chose to focus on PdgfrB+ population in Figure 7. Was it chosen as a target of Foxc1 or as a marker of proliferative cells? Any other reason? This should be explained as well as significance of this part in connection to the rest of the article.

We decided to focus on Pdgfrb for several reasons stated in the text: 1) it is a component of the matrisome we have validated; 2) it stands out in the Reactome pathways/Molecular function analysis; 3) it is a target of Foxc1; 4) it is a marker of cells with mesenchymal characteristics. We also chose this marker, for which antibodies for FACs exist, as proof of principle validation of the EOM mesenchymal phenotype. Similarly, sc-RNAseq analysis of human muscle, used Cav1 antibodies to isolate a functionally different subpopulation of MuSCs (Barruet elife 2020). To make the flow of the manuscript more consistent this data is now compiled on Figure 4.

Please explain the significance and relevance of the results presented in Figure 8A-C.

These figures highlight the potential of EOM progenitors to specifically regulate matrisome genes with respect to the TA. The 90 top active regulons of the EOM potentially regulate more matrisome genes than the TA, and the ratio (Number of EOM presumptive regulations/ Number of TA presumptive regulations) peaks when looking at the top 5 regulons (a 3-fold difference), and progressively reduces.

Prior studies are (not) referenced appropriately

We have corrected reference duplications.

Figures contain small fonts that are not legible. In particular, networks such as the one in Fig 5C are difficult to read.

The font size, font color and/or size of networks has been changed for Fig 2E-F, 5C-D, 8G-J and Suppl Fig 2A, 7C.

General assessment: Functional differences between EOM and TA progenitors have been previously described, but a deep mechanistic understanding of the underlying molecular pathways is lacking. This manuscript takes a step towards elucidating these molecular pathways.

We now provide more mechanistic data regarding the differential role that Foxc1 plays in extraocular compared to limb myogenic cells. Our revised plan will consolidate and extend these observations.

Advance: Tajbakhsh group recently published a paper Evano et al. Plos Genetics 2020, that also explored the differences between EOM and TA stem cells. Among other things this paper showed that EOM stem cells have intrinsic molecular mechanisms/programs that differentiate them from TA stem cells. In that context, the experimental insights that Benavente-Diaz et al. provide are not novel. Benavente-Diaz et al. would extend the knowledge in the field if they experimentally confirmed that the proposed master regulators indeed determine progenitor proliferation and regeneration capacity.

We respectfully disagree regarding the novelty of our work. In our previous study cited above, we showed that EOM stem cells have in vivo transcriptional differences with those in the limb (bulk RNAseq) and they can adopt the fate of limb cells when transplanted into the limb. Interestingly, full “reprogramming” was not achieved in those experiments pointing to limited plasticity.

Here, we show by scRNAseq and bioinformatic analysis of regulons that Foxc1 and extracellular matrix signature are features that are unique to EOM stem cells. Also, our overall single cell transcriptome analysis and comparative studies show that EOM muscle stem cells have adopted a signature that overlaps with mesenchymal stem cells – a property that has to date not been reported.

Muscle is arguably the best system to investigate stem cell heterogeneity, and our study provides mechanistic insights into the extend of this diversity.

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

Summary:

Benaverte-Diaz et al. address a question of why do progenitors isolated from extra ocular muscles have higher proliferative and regenerative capacity compared to progenitors from limb muscles. They perform transcriptomic profiling of the two progenitor populations and report which genes and functional groups are differentially expressed. They perform multiple bioinformatic analyses aimed at providing insights into which differentially expressed genes may be master regulators of these differences.

Major comments:

• Many conclusions in this paper are based on bioinformatic analyses with little experimental data provided as a confirmation. Overall, to increase significance of the manuscript I would suggest expanding the experimental validation part of the study. By following up on the findings from the bioinformatic analyses and confirming them in vitro this manuscript could move from beyond preliminary and only potentially interesting.
• Authors suggest a number of TFs and ECM components as master regulators of progenitor identity. An experiment of a rather limited scope was provided in Figure 7E-F, where effects of silencing TF Foxc1 on EOM progenitor proliferation was assessed. It would be highly beneficial to expand these experiments to include other TFs found in their dataset. Importantly, by overexpressing the proposed master regulators in TA progenitors authors should investigate whether these TFs indeed confer higher proliferative and regenerative capacity. Otherwise, the authors should make it clear that their conclusions about TFs regulating or maintaining EOM progenitors are preliminary and based on bioinformatic analyses.

Minor comments:

• In Figure 2, authors describe single-cell RNAseq analysis of EOM and TA progenitors. They report a number of markers differentially expressed between these two populations. It would be good to perform higher-resolution subclustering of each these populations to understand whether markers are expressed in all EOM progenitors or whether there is a specific subpopulation that is characterized by Mgp/Bgn/... expression. A paper by Yartseva et al. (Cell Reports 2020) did describe a population of activated satellite cells that express extracellular matrix components. In addition, from the data presented it is not clear whether proposed EOM markers are uniquely expressed or only enriched in EOM progenitors.
• Actinomycin D is commonly used in single cell preparations for RNAseq to mitigate stress gene activation due to isolation procedure (for example see Wu YE et al. Neuron 2017). It is possible that EOM progenitors are particularly responsive to the isolation procedure, which would imply that stress genes are not involved in EOM progenitor maintenance, but that their expression is an artifact of isolation. These experiments should be repeated with actinomycin D to exclude potential artifacts.
• It is unclear why authors chose to focus on PdgfrB+ population in Figure 7. Was it chosen as a target of Foxc1 or as a marker of proliferative cells? Any other reason? This should be explained as well as significance of this part in connection to the rest of the article.
• Please explain the significance and relevance of the results presented in Figure 8A-C.
• Prior studies are referenced appropriately
• Figures contain small fonts that are not legible. In particular, networks such as the one in Fig 5C are difficult to read.

Significance

• General assessment: Functional differences between EOM and TA progenitors have been previously described, but a deep mechanistic understanding of the underlying molecular pathways is lacking. This manuscript takes a step towards elucidating these molecular pathways.
• Advance: Tajbakhsh group recently published a paper Evano et al. Plos Genetics 2020, that also explored the differences between EOM and TA stem cells. Among other things this paper showed that EOM stem cells have intrinsic molecular mechanisms/programs that differentiate them from TA stem cells. In that context, the experimental insights that Benaverte-Diaz et al. provide are not novel. Benaverte-Diaz et al. would extend the knowledge in the field if they experimentally confirmed that the proposed master regulators indeed determine progenitor proliferation and regeneration capacity.
• Audience: basic research, translational. This research would be of interest to regenerative therapy field.

My field of expertise: molecular and cellular biology, muscle regeneration. I am not a bioinformatician and I am not able to technically evaluate in silico approaches used in this study.

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, Girolamo et al., describe the differences in molecular signature and biological features of MuSC populations between extraocular (EOM) and limb (Tibialis anterior, TA) muscles. Comprehensive approaches including scRNA-seq, bioinformatics, and live-cell imaging reveal that EOM MuSCs is more proliferative in vitro and express ECM components at high levels as well as non-myogenic markers such as Foxc1 and Pdgfrb. The transcription factor network described in this study will characterize the identity of EOM MuSCs, providing insights into stem cell-based therapies for muscle diseases.

Major comments

1. In Figure 1A, EOM-MuSCs appear to form myotubes as efficiently as TA-MuSCs. If so, why myotube formation is not affected in EOM-MuSCs even with lower Myog expression compared to TA-MuSCs? The authors should discuss about this point.
2. In Figure 2, activated EOM-MuSCs express Myod1e at lower levels compared to TA-MuSCs. Are MyoD protein levels are also lower in EOM-MuSCs?
3. In Figure 7E-G: Does the Foxc1 knockdown change the myogenic potential, especially on myotube formation? In addition to FOXC1 and Myog, the mRNA levels of Pax7 and Myod1 would be better to be provided.
4. The authors mention that Foxc1 maintains EOM MuSCs in a progenitor-like state through matrix remodeling and cooperation with other TFs. However, such functional analysis does not seem to be provided sufficiently. For example: does Foxc1 siRNA lower matrisome and Pdgfrb genes in EOM-MuSCs; does forced expression of Foxc1in TA-MuSCs acquire the EOM-like state? Although the authors can predict the role and function of Foxc1 in the MuSC population based on the pySCENIC and previous studies, the experimental approach by the authors would be more convincing. The reviewer would like to emphasize that these experiments are not entirely necessary in this manuscript.
5. Please state a reasonable explanation for the physiological role of high amounts of ECM produced by EOM-MuSCs.

Minor comments

On p7, the 3rd line: "marker" is redundant.

On p9: what is meant by "force-directed environment"? Is this the aggregation of regulons and targets based on interaction strength determined by the algorithmic arrangement used by pySCENIC?

On p11: add space between (Sambasivan et al., 20009a)and.

On p11: 1.7 fold increase → 2.7 fold?

On p13: please describe the GRN abbreviation since it is the first used here and not clarified beforehand.

On p15: add space between (Vallejo et al., 2018)Other.

On p16: add "a" after "Klf4 is" → Klf4 is a pioneer...

In Fig1: it would be better to show the quantification of EdU incorporation, especially at D5 for highlighting the difference between EOM and TA.

In Fig1: MF20 staining seemingly describes larger myotubes in EOM compared with TA at D5. Is this most likely due to the higher number of cells to start with in EOM rather than having more fusion ability?

In Fig 2E,F: some font is very faint in color and hard to read in printed format e.g. "TP53 regulates transcription of". May want to change color. In addition, the bottom edge of the Figure is slightly cut off.

In Fig 4D: COLVI should be changed to COLIV.

In Fig 5B: Myod1 is redundant.

In Fig 6D; please specify "Amp". Is Hey1 a myogenic marker?

In Fig 7F; there are Ctrl and control- please unify them.

In Fig7C,D: When does Foxc1 start to be expressed in EOM progenitors in the embryo? If the authors tested, please mention it.

In Supp Fig 7A: graphs are lacking color-coded legends.

In Supp Fig 7C: Font is very small and illegible in a paper format.

In Supp Fig 7F: what are the dynamics of PDGFβ+ cell populations in successive passages in culture? If the authors tested please mention it.

In Supp Fig 8 A: Are the colors reversed for EOM? In Fig 8 E, EOM progenitor and differentiation stream are light blue and yellow respectively but in Supp 8A the colors are flipped and don't seem to match the Map laid out in Fig 8E.

Some references are listed as redundant.

Significance

Skeletal muscle stem cells (MuSCs) play an indispensable role in muscle regeneration in adults. MuSCs are distributed in all muscles throughout the body and their function and molecular properties are diverse among muscles. Extraocular muscles (EOMs) are known to be preferentially spared in muscular dystrophies and during aging. In addition, EOM-derived MuSCs are highly transplantable compared to those of limb muscles. However, intrinsic regulators of EOM MuSCs have not been fully characterized yet. This study by Girolamo et al. shows the differences in molecular signature and biological features of MuSC populations between EOM and limb (Tibialis anterior, TA) muscles. Comprehensive approaches including scRNA-seq, bioinformatics, and live-cell imaging revealed that a subset of the EOM-derived MuSC population is highly proliferative and expresses extracellular matrix components at high levels. The analysis also shows that EOM-derived MuSCs have non-myogenic signatures such as Foxc1 and Pdgfrb. A transcriptional factor Foxc1 is described as a pro-mitogenic factor in the cancer field and is known as a driver of endothelial/smooth muscle fate. In this study, the authors find that Foxc1 is expressed in EOM MuSCs but not TA MuSCs. A siRNA-mediated gene silencing study shows that Foxc1 is important for the population expansion of EOM MuSCs. Furthermore, the authors demonstrated that the EOM MuSCs contain a PDGFRβ+ve cell population that is more proliferative and less myogenic compared to a PDGFRβ-ve cell population. Altogether, this study provides new insights into the regional differences in MuSCs and will contribute to the development of stem cell-based therapies for muscle-wasting diseases including muscular dystrophies and age-related sarcopenia.

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

---

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

Major points:

1. Although the role of mitofusin on mitochondrial morphology has been established by others and comprehensively assessed in the present study, the authors should determine the functional outcome from the genetic manipulations on Mfn2 and Mfn1. As observed by increased glucose uptake, one could hypothesize an impairment in mitochondrial oxidative phosphorylation, leading the cells to rely uniquely or heavily on glycolysis as a fuel. Also, as mentioned by the authors in the discussion, ROS play a fundamental role in adipogenesis, and, therefore, mitochondrial ROS emission and/or cellular redox balance should also be assessed. I believe these two experiments will add insightful information to the current dataset.

__Thank you for these suggestions. Whilst we agree with the general premise of this point, unfortunately quantifying oxidative phosphorylation and ROS production with sufficient precision to detect relatively subtle changes remains very challenging. We have attempted these experiments but they require considerable optimisation (particularly using adipocytes). Preliminary studies done in MEFs (Cover letter figure 1) suggest that under some stimuli there may be higher ROS in Mfn1 and Mfn2 knock-out lines. However this preliminary data would require further optimisation and repetition in adipocytes, which is more challenging. __

For now, we have amended the Discussion to specify that these experiments are of particular interest.

Cover letter figure 1. Levels of reactive oxygen species (ROS) in mouse embryonic fibroblasts measured by flow cytometry for fluorometric dyes CellROX (total cellular ROS), D2-HDCFA (total cellular ROS), and MitoSOX (mitochondrial ROS). Levels are expressed relative to wild-type. MEFs were treated with antimycin A (or media only) for 20minutes prior to incubation with the ROS dyes, then washed three times before assayed. AntA, Antimycin; CR, CellROX; M1, Mfn1-/- MEFs; M2, Mfn2-/- MEFs; MS, MitoSOX; WT, wild-type.

The insulin effect on glucose uptake does not allow to conclude any impairment in insulin responsivity. The fold change of glucose uptake mediated by insulin was roughly 1.2 in undifferentiated adipocytes, 2.3 in differentiated WT, and 2.5 in Mfn1KO differentiated adipocytes. The absolute increase in glucose uptake could be a compensatory mechanism due to impairment in mitochondrial bioenergetics (see point #1), given that the cells can still respond to insulin. Measuring Akt phosphorylation levels following insulin treatment would help solve this issue.

__As requested, we have assessed the effect of insulin treatment on Ser 473 phosphorylation of Akt2 (Pkb) in wild-type and knock-out MEFs differentiated into adipocytes (Fig 2D). Mfn1_-/-_ MEFs show an increase in Akt phosphorylation relative to the other cell lines. They also have higher expression of insulin receptor and Glut4, consistent with their degree of adipogenic differentiation. __

We agree that impaired mitochondrial bioenergetics could account for the observations in perturbed glucose uptake in the knockout cell lines (especially Mfn2-null) and have therefore amended the text throughout to reflect this.

Usually, working with clonal transgenic cells lines has the limitations that the cells might behave differently in terms of adipogenic potential over passages. A transient loss of function in the same cells would solve this concern. Also, introducing the patient mutations might be closer to the human situation than working with KO mouse fibroblasts.

__We agree with this potential concern, which is why we conducted knock-down studies in 3T3-L1 cells in addition to the work in knockout MEFs. These data were concordant with what we observed in the KO MEFs so we don’t think it is necessary to conduct repeat KD experiments in WT MEFs. __

In our previous study we observed that human fibroblasts with biallelic MFN2-R707W mutations did not have any obvious phenotype (____https://elifesciences.org/articles/23813____). We have separate work studying these mutations in vivo where we provide further characterisation of murine adipocytes harbouring Mfn2-R707W; this work is now published here: https://elifesciences.org/articles/82283

Minor points:

1. Although the authors mention in the introduction that the differentiation of adipocytes is followed by an increase in mitochondrial mass, it would be interesting the determine the expression profile of mfn1 and mfn2 during the differentiation process.

We have found that there is an increase in markers of mitochondrial fusion (Mfn1 & Mfn2) as well as fission (Fis1) throughout differentiation of 3T3-L1s. ____We have included this data in the manuscript (Supplementary Figure ____6A ).

The authors should discuss other models, even though pre-clinical, of mitochondrial dysfunction that results in lipodystrophy but with different metabolic outcomes. To cite a few but not only PMID: 29588285; PMID: 21368114; PMID: 31925461.

Thank you for this suggestion. We have added a section on this in the introduction.

It would be interesting to discuss the role of Mfn1/2 in the context of cold-induced adipogenesis, given the prominent role of mitochondrial dynamics, as mentioned by the authors in the reference list, on cold-induced adaptative thermogenesis (Mahdaviane et al. 2017; Boutant et al. 2017).

Thank you for this suggestion. We have added a section on this in the introduction.

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

• In Fig.2A, the authors report "increased lipid accumulation in Mfn1-/- MEFs, but not in Mfn2-/- MEFs". While the overall content might be similar, the pattern of lipid accumulation seems to be different. Indeed, differences in lipid droplet morphology have been observed in Mfn2 KO MEFs upon oleate treatment (McFie et al., 2016). The manuscript would benefit from having quantifications of lipid droplet size and number.

Thank you for highlighting this. We have quantified lipid droplet size and, consistent with McFie et al have found increased size in Mfn2 knock-down. This data is now included in Supplementary Figure 6B.

• Following the above point, McFie et al. also reported that Mfn1/Mfn2 double KO MEFs could differentiate into adipocytes. The authors should discuss these opposing observations. The contrasting observation may be due to acquired clonal differences in MEF lines. We have attempted ‘double’ knock down (of both Mfn1 and Mfn2 concurrently) in 3T3-L1 cells however this was essentially lethal and also did not generate any cells capable of differentiation. We have added a section in the Discussion regarding this point.

• In relation to the effects of Mitofusin deletions on glucose uptake, the authors mention that Mfn2 KO MEFs show impaired insulin stimulated glucose uptake. The interpretation of the result is not straight forward, as basal glucose uptake is highly increased in Mfn2 KO MEFs. Maybe there is simply a treshold for maximal glucose uptake capacity in MEF-derived adipocytes. In any of these cases, the authors might want to check GLUT1 levels, in line of their suggestion that the increased basal glucose uptake might be related to higher GLUT1. Alternatively, the authors might also want to check elements of the insulin signaling path, in case there are alterations that could explain the phenomenon.

As mentioned above in response to reviewer 1, we have now ____performed immunoblots to quantify some components of the insulin signalling cascade (Fig 2D). We observed lower expression of both Glut1 and Glut4 in the Mfn2-/- cells. Mfn2-/- cells did demonstrate some Akt phosphorylation but considerably less than Mfn1-/- cells. These results are now included in the revised manuscript (Figure 2D).

• In line with the above point, one would have wished that mitochondrial biology was better characterized in the different MEF models. While mitochondrial shape analyses are provided, some information on, at least, mitochondrial respiratory capacity, glucose oxidation and/or fatty acid oxidation rates, would be important. This would allow for a more solid discussion on why Mfn2 KO MEFs display such high basal glucose uptake rates.

We have responded to a similar suggestion from Reviewer 1, above.

• In relation to the experiments in MEFs, one should never forget that WT, Mfn1 and Mfn2 KO MEFs derive from different mice. Hence, the phenotypes could be related to trait variabilities in the origin mice themselves, and not just the gene deletion. To control for this aspect, the authors could simply re-introduced Mfn1 or Mfn2 in their respective MEFs and evaluate if their alterations are normalized.

__Yes one could try this but we have addressed this general concern by replicating the impact of Mfn1/2 KD in 3T3L1 cells so are not inclined to pursue this at this time. __

• Transcriptomic analyses reveals a decrease in adipogenic gene expression in Mfn2 KO MEFs. However, lipid accumulation is comparable to WT MEFs is normal. This could be due to defects in lipolytic capacity, leading to similar lipid accumulation despite lower adipogenic capacity. This could be tested by evaluating the adrenergic response of these cells (e.g.: glycerol release).

Thank you for this suggestion. We have commented in the Discussion to explain that we have not fully characterised this mechanism.

• The experiments in 3T3-L1 would also benefit from some gene expression analyses to evaluate if Mfn1 depletion leads to acceleration and/or magnification of the differentiation stages. In relation to this, 3T3-L1 cells could be used to monitor Mfn1 and Mfn2 through differentiation, which in itself would be valuable information.

We have performed a protein-level time course for markers of mitochondrial fusion (Mfn1 & Mfn2) as well as fission (Fis1) throughout differentiation of 3T3-L1s. We have included this data in the manuscript (Supplementary Figure 6A). We think that changes in protein expression are more relevant than changes in mRNA so have not included gene expression changes at this time.

CROSS-CONSULTATION COMMENTS The comments from the three independent reviewers are extremely well aligned and agree that improving the following aspects could largely benefit the manuscript:

• A better metabolic characterisation of the models used
• Provide measurements in relation to mitochondrial bioenergetics and ROS production – we have attempted this but the data is not very clear in our view and warrants further optimisation which we are not inclined to pursue currently. - Explorations of insulin signaling - done thank-you.
• Improve the validation and significance of the cellular models used, following the different suggestions from the three reviewers. Most notably, considering the introduction of human Mfn2 mutation forms – we have published a separate manuscript on follow up work on the human MFN2 variant as mentioned above.

A number of additional comments are raised, all of which are very reasonable and, in my opinion, should not be difficult to address. I think we can all agree that a mechanistic underpinning of the observations would give a larger degree of novelty to the work. Also, none of us would like the revision's quality to be constraint by a tight deadline. I would therefore be totally OK to extend the timeframe for the revision beyond the original 3 months proposed.

Reviewer #2 (Significance (Required)):

This is an interesting and well-crafted manuscript. Mice deficient for Mfn2 or Mfn1 have been reported by different laboratories, yet most of them fail to explore the effects on early adipogenesis. The study is limited to cultured cells, but this is well acknowledged by the authors Given the existence of human mutations in the mitofusin-2 gene that largely alter fat mass distribution, this work provides new clues on how these mutations might impact adipose tissue.

Reviewer #3 (Evidence, reproducibility and clarity (Required)):

Mann et al. The objective of this study is to determine the extent to which mitofusins (Mfn1 and Mfn2) have redundant functions and assess their contributions to adipocyte differentiation. While a point mutation in the Mfn2 gene has been associated with severe adipose tissue dysfunction and lipodystrophy, no disease phenotypes have been linked to mutations in Mfn1. To address these objectives, the authors sought to characterize how adipocyte differentiation and function is affected in Mfn1, Mfn2 or double knockout adipocytes in two distinct in vitro models. Their findings indicate divergent effects of Mfn1 and Mfn2 on adipocyte differentiation and function despite similar alterations to mitochondrial morphology. Loss of Mfn1 promotes adipogenesis while Mfn2 decreases it. The authors conclude that these findings are indicative of non-redundant functions in Mfn1 and Mfn2.

Major comments: The observation that Mfn1 KO/KD leads to increased adipogenesis in vitro is somehow novel and, perhaps, surprising, as the author say. However, the molecular understanding underlying this phenotype remains unexplored. The analyses performed are mainly descriptive and don't dig deeper into the identification of the molecular mechanism. They do hypothesize that ROS production may be responsible for the observed effects, but that's how far they go.

The authors do highlight the limitations of this work, but these limitations need careful consideration, for not addressing them seriously limits the novelty of this study, especially not testing these conditions in human cells. The current version of this work seems too preliminary to suggest useful experiments that could strengthen the study, since future analyses could take many different directions.

Yes, we accept that the findings are rather preliminary but our initial efforts suggest that precisely elucidating the underlying mechanism/s is likely to be more difficult and complicated than alluded to by the reviewers. We would therefor prefer to share our initial observations so that others can also attempt to clarify the underlying mechanisms.

A few unanswered questions that the authors might consider are: What is the difference between the Arg707Trp mutation and the KO/KD? Mfn1 and 2 deletions lead to fragmented mitochondria, but opposite adipogenic potentials. What other mitochondrial defects can explain it? Are organelle contact site disrupted only with Mfn2? How does Mfn1 and 2 KO/KD affect mitochondrial proteome? What does mitochondrial bioenergetics look like? How is ROS production affected? Is the increased glucose uptake (basal) a compensatory mechanism for mitochondrial dysfunction? Thank you for these suggestions. We acknowledge that this work is largely descriptive in nature. These are all questions that should be addressed to improve mechanistic understanding of our observations.

__The difference between p.Arg707Trp and KO/KD is challenging to address because in the non-adipose cell lines studied so far (human and mouse fibroblasts) there has been no evidence of perturbation of the mitochondrial network. __

As discussed above, we have done preliminary studies into ROS production but are unable to provide a complete characterisation at this time. Similarly, we have not been able to perform bioenergetic studies (e.g. Seahorse, Oxyboros) that would provide more insight into differences between Mfn1 and Mfn2 KO cell lines.

CROSS-CONSULTATION COMMENTS I agree the work is interesting, but is too preliminary and merely descriptive. the experiments suggested will significantly improve the manuscript. However, I don't think they will take only three months to be completed. This work needs a significant amount of work including the study of the mechanism, at least an idea of what the mechanism could be, to be considered novel.

We accept this limitation and have responded to this general point above.

Reviewer #3 (Significance (Required)):

Understanding how mitochondrial dynamics affect adipogenic differentiation is critical to better understand how metabolism impact cell signaling, cell fate and function.

Strengths: this work reveals an interesting phenotype for Mfn1 and Mfn2 mutant preadipocytes. Weaknesses: this work is merely descriptive and preliminary to provide a clear understanding of the observed phenotypes

Advance: Although the performed experiments are accurate, well designed, and well controlled, the fact that Mfn1 and 2 have distinct functions and cannot compensate for one another was already clear based on the embryonic lethality of either Mfn1 and Mfn2 KO mice as well as the Mfn2 mutation in humans that leads to a pathological condition.In the current verison, this work minimally contributes to advancing the field.

Audience: an extensively revised version of this work including deeper phenotyping of thier models and human cell work would be of interest for sceintists studying mitchondrial biology, adipose tissue, metabolic diseases, and human genetic diseases.

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 #3

Evidence, reproducibility and clarity

Mann et al. The objective of this study is to determine the extent to which mitofusins (Mfn1 and Mfn2) have redundant functions and assess their contributions to adipocyte differentiation. While a point mutation in the Mfn2 gene has been associated with severe adipose tissue dysfunction and lipodystrophy, no disease phenotypes have been linked to mutations in Mfn1. To address these objectives, the authors sought to characterize how adipocyte differentiation and function is affected in Mfn1, Mfn2 or double knockout adipocytes in two distinct in vitro models. Their findings indicate divergent effects of Mfn1 and Mfn2 on adipocyte differentiation and function despite similar alterations to mitochondrial morphology. Loss of Mfn1 promotes adipogenesis while Mfn2 decreases it. The authors conclude that these findings are indicative of non-redundant functions in Mfn1 and Mfn2.

Major comments:

The observation that Mfn1 KO/KD leads to increased adipogenesis in vitro is somehow novel and, perhaps, surprising, as the author say. However, the molecular understanding underlying this phenotype remains unexplored. The analyses performed are mainly descriptive and don't dig deeper into the identification of the molecular mechanism. They do hypothesize that ROS production may be responsible for the observed effects, but that's how far they go.

The authors do highlight the limitations of this work, but these limitations need careful consideration, for not addressing them seriously limits the novelty of this study, especially not testing these conditions in human cells.<br /> The current version of this work seems too preliminary to suggest useful experiments that could strengthen the study, since future analyses could take many different directions. A few unanswered questions that the authors might consider are: What is the difference between the Arg707Trp mutation and the KO/KD? Mfn1 and 2 deletions lead to fragmented mitochondria, but opposite adipogenic potentials. What other mitochondrial defects can explain it? Are organelle contact site disrupted only with Mfn2? How does Mfn1 and 2 KO/KD affect mitochondrial proteome? What does mitochondrial bioenergetics look like? How is ROS production affected? Is the increased glucose uptake (basal) a compensatory mechanism for mitochondrial dysfunction?

Referees cross-commenting

I agree the work is interesting, but is too preliminary and merely descriptive. the experiments suggested will significantly improve the manuscript. However, I don't think they will take only three months to be completed. This work needs a significant amount of work including the study of the mechanism, at least an idea of what the mechanism could be, to be considered novel.

Significance

Understanding how mitochondrial dynamics affect adipogenic differentiation is critical to better understand how metabolism impact cell signaling, cell fate and function.

Strenghts: this work reveals an interesting phenotype for Mfn1 and Mfn2 mutant preadipocytes. Weaknesses: this work is merely descriptive and preliminary to provide a clear understanding of the observed phenotypes

Advance: Although the performed experiments are accurate, well designed, and well controlled, the fact that Mfn1 and 2 have distinct functions and cannot compensate for one another was already clear based on the embryonic lethality of either Mfn1 and Mfn2 KO mice as well as the Mfn2 mutation in humans that leads to a pathological condition.In the current verison, this work minimally contributes to advancing the field.

Audience: an extensively revised version of this work including deeper phenotyping of thier models and human cell work would be of interest for sceintists studying mitchondrial biology, adipose tissue, metabolic diseases, and human genetic diseases.

Reviewer expertise: adipose tissue function, metabolic disorders, mitochondrial bioenergetics.

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

The work by Mann and colleagues explores the adipogenic potential of MEF cells derived from Mfn1, Mfn2, Mfn1/Mfn2 and OPA1 KO mice. Two of these cell lines (Mfn1/Mfn2 KO MEFs and OPA1 MEFs) failed to differentiate, so only Mfn1 and Mfn2 were considered for most of the work. The experiments revealed that Mfn1 deletion lead to a faster and more prominent differentation of MEFs into adipocytes, which did not occur on Mfn2 KO MEFs. In contrast Mfn2 KO MEFs showed some signs of impaired adipogenesis, including lower GLUT4 gene expression and reduced insulin-stimulated glucose transport. Most of these observations were verified using a second cell model, in which Mfn1 or Mfn2 were knocked-down in 3T3-L1 adipocytes. This led the authors to conclude that Mfn1, but not Mfn2, enhances adipogenesis.

The manuscript is very well written and the experiments are proficiently designed. One might have wished confirmation of these findings in primary adipocyte cultures or in model organisms, but this limitation is duly acknowledged by the authors. The methods are well described and should allow other labs to easily reproduce the experiment. A few suggestions that could improve the manuscript can be found below.

• In Fig.2A, the authors report "increased lipid accumulation in Mfn1-/- MEFs, but not in Mfn2-/- MEFs". While the overall content might be similar, the pattern of lipid accumulation seems to be different. Indeed, differences in lipid droplet morphology have been observed in Mfn2 KO MEFs upon oleate treatment (McFie et al., 2016). The manuscript would benefit from having quantifications of lipid droplet size and number.
• Following the above point, McFie et al. also reported that Mfn1/Mfn2 double KO MEFs could differentiate into adipocytes. The authors should discuss these opposing observations.
• In relation to the effects of Mitofusin deletions on glucose uptake, the authors mention that Mfn2 KO MEFs show impaired insulin stimulated glucose uptake. The interpretation of the result is not straight forward, as basal glucose uptake is highly increased in Mfn2 KO MEFs. Maybe there is simply a treshold for maximal glucose uptake capacity in MEF-derived adipocytes. In any of these cases, the authors might want to check GLUT1 levels, in line of their suggestion that the increased basal glucose uptake might be related to higher GLUT1. Alternatively, the authors might also want to check elements of the insulin signaling path, in case there are alterations that could explain the phenomenon.
• In line with the above point, one would have wished that mitochondrial biology was better characterized in the different MEF models. While mitochondrial shape analyses are provided, some information on, at least, mitochondrial respiratory capacity, glucose oxidation and/or fatty acid oxidation rates, would be important. This would allow for a more solid discussion on why Mfn2 KO MEFs display such high basal glucose uptake rates.
• In relation to the experiments in MEFs, one should never forget that WT, Mfn1 and Mfn2 KO MEFs derive from different mice. Hence, the phenotypes could be related to trait variabilities in the origin mice themselves, and not just the gene deletion. To control for this aspect, the authors could simply re-introduced Mfn1 or Mfn2 in their respective MEFs and evaluate if their alterations are normalized.
• Transcriptomic analyses reveals a decrease in adipogenic gene expression in Mfn2 KO MEFs. However, lipid accumulation is comparable to WT MEFs is normal. This could be due to defects in lipolytic capacity, leading to similar lipid accumulation despite lower adipogenic capacity. This could be tested by evaluating the adrenergic response of these cells (e.g.: glycerol release).
• The experiments in 3T3-L1 would also benefit from some gene expression analyses to evaluate if Mfn1 depletion leads to acceleration and/or magnification of the differentiation stages. In relation to this, 3T3-L1 cells could be used to monitor Mfn1 and Mfn2 through differentiation, which in itself would be valuable information.

Referees cross-commenting

The comments from the three independent reviewers are extremely well aligned and agree that improving the following aspects could largely benefit the manuscript:

• A better metabolic characterisation of the models used
• Provide measurements in relation to mitochondrial bioenergetics and ROS production
• Explorations of insulin signaling
• Improve the validation and significance of the cellular models used, following the different suggestions from the three reviewers. Most notably, considering the introduction of human Mfn2 mutation forms

A number of additional comments are raised, all of which are very reasonable and, in my opinion, should not be difficult to address. I think we can all agree that a mechanistic underpinning of the observations would give a larger degree of novelty to the work. Also, none of us would like the revision's quality to be constraint by a tight deadline. I would therefore be totally OK to extend the timeframe for the revision beyond the original 3 months proposed.

Significance

This is an interesting and well-crafted manuscript. Mice deficient for Mfn2 or Mfn1 have been reported by different laboratories, yet most of them fail to explore the effects on early adipogenesis. The study is limited to cultured cells, but this is well acknowledged by the authors Given the existence of human mutations in the mitofusin-2 gene that largely alter fat mass distribution, this work provides new clues on how these mutations might impact adipose tissue.

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

Mann et al. described the effects of Mfn1 and Mfn2 deletion on adipogenesis. The authors describe a surprisingly pro-adipogenic effect of Mfn1 deletion despite massive mitochondrial fragmentation whilst conversely, loss of Mfn2 led to mitochondria fragmentation and impairment of adipogenesis. Overall, the research is well-designed and properly presented. Besides the lack of in vivo data (which is difficult due to the lack of a specific preadipocyte Cre), as acknowledged by the authors as a limitation, the study would benefit from a few experimental data in order to make the conclusions more robust. Please, find my comments in a point-by-point manner, which I hope will be useful for the authors.

Major points:

1. Although the role of mitofusin on mitochondrial morphology has been established by others and comprehensively assessed in the present study, the authors should determine the functional outcome from the genetic manipulations on Mfn2 and Mfn1. As observed by increased glucose uptake, one could hypothesize an impairment in mitochondrial oxidative phosphorylation, leading the cells to rely uniquely or heavily on glycolysis as a fuel. Also, as mentioned by the authors in the discussion, ROS play a fundamental role in adipogenesis, and, therefore, mitochondrial ROS emission and/or cellular redox balance should also be assessed. I believe these two experiments will add insightful information to the current dataset.
2. The insulin effect on glucose uptake does not allow to conclude any impairment in insulin responsivity. The fold change of glucose uptake mediated by insulin was roughly 1.2 in undifferentiated adipocytes, 2.3 in differentiated WT, and 2.5 in Mfn1KO differentiated adipocytes. The absolute increase in glucose uptake could be a compensatory mechanism due to impairment in mitochondrial bioenergetics (see point #1), given that the cells can still respond to insulin. Measuring Akt phosphorylation levels following insulin treatment would help solve this issue.
3. Usually, working with clonal transgenic cells lines has the limitations that the cells might behave differently in terms of adipogenic potential over passages. A transient loss of function in the same cells would solve this concern. Also, introducing the patient mutations might be closer to the human situation than working with KO mouse fibroblasts.

Minor points:

1. Although the authors mention in the introduction that the differentiation of adipocytes is followed by an increase in mitochondrial mass, it would be interesting the determine the expression profile of mfn1 and mfn2 during the differentiation process.
2. The authors should discuss other models, even though pre-clinical, of mitochondrial dysfunction that results in lipodystrophy but with different metabolic outcomes. To cite a few but not only PMID: 29588285; PMID: 21368114; PMID: 31925461.
3. It would be interesting to discuss the role of Mfn1/2 in the context of cold-induced adipogenesis, given the prominent role of mitochondrial dynamics, as mentioned by the authors in the reference list, on cold-induced adaptative thermogenesis (Mahdaviane et al. 2017; Boutant et al. 2017).

Referees cross-commenting

I agree with the statement of reviewer #2. I agree with reviewer #3, this is not the first paper on Mfns in adipocytes, so the novelty is limited but TMO sufficient for publication. Also, I tend to first look at what is there, not what is not there, and to my opinion, based on quality control measures, this work has merit.

Significance

See above

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

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

Reply to reviewers.

We deeply thank the reviewers for the time spent on evaluating our manuscript as well as providing comments and suggestions to improve our study.

__Reviewer #1 (Evidence, reproducibility and clarity (Required)): __

*In this manuscript Lebdy et al. describe a new role of GNL3 in DNA replication. They show that GNL3 controls replication fork stability in response to replication stress and they propose this is due to the regulation of ORC2 and the licensing of origins of replication. Their data suggest that GNL3 regulates the sub nuclear localization of ORC2 to limit the number of licensed origins of replication and to prevent resection of DNA at stalled forks in the presence of replication stress.

While many of the points of the manuscript are proven and well supported by the results, there are some experiments that could improve the quality and impact of the manuscript. The main issue is that the connection between the role of GNL3 in controlling ORC2, the firing of new origins and the protection of replication forks is not clearly established. At the moment the model relies on mainly correlative data. In order to further substantiate the model, we propose to address some of the following issues:*

1. *The authors indicate that RPA and RAD51 accumulation at stalled forks is not affected by GNL3 depletion. These data should be included and other proteins should be analysed. In addition, the role of helicases could be explored through the depletion of the main helicases involved in the remodelling of the forks. * Response: As asked by the reviewer we will add the fractionation experiments that show that the level of RAD51 and RPA on chromatin is not affected by GNL3 depletion. So far, the other proteins we checked (RIF1 and BRCA1), both involved in nascent strand protection, did not show clear differences. Therefore, we concluded that depletion of GNL3 does not seem to affect the recruitment of major proteins required for protection of nascent DNA. Of course, we cannot exclude that other proteins may be affected by GNL3 depletion, but testing all the possible candidates would be time consuming with a very low chance of success. In addition, fractionation experiments are possibly not quantitative enough to uncover small differences and may be not that informative. Thus it remains possible that RPA exhaustion may be the cause of resection in absence of GNL3 as suggested by the work conducted in Lukas’ lab (Toledo et al. 2013. https://pubmed.ncbi.nlm.nih.gov/24267891/). To test this hypothesis, we will analyze if resection in absence of GNL3 is still occurring in a well-characterized cell line that overexpress the three RPA subunits that we obtained from Lukas’ lab.

To our knowledge not many helicases have been shown to be involved in remodeling of stalled forks. The best example is RECQ1, however we feel that testing RECQ1 involvement in resection upon GNL3 depletion will complicate our story without adding much regarding the mechanism. We hope the reviewer understands our concern.

• The proposed model implies that GNL3 depletion leads to increased origin licensing. FThe authors should address if the primary effect of GNL3 depletion is on origin firing by using CDC7 inhibition in the absence of stress (Rodríguez-Acebes et al., JBC 2018). *

__Response: __This is an excellent point raised by the reviewer. To test if the primary effect of GNL3 depletion in on origin firing we will test if the defect in replication fork progression is dependent on CDC7 using DNA fibers experiments and CDC7 inhibitor.

• A way to prove that origin firing mediates the effect of GNL3 on fork protection would be to reduce the number of available origins. The depletion of MCM complexes has been shown to limit the number of back-up origins that are licensed and leads to sensitivity to replication stress (Ibarra et al., PNAS 2008). If GNL3 depletion results in increased number of origins, this effect should be prevented by the partial depletion of MCM complexes. *

__Response: __This is also an excellent point. We will test if MCM depletion decreases resection upon GNL3 depletion and treatment with HU. In addition, we will integrate in the manuscript experiments that we have done recently that show that treatment with roscovitine, a CDK inhibitor that impairs origin firing, decreases the level of resection observed in absence of GNL3. We think this experiment strengthens the results obtained with CDC7 inhibitors.

*Alternatively, the authors could try to modulate the depletion of GNL3. Origin licensing takes place in the G1 phase and thus the depletion of GNL3 by siRNA could affect the following S phase. Using an inducible degron for GNL3 depletion would allow to deplete GNL3 in G1 or S phase specifically. If the model is correct, the removal of GNL3 in S phase should not affect fork protection but removing GNL3 in the previous G2/M phase should reduce the number of licensed origins and lead to impaired fork protection. *

__Response: __This is obviously a good point given the fact that GNL3 deletion is not viable (see responses to reviewer 2). We tried to develop an auxin induced degron of GNL3, but we could not obtain homozygous clones, meaning that our clones had always an untagged GNL3 allele. Since GNL3 is essential its tagging may impair its function, explaining why we could not obtain homozygous clones. However, we are planning to optimize the design using other degrons system (for instance Halo-tag) to address the role of GNL3 specifically during S-phase. But we think this is above the scope of the present study.

*In addition to the connection GNL3-origin firing-fork protection, it is unclear how the lack of GNL3 in the nucleolus and the change in the sub nuclear localization of ORC2 controls origin firing and resection. The strong interaction observed between GNL3-dB and ORC2, and the subsequent change in ORC2 localization does not explain how origin licensing can be affected. In this sense, the authors could address: *

1. *Does the depletion of GNL3 and the expression of GNL3-dB affect the formation of the ORC complex, its subnuclear localization or its binding to chromatin? The authors have not explored if the interaction of GNL3 with ORC2 is established in the context of the ORC complex. An IF showing NOP1 with PLA data from GNL3-dB and ORC2 is needed to analyse how the expression of increasing amounts of GNL3-dB affects ORC2. * __Response: __We tested if GNL3 depletion impacts ORC2 and ORC1 recruitment on chromatin, but we could not observe significant differences. No clear differences were observed upon GNL3-dB expression either. One reason for this may be due to the excess of ORC complex on the chromatin, in addition chromatin fractionation is likely not sensitive enough to observe small differences. We think that quantitative ChIP-seq of ORC2 or other ORC subunits upon GNL3 depletion is required to visualize such differences, but this is above the scope of the study, and this constitutes the following of this project. We also tried to look at subnuclear localization of ORC2 using immunofluorescence, but the signal was not specific enough to observe differences. We think that the increased interaction (PLA) of ORC2 with GNL3-dB (Figure 5E) demonstrates a change in ORC2 subnuclear localization. To confirm this, we will perform the excellent experiment proposed by the reviewer to test if increasing level of GNL3-dB affects its interaction with ORC2 using PLA.

We do not think that the interaction between ORC2 and GNL3 is established in the context of the ORC complex since only ORC2 (and not the other ORC) was significantly enriched in the GNL3 Bio-ID experiment. The full list of proteins from the Bio-ID experiment (Figure 4A) will be provided in the revised version. Therefore, we think that either GNL3 regulates ORC2 subnuclear localization that in turns impact the ORC complex or GNL3 regulates ORC2-specific functions. More and more evidences show that ORC2 plays roles possibly independently of the ORC complex (see Huang et al. 2016 https://doi.org/10.1016/j.celrep.2016.02.091 or Richards et al. 2022 https://doi.org/10.1016/j.celrep.2022.111590 for instance). Future work should uncover how these ORC2 functions may regulate origins activity.

*In order to confirm if the mislocalization of ORC2 by the expression of GNL3-dB increases origin firing and mediates the effects on fork protection the authors could check DNA resection levels inhibiting CDC7 in high GNL3-dB conditions. Also, the levels of MCM2, phosphor-MCM2, CDC45, have not been analysed upon expression of GNL3-dB. *

__Response: __This is a good point; we will test if the resection observed upon expression of GNL3-dB is dependent on origin firing using CDC7 inhibitor. We have not measured the level of the cited proteins but instead we performed DNA combing to measure Global Instant Fork Density. We now show that expression of GNL3-WT suppresses the increased origin firing observed upon GNL3 depletion, in contrast expression of GNL3-dB does not suppress it. This important result indicates that origin firing is increased upon GNL3-dB expression, providing a link between aberrant localization and increased firing. These data will be part of the revised version of the manuscript.

The data in the paper suggest that GNL3 may affect the role of ORC2 in centromeres. Since depletion of GNL3 leads to increased levels of gH2AX, it would be interesting to address if this damage is due to incomplete replication in centromeres by analysing the co-localization of g*H2AX and centromeric markers both in unstressed conditions and upon the induction of replication stress. *

__Response: __This is indeed and interesting comment, however since it has been previously shown that gH2AX signal is rather strong upon GNL3 depletion (see Lin et al. 2013. https://pubmed.ncbi.nlm.nih.gov/24610951/ ; Meng et al. 2013. https://pubmed.ncbi.nlm.nih.gov/23798389/) we do not think that co-localization experiments with CENP-A for instance will be informative given the high number of gH2AX foci.

*Minor points: *

1. In the initial esiRNA screen the basal levels of g*H2AX should also be shown. * Response: Our negative control is the transfection of an esiRNAs that targets EGFP (a gene that is not expressed in the tested cell line). This esiRNAs is ranked at the end of the list and therefore constitutes the basal level of gH2AX signal. In any case it is well-established that GNL3 depletion increases gH2AX signal (see Lin et al. 2013. https://pubmed.ncbi.nlm.nih.gov/24610951/ ; Meng et al. 2013. https://pubmed.ncbi.nlm.nih.gov/23798389/).

*Figure EV1B: I think the rank needs another RS mark to see better the effect of each esiRNA on DNA lesions (high variability in all the conditions showed). *

__Response: __We understand this issue, but we cannot repeat this set of experiments for technical reasons (reagents and cost mainly). Anyway, we believe that the role of GNL3 is response to replication stress is extensively addressed by other experiments of this manuscript.

*Figure 1C and Figure EV1D/E: the quantification of the pCHK1/CHK1 levels could be included to show that there are no changes in phosphorylation upon GNL3 depletion. *

Response: it is a good point; we will put quantification in the revised version.

*In the first section of the results, at the end Figure 4B is incorrectly called for. *

__Response: __Thanks for the comment, we will modify accordingly.

The levels of GLN3 expression in 293 cells should be already included in section GNL3 interacts with ORC2.

__Response: __We will add a figure that shows the level of expression in 293 cells.

The full MS data needs to be included for both GNL3 and ORC2.

__Response: __This will be integrated in the revised version.

Figure 4B should be improved, since there is a faint band in the IgG mouse control.

__Response: __it is true that the figure is not perfect, but we believed that our Bio-ID and PLA experiments fully demonstrate the interaction between GNL3 and ORC2.

__Reviewer #1 (Significance (Required)): __

*The work is nicely written, the figures are well presented and the experiments have the necessary controls. It provides relevant information to understand how replication stress is controlled and linked to replication fork protection through origin firing. These results are relevant to the field, linking GNL3 to origin firing and with potential to help understand the role of GNL3 in cancer. They provide new information and can give rise to new studies in the future. Many of the conclusions of the manuscript are well supported. Additional support for some of the main claims would strengthen the results and also increase the impact providing a bigger conceptual advance by performing some of the suggested experiments. *

__Reviewer #2 (Evidence, reproducibility and clarity (Required)): __

*This manuscript explores the role of GNL3/nucleostemin in DNA replication and specifically in the response of DNA replication to DNA damage. GNL3 is a predominantly nucleolar protein, previously characterised as a GTP-binding protein and shown to be necessary for effective recruitment of the RAD51 recombinase to DNA breaks. The entry point for this report is a mini screen, based on proteins identified previously by the authors to associate with replication forks by iPOND, for factors that increase gamma-H2Ax (an indicator of DNA damage) after treatment with the Top1 inhibitor camptothecin (CPT). In this mini-screen GNL3 emerged as the top hit.

The authors put forward the hypothesis that GNL3 is able to sequester the replication licensing factor ORC2 in the nucleolus and that failure of this mechanism leads to excessive origin firing and DNA resection following CPT treatment.*

• The model put forward is interesting, but currently rather confusing. However, for the reasons upon which I expand below, I do not believe that the data provide a compelling mechanistic explanation for the effects that are reported and I am left not being certain about some of the links that are made between the various parts of the study, even though individual observations appear to be of good quality. *

*Specific points: *

*The knockdown of GNL3 is very incomplete. In this regard, the complementation experiments are welcome and important. However, is it an essential protein? Can it be simply deleted with CRISPR-Cas9?

*__Response: __There are obviously variations between experiments but overall, the depletion of GNL3 using siRNA seems good in our opinion. Deletion of GNL3/nucleostemin leads to embryonic lethality in mouse (Beekman et al. 2006. https://pubmed.ncbi.nlm.nih.gov/17000755/ ; Zhu et al. 2006. https://pubmed.ncbi.nlm.nih.gov/17000763/). ES cells deleted for GNL3 can be obtain but do not proliferate probably because of their inability to enter in S-phase (Beekman et al. 2006. https://pubmed.ncbi.nlm.nih.gov/17000755/). We wanted to test if it was the case in our cellular model and we tried to delete it using CRISPR-Cas9. We managed to obtain few clones deleted for GNL3, but they grow really poorly prevented us to do experiments. To bypass this, and as suggested by the reviewer 1, we tried to make an auxin-induced degron of GNL3. Unfortunately, we did not manage to obtain homozygous clones, only heterozygous. One possibility could be that the tagging induced a partial loss of function of GNL3, and since GNL3 is essential, it may explain why we did not obtain homozygous clones. We may also want to use alternative degron systems such as Halo-Tag, but we believe this is out of the scope of the study.

__ __*Global instant fork density is not quite the same as actually measuring origin firing. Ideally, it would be good to see some more direct evidence of addition origin firing e.g. by EdU-seq (Macheret & Halazonetis Nature 2018) but this would be quite a significant additional undertaking. However, given the authors have performed DNA combing with DNA counterstain, they should be able to provide accurate measurements of origin density and inter-origin distance. *

__Response: __As indicated by the reviewer EdU-seq would need a lot of development since we are not using this approach in our team. In addition, this method can detect replication origins only if performed in the beginning of S-phase, meaning that only the early firing origins will be detected and not the others. GIFD measurement is actually directly linked with origin firing since it is counting the forks to duplicate the genome. The measurements of IODs have at least two main limitations: (1) there is a bias for short IODs due to the length of analyzed fibers and (2) it focuses only on origins within a cluster not globally. Overall, we believe that GIFD is the method of choice to measures origins firing. In addition, these experiments have been done by the lab of Etienne Schwob (see acknowledgments), a leader in the field.

*'Replication stress' is induced with CPT. This term is frequently used to describe events that lead to helicase-polymerase uncoupling (e.g. O'Connor Mol Cell 2015) but that is not the case with CPT, which causes fork collapse and breaks. Are similar effects seen with e.g. UV or cisplatin? Additionally, a clear statement of the authors definition of replication stress would be welcome. *

__Response: __We will better define the term ‘replication stress’ in the revised version of the manuscript. It should be understood, in our case, that any impediment that leads to replication fork stalling and measurable by DNA combing or Chk1 phosphorylation. We have not performed experiments using UV and cisplatin.

*It is really not clear how the authors explain the link between potential changes in origin firing and resection. i.e. What is the relationship between global origin firing and resection at a particular fork, presumably broken by encounter with a CPT-arrested TOP1 complex. What is the link mechanistically? This link needs elaborating experimentally or clearly explaining based on prior literature. *

• *__Response: __Most of our results on resection has been performed with hydroxyurea, but it is true that we saw resection in absence of GNL3 in response to CPT. Treatment with HU or CPT reduces fork speed and activates additional replication origins (see Ge et al. 2007 https://pubmed.ncbi.nlm.nih.gov/18079179/ for HU or Hayakawa et al. 2021 https://pubmed.ncbi.nlm.nih.gov/34818230/ for CPT ). When GNL3 is depleted, more forks are active, meaning more targets for HU and CPT. In addition, it is likely that the firing of additional origins in response to HU and CPT is stronger in absence of GNL3. Because of this we believe that factors required to protect stalled forks may be exhausted explaining why resection is observed. This is inspired by the work of Lukas’ lab (Toledo et al. 2013 https://pubmed.ncbi.nlm.nih.gov/24267891/) and is described in the figure 6. One obvious candidate that may be exhausted is RPA, to test this we will check if resection upon GNL3 depletion and treatment with HU is still occurring in cell lines provided by Lukas’ lab that overexpress RPA complex (described in Toledo et al.). We will explain our model more carefully in the revised version.

*Related to this, I remain unconvinced that the experiments in Figure 3 show that the effects of ATRi and Wee1i on origin firing and on resection are contingent on each other. I do not believe that the authors have adequately supported the statement (end of pg 9) 'We conclude that the enhanced resection observed upon GNL3 depletion is a consequence of increased origin firing.' The link between origin firing and resection needs really needs further substantiation and / or explanation.

*__Response: __Our rational was the following. Inhibition of ATR or WEE1 increase replication origin firing, a situation that may be like the one observed for GNL3 depletion. In Toledo et al, they show that inhibition of WEE1 or ATR induces exhaustion of RPA. This exhaustion is reduced in presence of CDC7 inhibitor, roscovitine (a CDK inhibitor that inhibits origin firing) or depletion of CDC45, indicating that this is due to excessive origin activation. In our case we show that the resection observed upon WEE1 or ATR inhibition is reduced upon treatment with CDC7 inhibitor. We conclude that excessive replication origin firing induces DNA resection. Since we observed the same thing upon GNL3 depletion (but not upon BRCA1 depletion) we conclude that excessive origin firing favors DNA resection likely through exhaustion of RPA. As indicated above we will test this hypothesis by overexpressing RPA. In addition, we now show that treatment with roscovitine decreases resection upon GNL3 depletion (this will be part of the revised manuscript), an experiment that we believe confirms that excessive replication origins firing is responsible for resection upon GNL3 depletion. As suggested by reviewer 1, we will also test if depletion of MCM also reduces resection observed in absence of GNL3.

*It is not clear whether the binding of ORC2 to GNL3 also sequesters other components of the origin recognition complex? Does loss of the ability of GNL3 to bind ORC2 actually lead to more ORC bound to chromatin? How does GNL3 contribute to regulation of origin firing under normal conditions? Is it a quantitatively significant sink for ORC2 and what regulates ORC2 release? *

Response: The results of GNL3 Bio-ID were extremely clear, we could not significantly detect any other ORC subunits than ORC2 (these data were not present in the manuscript but will be added in the revised version), therefore we believe that GNL3 may sequester/regulate only ORC2. We tried to see if GNL3 depletion was changing the binding of ORC1 and ORC2 to the chromatin, but we could not see any difference, one possibility may be that small differences are not detectable by chromatin fractionation. We believe that ChIP-seq or ORC2 or other ORC subunits in absence of GNL3 is required but this it out of the scope of the study. GNL3 may regulates the stability of the ORC complex on chromatin via ORC2 but GNL3 may also regulates other ORC2 functions, at centromeres for instance. It has been shown indeed that ORC2 plays roles possibly independently of the ORC complex (see Huang et al. 2016 https://doi.org/10.1016/j.celrep.2016.02.091 or Richards et al. 2022 https://doi.org/10.1016/j.celrep.2022.111590 for instance). How exactly this is affecting origin firing is still mysterious. This is something we are planning to address in the future.

We do not know if it is a quantitatively sink for ORC2 or how this is regulated, however we believe that the ability of GNL3 to accumulate in the nucleolus may sequester ORC2. Consistent with this, we show that a mutant of GNL3 (GNL3-dB) that diffuses in the nucleoplasm interacts more with ORC2 in the nucleoplasm suggesting a release. As suggested by reviewer 1 we will now test if the interaction between ORC2 and GNL3-dB is dependent on the level of expression of GNL3-dB. In addition, we now show that expression of GNL3-dB increases replication origin firing like GNL3 depletion (data that will be added in the revised version), suggesting that regulation of ORC2 is the major cause of increased firing upon GNL3 depletion.

*Minor points: *

*All blots should include size markers *

__Response: __We will add them

*Some use of language is not sufficiently precise. For instance: ** - the meaning of 'DNA lesions' at the end of the first paragraph of the introduction needs to be more explicit. *

* - the approach to measurement of these 'lesions' (monitoring gamma-H2Ax) needs to be spelled out explicitly, e.g. line 4 of the last paragraph of the introduction. *

*

• 'we observed that the interaction between GNL3-dB and ORC2 was stronger' ... I do not see how number of foci indicates necessarily the strength of an interaction. *

* - in many places throughout 'replication origins firing' should be 'replication origin firing' (or 'firing of replication origins'). *

__Response: __We will correct these language mistakes.

__Reviewer #2 (Significance (Required)): __

The model put forward here has the potential to shed light on an important facet of the cellular response to DNA damage, namely the control of origin firing in response to replication stress that will certainly be of interest to the DNA repair / replication community and possibly more widely. The roles of GNL3 are poorly understood and this study could improve this state of affairs. However, the gaps in the mechanism outlined above and somewhat confusing conclusions do limit the ability of the paper to achieve this at present.

__Reviewer #3 (Evidence, reproducibility and clarity (Required)): __

*In this study, Lebdy et al propose a new mechanism to regulate the resection of nascent DNA at stalled replication forks. The central element of this mechanism is nucleolar protein GNL3, whose downregulation with siRNA stimulates DNA resection in the presence of stress induced by HU (Figure 1). Resection depends on the activity of nucleases MRE11 and CtIP, and can be rescued by reintroducing exogenous GNL3 protein in the cells (Figure 1G). GNL3 downregulation decreases fork speed and increases origin activity, without any strong effect on replication timing (Figure 2). Inhibition of Dbf4-dependent kinase CDC7 (a known origin-activating factor) also restricts fork resection (Figure 3). GNL3 interacts with ORC2, one of the subunits of the origin recognition complex, preferentially in nucleolar structures (Figure 4). A mutant version of GNL3 (GNL3-dB) that is not sufficiently retained in the nucleoli fails to prevent fork resection as the WT protein (Figure 5). In the final model, the authors propose that GNL3 controls the levels of origin activity (and indirectly, stalled fork resection) by maintaining a fraction of ORC2 in the nucleoli (Figure 6). *

This model is interesting and provocative, but it also relies on a significant degree of speculation. The authors are not trying to "oversell" their observations, because the Discussion section entertains different interpretations and possibilities, and the model itself contains several interrogative statements (e.g. "ORC2-dependent?"; "exhaustion of factors?").

• While the article is honest about its own limitations, the major concern remains about its highly speculative nature. I have some questions and suggestions for the authors to consider that could contribute to test (and hopefully support) their model. *

• *If GNL3 downregulation induces an excess of licensed origins and mild replicative stress resulting in some G2/M accumulation (Figure 2), what is the consequence of longer-term GNL3 ablation? Do the cells adapt, or do they accumulate signs of chromosomal instability? (micronuclei, chromosome breaks and fusions, etc) * __Response: __This is an important point also raised by Reviewer 2: deletion of GNL3 leads to embryonic lethality in mouse and ES cells deleted for GNL3 do not proliferate and fail to enter into S-phase. Consistent with this, the clones deleted for GNL3 that we obtained using CRISPR-Cas9 grow poorly, thus preventing us to do experiments. To our knowledge micronuclei and chromosome breaks have never been analyzed upon transient depletion of GNL3 using siRNA. However, it is well established that depletion of GNL3 induces phosphorylation of H2A.X) and the formation of ATR, RPA32 and 53BP1 foci due to S-phase arrest (Lin et al. 2013. https://pubmed.ncbi.nlm.nih.gov/24610951/ ; Meng et al. 2013. https://pubmed.ncbi.nlm.nih.gov/23798389/). DNA lesions have also been visualized by comet assay (Lin et al. 2019. https://pubmed.ncbi.nlm.nih.gov/30692636/). Consistent with this we observed a weak increased of DNA double-strand breaks upon GNL3 depletion using pulse-field gel electrophoresis as well as mitotic DNA synthesis (MiDAS). We can integrate this data in the revised version of the manuscript if required. To sum up, it is clear that GNL3 depletion is inducing problems during S-phase that may lead to possible genomic rearrangements.

• The model relies on the link between origin activity and stalled fork resection that is almost exclusively based on the results obtained with CDC7i (Figure 3). But CDC7 has other targets besides pre-RC components at the origins, such as Exo1 (from the Weinreich lab, cited in the study), MERIT40 and PDS5B (from the Jallepalli lab, also cited). The effect of CDC7i could be exerted through these factors, which are linked to fork stability and DNA resection. The loss of BRCA1 (Figure 3F) could somehow entail the loss of control over these factors. Could the authors check the possible participation of these proteins?*

__Response: __It is true that CDC7 has other targets than pre-RC components. We therefore decided to inhibit origin firing using roscovitine, a broad CDK inhibitor, a strategy previously used in Lukas lab (Toledo et al. 2013. https://pubmed.ncbi.nlm.nih.gov/24267891/). We observed that treatment with roscovitine decreased significantly resection observed upon GNL3 depletion, confirming the link between origin activity and stalled fork resection. This will be integrated in the revised version of the manuscript. As asked by Reviewer 1, we will also perform depletion of MCM to strength our model.

Exo1 is indeed a target of CDC7 as shown by the Weinreich lab (Sasi et al. 2018. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6111017/) however the authors do not formally demonstrate that Exo1 phosphorylation is required for its activity. We observed that depletion of Exo1 significantly reduced resection upon GNL3 depletion (data that will be added in the revised version), indicating that the effect of CDC7 inhibitor could be exerted via the control of Exo1. This is why our BRCA1 control is important, it is well stablished that Exo1 is required for nascent strand degradation upon BRCA1 depletion (Lemaçon et al. 2017. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5643552/) but CDC7 inhibition has no effect on resection upon BRCA1 depletion suggesting that resection by Exo1 may not be regulated by CDC7 in our context.

As stated by the reviewer MERIT40 and PDS5B are targets of DDK kinases (Jones et al. 2021 https://doi-org.insb.bib.cnrs.fr/10.1016/j.molcel.2021.01.004) and seem to be required for protection of nascent DNA and in response to HU. However, little is known about the role(s) of these proteins and we think that adding them will complicate message. We hope the reviewer understands this.

The model also relies on the fact that GNL3-dB mutant (not retained in the nucleoli) is not sufficient to counteract fork resection induced by HU (Figure 5G). The authors should test directly whether GNL3-dB induces extra origin activation, using their available DNA fibers-based technique.

__Response: __This is an excellent point. We have now GIFD (Global Instant Fork Density) data that shows that the number of active forks is increased upon dB GNL3-dB expression. It demonstrates that when GNL3 is no longer retained in the nucleolus more origins are active. These data will be integrated in the revised version of the manuscript, and we believe further support the regulation of ORC2 by GNL3.

*Finally, the model implies an exquisite regulation of the amount of ORC2 protein, which could influence the number of active origins and the extent of fork resection in case of stress. In this scenario, one could predict that ORC2 ectopic expression would have similar, or even stronger effects, than GNL3 downregulation. Is this the case? *

__Response: __We completely agree with this prediction. However, we are afraid that overexpression of ORC2 may have indirect effects due to the many described functions of ORC2, therefore it may be difficult to interpret the data. We will give a try anyway.

*Even if the connection between origins and fork resection could be firmly established, the molecular link between them remains enigmatic. The authors hint (as "data not shown") that it is neither mediated by RPA nor RAD51. Unfortunately, the reader is left without a clear hypothesis about this point. *

__Response: __We will add data that show that RPA and RAD51 recruitment is not affected by GNL3 depletion. However, the sensitivity of chromatin fractionation approach may be too weak to detect low differences. Based on the work of Lukas Lab (Toledo et al. 2013 https://pubmed.ncbi.nlm.nih.gov/24267891/) one possible mechanism may be exhaustion of the pool of RPA. This may link the excessive activation of origins observed upon GNL3 depletion and resection. To test this, we will check if resection upon GNL3 depletion and treatment with HU is still occurring in cell lines that overexpress RPA complex (described in Toledo et al.) that we obtained from Lukas’ lab.

__ __ **Referees cross-commenting**

__ __In addition to each reviewer's more specific comments, the three reviews share a main criticism: the lack of mechanistic information about the proposed link between origin activity and resection of nascent DNA at stalled forks.

__Reviewer #3 (Significance (Required)): __

In principle, this study would appeal to the readership interested in fundamental mechanisms of DNA replication and the cellular responses to replicative stress.

For the reasons outlined in the previous section, I believe that in its current version the study is not strong enough to provide a new paradigm about origins being regulated by partial ORC2 sequestering at nucleoli. The other potentially interesting advance is the connection between frequency of origin activity and the extent of nascent DNA resection at stalled forks, but the molecular link between both remains unknown.

---

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 #3

Evidence, reproducibility and clarity

In this study, Lebdy et al propose a new mechanism to regulate the resection of nascent DNA at stalled replication forks. The central element of this mechanism is nucleolar protein GNL3, whose downregulation with siRNA stimulates DNA resection in the presence of stress induced by HU (Figure 1). Resection depends on the activity of nucleases MRE11 and CtIP, and can be rescued by reintroducing exogenous GNL3 protein in the cells (Figure 1G). GNL3 downregulation decreases fork speed and increases origin activity, without any strong effect on replication timing (Figure 2). Inhibition of Dbf4-dependent kinase CDC7 (a known origin-activating factor) also restricts fork resection (Figure 3). GNL3 interacts with ORC2, one of the subunits of the origin recognition complex, preferentially in nucleolar structures (Figure 4). A mutant version of GNL3 (GNL3-dB) that is not sufficiently retained in the nucleoli fails to prevent fork resection as the WT protein (Figure 5). In the final model, the authors propose that GNL3 controls the levels of origin activity (and indirectly, stalled fork resection) by maintaining a fraction of ORC2 in the nucleoli (Figure 6).

This model is interesting and provocative, but it also relies on a significant degree of speculation. The authors are not trying to "oversell" their observations, because the Discussion section entertains different interpretations and possibilities, and the model itself contains several interrogative statements (e.g. "ORC2-dependent?"; "exhaustion of factors?").

While the article is honest about its own limitations, the major concern remains about its highly speculative nature. I have some questions and suggestions for the authors to consider that could contribute to test (and hopefully support) their model.

1. If GNL3 downregulation induces an excess of licensed origins and mild replicative stress resulting in some G2/M accumulation (Figure 2), what is the consequence of longer-term GNL3 ablation? Do the cells adapt, or do they accumulate signs of chromosomal instability? (micronuclei, chromosome breaks and fusions, etc)
2. The model relies on the link between origin activity and stalled fork resection that is almost exclusively based on the results obtained with CDC7i (Figure 3). But CDC7 has other targets besides pre-RC components at the origins, such as Exo1 (from the Weinreich lab, cited in the study), MERIT40 and PDS5B (from the Jallepalli lab, also cited). The effect of CDC7i could be exerted through these factors, which are linked to fork stability and DNA resection. The loss of BRCA1 (Figure 3F) could somehow entail the loss of control over these factors. Could the authors check the possible participation of these proteins?
3. The model also relies on the fact that GNL3-dB mutant (not retained in the nucleoli) is not sufficient to counteract fork resection induced by HU (Figure 5G). The authors should test directly whether GNL3-dB induces extra origin activation, using their available DNA fibers-based technique.
4. Finally, the model implies an exquisite regulation of the amount of ORC2 protein, which could influence the number of active origins and the extent of fork resection in case of stress. In this scenario, one could predict that ORC2 ectopic expression would have similar, or even stronger effects, than GNL3 downregulation. Is this the case?
5. Even if the connection between origins and fork resection could be firmly established, the molecular link between them remains enigmatic. The authors hint (as "data not shown") that it is neither mediated by RPA nor RAD51. Unfortunately, the reader is left without a clear hypothesis about this point.

Referees cross-commenting

In addition to each reviewer's more specific comments, the three reviews share a main criticism: the lack of mechanistic information about the proposed link between origin activity and resection of nascent DNA at stalled forks.

Significance

In principle, this study would appeal to the readership interested in fundamental mechanisms of DNA replication and the cellular responses to replicative stress.

For the reasons outlined in the previous section, I believe that in its current version the study is not strong enough to provide a new paradigm about origins being regulated by partial ORC2 sequestering at nucleoli. The other potentially interesting advance is the connection between frequency of origin activity and the extent of nascent DNA resection at stalled forks, but the molecular link between both remains unknown.

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

This manuscript explores the role of GNL3/nucleostemin in DNA replication and specifically in the response of DNA replication to DNA damage. GNL3 is a predominantly nucleolar protein, previously characterised as a GTP-binding protein and shown to be necessary for effective recruitment of the RAD51 recombinase to DNA breaks. The entry point for this report is a mini screen, based on proteins identified previously by the authors to associate with replication forks by iPOND, for factors that increase gamma-H2Ax (an indicator of DNA damage) after treatment with the Top1 inhibitor camptothecin (CPT). In this mini-screen GNL3 emerged as the top hit.

The authors put forward the hypothesis that GNL3 is able to sequester the replication licensing factor ORC2 in the nucleolus and that failure of this mechanism leads to excessive origin firing and DNA resection following CPT treatment.

The model put forward is interesting, but currently rather confusing. However, for the reasons upon which I expand below, I do not believe that the data provide a compelling mechanistic explanation for the effects that are reported and I am left not being certain about some of the links that are made between the various parts of the study, even though individual observations appear to be of good quality.

Specific points:

The knockdown of GNL3 is very incomplete. In this regard, the complementation experiments are welcome and important. However, is it an essential protein? Can it be simply deleted with CRISPR-Cas9?

Global instant fork density is not quite the same as actually measuring origin firing. Ideally, it would be good to see some more direct evidence of addition origin firing e.g. by EdU-seq (Macheret & Halazonetis Nature 2018) but this would be quite a significant additional undertaking. However, given the authors have performed DNA combing with DNA counterstain, they should be able to provide accurate measurements of origin density and inter-origin distance.

'Replication stress' is induced with CPT. This term is frequently used to describe events that lead to helicase-polymerase uncoupling (e.g. O'Connor Mol Cell 2015) but that is not the case with CPT, which causes fork collapse and breaks. Are similar effects seen with e.g. UV or cisplatin? Additionally, a clear statement of the authors definition of replication stress would be welcome.

It is really not clear how the authors explain the link between potential changes in origin firing and resection. i.e. What is the relationship between global origin firing and resection at a particular fork, presumably broken by encounter with a CPT-arrested TOP1 complex. What is the link mechanistically? This link needs elaborating experimentally or clearly explaining based on prior literature.

Related to this, I remain unconvinced that the experiments in Figure 3 show that the effects of ATRi and Wee1i on origin firing and on resection are contingent on each other. I do not believe that the authors have adequately supported the statement (end of pg 9) 'We conclude that the enhanced resection observed upon GNL3 depletion is a consequence of increased origin firing.' The link between origin firing and resection needs really needs further substantiation and / or explanation.

It is not clear whether the binding of ORC2 to GNL3 also sequesters other components of the origin recognition complex? Does loss of the ability of GNL3 to bind ORC2 actually lead to more ORC bound to chromatin? How does GNL3 contribute to regulation of origin firing under normal conditions? Is it a quantitatively significant sink for ORC2 and what regulates ORC2 release?

Minor points:

All blots should include size markers

Some use of language is not sufficiently precise. For instance:

• the meaning of 'DNA lesions' at the end of the first paragraph of the introduction needs to be more explicit.
• the approach to measurement of these 'lesions' (monitoring gamma-H2Ax) needs to be spelled out explicitly, e.g. line 4 of the last paragraph of the introduction.
• 'we observed that the interaction between GNL3-dB and ORC2 was stronger' ... I do not see how number of foci indicates necessarily the strength of an interaction.
• in many places throughout 'replication origins firing' should be 'replication origin firing' (or 'firing of replication origins').

Significance

The model put forward here has the potential to shed light on an important facet of the cellular response to DNA damage, namely the control of origin firing in response to replication stress that will certainly be of interest to the DNA repair / replication community and possibly more widely. The roles of GNL3 are poorly understood and this study could improve this state of affairs. However, the gaps in the mechanism outlined above and somewhat confusing conclusions do limit the ability of the paper to achieve this at present.

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 Lebdy et al. describe a new role of GNL3 in DNA replication. They show that GNL3 controls replication fork stability in response to replication stress and they propose this is due to the regulation of ORC2 and the licensing of origins of replication. Their data suggest that GNL3 regulates the sub nuclear localization of ORC2 to limit the number of licensed origins of replication and to prevent resection of DNA at stalled forks in the presence of replication stress.

While many of the points of the manuscript are proven and well supported by the results, there are some experiments that could improve the quality and impact of the manuscript. The main issue is that the connection between the role of GNL3 in controlling ORC2, the firing of new origins and the protection of replication forks is not clearly established. At the moment the model relies on mainly correlative data. In order to further substantiate the model, we propose to address some of the following issues:

1. The authors indicate that RPA and RAD51 accumulation at stalled forks is not affected by GNL3 depletion. These data should be included and other proteins should be analysed. In addition, the role of helicases could be explored through the depletion of the main helicases involved in the remodelling of the forks.
2. The proposed model implies that GNL3 depletion leads to increased origin licensing. FThe authors should address if the primary effect of GNL3 depletion is on origin firing by using CDC7 inhibition in the absence of stress (Rodríguez-Acebes et al., JBC 2018).
3. A way to prove that origin firing mediates the effect of GNL3 on fork protection would be to reduce the number of available origins. The depletion of MCM complexes has been shown to limit the number of back-up origins that are licensed and leads to sensitivity to replication stress (Ibarra et al., PNAS 2008). If GNL3 depletion results in increased number of origins, this effect should be prevented by the partial depletion of MCM complexes.
4. Alternatively, the authors could try to modulate the depletion of GNL3. Origin licensing takes place in the G1 phase and thus the depletion of GNL3 by siRNA could affect the following S phase. Using an inducible degron for GNL3 depletion would allow to deplete GNL3 in G1 or S phase specifically. If the model is correct, the removal of GNL3 in S phase should not affect fork protection but removing GNL3 in the previous G2/M phase should reduce the number of licensed origins and lead to impaired fork protection. In addition to the connection GNL3-origin firing-fork protection, it is unclear how the lack of GNL3 in the nucleolus and the change in the sub nuclear localization of ORC2 controls origin firing and resection. The strong interaction observed between GNL3-dB and ORC2, and the subsequent change in ORC2 localization does not explain how origin licensing can be affected. In this sense, the authors could address:
   1. Does the depletion of GNL3 and the expression of GNL3-dB affect the formation of the ORC complex, its subnuclear localization or its binding to chromatin? The authors have not explored if the interaction of GNL3 with ORC2 is established in the context of the ORC complex. An IF showing NOP1 with PLA data from GNL3-dB and ORC2 is needed to analyse how the expression of increasing amounts of GNL3-dB affects ORC2.
   2. In order to confirm if the mislocalization of ORC2 by the expression of GNL3-dB increases origin firing and mediates the effects on fork protection the authors could check DNA resection levels inhibiting CDC7 in high GNL3-dB conditions. Also, the levels of MCM2, phosphor-MCM2, CDC45, have not been analysed upon expression of GNL3-dB.
   3. The data in the paper suggest that GNL3 may affect the role of ORC2 in centromeres. Since depletion of GNL3 leads to increased levels of H2AX, it would be interesting to address if this damage is due to incomplete replication in centromeres by analysing the co-localization of H2AX and centromeric markers both in unstressed conditions and upon the induction of replication stress.

Minor points:

1. In the initial esiRNA screen the basal levels of H2AX should also be shown.
2. Figure EV1B: I think the rank needs another RS mark to see better the effect of each esiRNA on DNA lesions (high variability in all the conditions showed).
3. Figure 1C and Figure EV1D/E: the quantification of the pCHK1/CHK1 levels could be included to show that there are no changes in phosphorylation upon GNL3 depletion.
4. In the first section of the results, at the end Figure 4B is incorrectly called for.
5. The levels of GLN3 expression in 293 cells should be already included in section GNL3 interacts with ORC2.
6. The full MS data needs to be included for both GNL3 and ORC2.
7. Figure 4B should be improved, since there is a faint band in the IgG mouse control.

Significance

The work is nicely written, the figures are well presented and the experiments have the necessary controls. It provides relevant information to understand how replication stress is controlled and linked to replication fork protection through origin firing. These results are relevant to the field, linking GNL3 to origin firing and with potential to help understand the role of GNL3 in cancer. They provide new information and can give rise to new studies in the future. Many of the conclusions of the manuscript are well supported. Additional support for some of the main claims would strengthen the results and also increase the impact providing a bigger conceptual advance by performing some of the suggested experiments.

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

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

1. General Statements [optional]

We are grateful for the thoughtful comments and suggestions from the reviewers which we feel have resulted in a manuscript that is both clearer to the reader and more rigorous. We have addressed the suggested revisions point by point below:

2. Point-by-point description of the revisions

Reviewer #1 Major comments:

Figure 2E. This shows the residence time of Cse4 without Ndc10 association. How does this compare to the residence time on mutant CEN3 (Supplemental Figure 1). It looks like Cse4 still binds to CEN3 with some specificity even in the absence of Ndc10. Does this suggest that Cse4 has some intrinsic ability to recognize CEN3? Alternatively, Ndc10 is still required for Cse4 binding but is below detection in the Cse4-alone residence lifetimes. Ideally, the authors would compare this with Cse4 binding in an Ndc10 mutant.

We thank the reviewer for this interesting question. As suggested, we analyzed Cse4 behavior on the mutant CDEIIImut CEN3 DNA, which does not stably recruit Ndc10 (Figure 1C), using the real-time colocalization assay. Although overall Cse4 associations were reduced, we still observed transient interactions, consistent with the possibility that Cse4 has some intrinsic ability to recognize CEN3. Kaplan-Meier analysis of the lifetimes of Cse4 colocalizations on CDEIIImut CEN3 DNA were significantly reduced when compared to CEN DNA (Figure EV1G, H). We have added these points to the text (p. 10, lines 29-31 and p. 11, lines 1-8).

Figure 3 explores the very interesting relationship between Scm3 dynamics and Cse4 binding but I feel that the data is not fully flushed out. A key finding is that Cse4 can potentially bind to CEN DNA prior to engaging with Scm3 to be incorporated. This predicts that Cse4 should bind first and then colocalize with Scm3. Can this be detected in the timing of the traces? How often does Scm3 bind to already prebound Cse4 and does this correlate with Cse4 residing longer?

We performed new and more rigorous analyses of the data to address these questions in the revised manuscript. After our analysis to calculate ternaryScm3 off-rates, we analyzed the relative timing of these ternary residences and found that indeed Cse4 can bind to CEN DNA prior to Scm3 and these events do correlate with Cse4 residing longer. Complete analysis of the binding order of Cse4 and Scm3 ternary residences revealed that Scm3 bound to CEN3 DNA prior to Cse4 more often than Cse4 preceding Scm3 (46% vs 34% of ternary residences) with the remaining 20% arriving simultaneously (Figure EV2E). Despite a difference in prevalence, the median lifetimes of both Scm3-Firstand Scm3-Last contexts were similar to each other (Figure EV2F) and both were significantly stabilized when compared to non-ternary residences. These results highlight a potential mechanism where Scm3 catalyzes stable Cse4 incorporation at centromeric DNA instead of delivering it to the centromere regardless of the order of arrival. These data are now reported and discussed in the revision (p. 12, line 11-p.13, line 10).

A related and perhaps even more interesting point is whether Scm3 is involved in "loading" of Cse4. If so, then one would expect that once Cse4 is assembled into nucleosomes it should be stable, even if Scm3 leaves. Can the authors extract this from the data? Alternatively, it is possible that Scm3 remains associated to Cse4 to maintain the nucleosome which would imply a more extended role for Scm3 apart from assembly alone. It would be interesting if information on this can be extracted from the data.

Using our updated analysis of ternaryScm3 Cse4 residences, feel we have addressed this possibility indirectly in a couple of ways. First, we found that in instances where ternary Scm3/Cse4 complexes are formed, Scm3 co-occupied the CEN DNA an average of 56.0% of the total Cse4 residence time, which is distinctly shorter than the 84% of the total Cse4 residence that was occupied by Ndc10 in Cse4/Ndc10 ternary residences. These findings are consistent with a Scm3-turnover mechanism that occurs post ternary complex formation with Cse4 as we found that Cse4 off-rates were significantly reduced after Scm3 association (Figure 3D).

Second, further analysis of Scm3 residence behavior revealed that there was no significant stabilization of Scm3 after ternary Cse4/Scm3 complex formation vs non-ternary Scm3 residences found in either off-rates (33 hr-1 vs 32 hr-1, Figure EV2C) or median lifetimes (45 s vs 52 s, Figure EV2D). These results indicate that Scm3 association is not stabilized like Cse4 after ternary complex formation and points to a potential catalytic role in Cse4 nucleosome formation, leaving a stable Cse4 nucleosome after turnover. We reported these findings in the revision results section (p. 12, lines 11-16) as well as briefly within the discussion.

Even in the presence of Scm3 and CCAN components, Cse4 appears to have a limited lifetime in the in vitro assay compared to in vivo stability. The authors should speculate on whether activities exist in their extract that actively disassembles nucleosomes. Perhaps ATP could be depleted to inactivate remodellers?

This is an excellent suggestion that we addressed with a new experiment. We performed an endpoint localization experiment with lysate containing fluorescently labeled Ndc10 and Cse4 and then removed the lysate and incubated the slide for 24hr in imaging buffer at RT. Strikingly, the proteins were maintained at the CEN DNA with a high protein total colocalization (~75% retention) (shown in Figure EV1B, C). These data suggest that the lysates may contain negative regulatory factors and we have added this point in the revised text (p. 9 lines 6-25).

We were not able to address whether the removal of ATP stabilized the proteins because we previously found that ATP depletion of the lysates completely abrogates kinetochore assembly in extracts. We will need to eventually dissect the role of remodelers in future work using a different approach.

For Figure 6, it is not clear why AT-track mutants of CDEII are labeled as genetically stable and genetically unstable. This is confusing as the "genetically stable" show a more than 10-fold increase in chromosome loss rates. Perhaps these can be renamed into "unstable" and "very unstable" or "weak" and "strong" mutants, just to make clear that these classes are both poorer than wild type.

We had deferred to the nomenclature used in the previous study (Baker and Rogers, 2005) which initially characterized these mutants. To avoid this confusion, we have renamed these mutants “unstable” and “very unstable” as suggested to make it clearer that none of these synthetic mutants fully recapitulate WT CEN3 behavior.

Finally, it would be wonderful to include data to assess whether a full Cse4 nucleosome is assembled or a partial nucleosome e.g. just Cse4/H4 tetrasomes. This could be done by tracking the accumulation of H2A or H2B at the CEN3. This would give further insight into what step Scm3 catalyses.

This is a very interesting suggestion that we were not able to directly address. Epitope tagging of these histone proteins in Saccharomyces cerevisiae with endogenous fluorophores has proved challenging due to gene duplication, overall sensitivity to histone levels within the cells and disruption of histone function by epitope tagging. We hope to find a workable method to address this in the future to address this question directly.

Minor comments:

Typo on page 5, line 1 "nucleosom" missing an e.

We have corrected this in the revised text.

Kaplan-Meyer should be spelled Kaplan-Meier

We have corrected this in the revised text.

The term "censored" is mentioned across many figures but comes up just ones in the methods where it is not clearly explained. Perhaps this could be clarified in the legend.

We have now provided a clear explanation of the term “censored” in the text on p. 28, lines 25-27. It has also been added to the figure legends and reported in the Statistical tests section of the methods section to address this point.

The abstract states that Cse4 can arrive at the centromere without its chaperone. More likely, Cse4 is in complex with other chaperones that may allow it to bind. Perhaps the abstract can be modified to read "Cse4 can arrive at the centromere without its dedicated centromere-specific chaperone Scm3..."

We updated the abstract to reflect this point in the revised text.

Related to this point, the discussion states the possibility that Cse4 can initially bind to CEN3 via other more general chaperones. However, it should be acknowledged that transient Cse4 binding in their assay may simply occur through mass action due to high concentrations of CEN3 DNA. In vivo, this transient binding may not be that relevant.

We acknowledge this potential caveat in the discussion section (p. 20 lines 15-18), although we feel this is somewhat unlikely due to our observation of significantly reduced Cse4 binding on CDEIIImut DNA despite DNA concentrations being similar to previous assays (Figure EV3A). We speculate that some of this transient behavior is at least in part driven by two major factors: negative regulatory factors within our cellular extracts that counter nucleosome formation (as explored in Figure EV1B-C) and photostability of the endogenous fluorophores used within the study (Figure EV1D-E). These points were highlighted within the second paragraph of page 9.

Reviewer #2 Major comments:

1. Figure S1A-D seem like some of the most compelling data in the paper to bolster the rigor of their experimental setup. There appears to be plenty of space to include these data in the main figure set in Figure 1 after panel D. The authors would be well served to consider moving S1A-D somewhere in the main figure set.

We appreciate the helpful feedback on the importance of the date found in Supplemental Figure 1 and have now incorporated it into Figure 1 within the main text as suggested.

The authors conclude that Ndc10 recruits HJURP(Smc3) to the yeast point centromeres. If this is the case, can the TIRFM assay measure ternary residence lifetimes complexes between Ndc10/HJURP(Smc3)/CenDNA?

We made the conclusion that Ndc10 recruits Scm3 based on previous publications showing this requirement in vivo. We have now attempted to address this in our assay indirectly by monitoring Scm3 behavior on the CDEIIImut CEN3 DNA that lacks Ndc10. Surprisingly, we found that Scm3 interacted similarly with CDEIIImut CEN3 DNA and actually showed an increase in median lifetimes vs. CEN3 DNA (Figure EV2B), suggesting its intrinsic DNA-binding activity may be the primary driver of its CEN DNA binding and that stable Cse4 association is required for its turnover. These data suggest that Ndc10 is not driving Scm3 interaction (or targeting) to CEN3. We are grateful to the reviewer for pointing this out and have adjusted our conclusions in the revised manuscript (p. 12, line 26-p.13 line 10).

Throughout the manuscript short- and long- term residence lifespans are mentioned, referencing the figures containing lines with lengths depicting residence times. This is a qualitative reference to short and long residences. Can the authors provide a graph for short-term ( 300 s) residence life-spans for, CENPA alone, CENPA/Ndc10, and CENPA/HJURP on CEN3 DNA? Or some figure similar to Figure 3C, but reporting the proportion of short-term vs long-term residence?

We typically used Kaplan-Meier survival function estimates to compare binding behavior but agree that quantification of residences within these contexts may be easier for the reader to follow. We have therefore quantified short-term ( 300 s) as suggested and added them as a panel (F) to Figure EV1 and as panel (E) in Figure 3.

The choice of CCAN components for analysis in Fig. 5 is interesting, but many readers may be curious why Mif2 wasn't selected for disruption, since it has such a cozy placement with CENP-A and CEN DNA. Can choice to not include Mif2 mutants/degrons be mentioned/justified in the text (unless, even better yet, they choose to address Mif2 role directly in new experimentation)?

We relied on structural models to choose CCAN proteins that are in close proximity to the DNA. Because Mif2 is not in these structures, we did not include it in our studies. We have explained this in the revised text (p. 16 lines 14-17) and agree it is an interesting future area of study.

Minor comments:

1. Are these whole cell extracts (WCE) DNA-free? I'm curious if there is any competition from endogenous DNA from the yeast cellular extract.

The extracts are not DNA-free so it is likely there is some competition from endogenous DNA. We have avoided enzymatically removing the DNA since the TIRF assay depends on the integrity of DNA.

In relation to Mif2 and comment #4 above, do the authors make any connection to their results with synthetic nucleosome sequences not being conducive to yeast centromere formation with the prior observation (Allu et al 2019) using recombinant components that the human version of Mif2 more easily saturates its binding sites on CENP-A nucleosomes when they are assembled with natural centromere DNA rather than the Widom 601 sequence?

We did not speculate on the role of Mif2 and stability of synthetic nucleosome sequences. This is an interesting point but the differences between the yeast and human systems combined with the fact we have not yet started to study Mif2 made it seem too premature to include in this manuscript.

Providing a gel (or other measure) of the DNA templates (750, 250, and 80 bp) used in TIRFM assay would be nice to show to confirm the designed size of the pre-tethered DNA.

We agree this is a helpful control and we have now included it as a panel in Figure EV5B.

1. Some of the references to figures/figure panels in the main text do not match the figures. (discussion, pg 16, paragraph 1 & 2;pg 18, paragraph 1).

We have updated references mentioned to reflect to the correct figure in both sections of the discussion.

Reviewer #3 Major comments: -->Statistics are highly recommended for all the data in the ms.

We have included log-rank analysis to instances where two survival functions were plotted together where appropriate. P-values for all these analyses were reported in the appropriate figure legends.

• At what rate is data collected in the TIRFM setup. For clarity for the reader, it is important to provide imaging details for time-lapse. What is the impact of photobleaching on the stability of the fluorophore signal? Please clarify.

This is a helpful suggestion and we have now included imaging details (like time intervals for each channel) when the real-time assay is introduced in the results section (p. 8, lines 10-12). We have also provided additional details for the photobleaching estimates and how these might censor data in turn (p. 9, lines 14-25). Although photobleaching is a primary limitation of the time-lapse assays, we point out that it is appropriate to compare protein behavior under identical imaging parameters within differing contexts. We also noted that we typically compared time-lapse behavior (which is affected by photobleaching) with endpoint assays to ensure consistent behavior.

• The power of single-molecule technique is precisely that such data can be made quantitative. Indeed, the Kaplan-Meyer graphs do show nice quantitative results. Unfortunately, in the text few quantitative measurements are reported. In fact, the Kaplan-Meyer graphs can be interpreted in a quantitative manner such as probability of residency time. Although in most cases the statistical significance between two conditions can be expected, this is not formally calculated and shown. What is the 50% survival time of Cse4 alone or with Ndc10, for instance? This manuscript would greatly benefit from a quantitative approach to the data, including a summary table of the results of the various conditions tested. Please add this important table.

We initially put the quantitative data in the figure legends but omitted it from the main text for simplicity but appreciate the Reviewer’s point. We note that we performed log-rank tests on all Kaplan-Meier analyses that are plotted on the same graph to provide statistical differences where applicable and have included all P-values in the figure legends. In response to the suggestion, we have now also included a table (Table 1) that contains the median survival time for all proteins tested as well as the median survival times for the differing contexts tested for quick reference and easier comparison for the reader.

• This reviewer would expect that the endpoint (90 min) would roughly reflect the occupancy results from time-lapse (45 min) experiments. Based on the data presented in Figures 1, 2, S1-3, this does not appear to be the case. 50% of Cse4-GFP and 70% Ndc10-mCherry colocalized with CEN3 DNA in the endpoint experiment, whereas in Fig 2C, ~30 and ~52 traces were positive for Cse4-GFP and Ndc10-mCherry, resp. with the former having drastically lower residency survival compared to Ndc10-mCherry. If indeed, 50% of Cse4-GFP makes it to the endpoint, about 50% of all traces should reach the end of the 45 minutes time-lapse window. Only about 1/3 of all positive Cse4-GFP traces can be seen at the end of the 45 min window. Could this be due to technical issues of photostability of GFP? Why does the colocalization of Cse4 signal with the DNA signal disappear so readily? Are Cse4 so unstable? Is the same true for canonical H3 nucleosomes? This unlikely true for nucleosomes in cells.

This is a valid concern, and we appreciate the thoughtful critique. The inconsistency noted between the initial endpoint colocalization and those reported later in the paper is mainly due to the difference between yeast strains carrying Cse4 tagged alone in comparison to multiple kinetochore proteins with tags that exhibit mild genetic interactions. This point is now explained in the revised text (p. 8, line 29-p. 9. line 3).

Photostability is also a factor in the live imaging experiments compared to the endpoint localization assays. However, our photobleaching estimates suggest that the Cse4 lifetimes are not limited by photobleaching (Figure EV1D, E) so we do not believe that accounts for the differences between experiments and it is mainly the presence of multiple epitope tags.

In regard to why Cse4 is not more stable, Reviewer 1 had the same question so we performed an experiment to address whether the lysate contains negative regulatory factors. We found that Cse4 is stable once the lysate is removed (Figure EV1B, C), consistent with the idea that there are factors that disrupt it in the lysate. We discuss these potential reasons for transient association in the revised text (p. 9, lines 4-25).

It should also be noted that there are clear differences in nucleosome formation in reconstitutions and within our extracts, as evident by the Widom-601 DNA data (Figure 6D). This was not necessarily unexpected, as extracts are a much more complex medium, but we are encouraged by the fact that at least once formed, these Cse4-containing particles on CEN DNA are perhaps more stable than their reconstituted counterparts that seem to be so far unsuitable for structural studies.

Along the same lines, in Suppl Fig 3 there is a disconnect between residency survival and endpoint colocalization on either CEN3, CEN7, or CEN9. What could be the underlying mechanism between the discordance of endpoint results and time-lapse results? Could this be the result of experimental differences?

We are grateful that this discrepancy was highlighted to us, as upon closer examination we discovered that endpoint colocalization analysis had not been correctly updated in the figure to include data from equivalent genetic backgrounds as the CEN3 and CEN9 assays. Updating the figure in Appendix Figure S2 to include this data remedied this discrepancy.

• What fraction of particles show colocalization between Cse4-GFP and Ndc10-mCherry? What fraction of occupancy time show colocalization between Cse4-GFP and Ndc10-mCherry? Altogether, understanding the limitation and benefits of endpoint analysis and time-lapse analysis in this particular experimental set-up is important to be able to interpret the results. Please clarify these points.

We have now added particle numbers to all survival estimate plots which makes it much easier for the reader to interpret the proportion of Cse4 residences that are ternary vs. non-ternary in instances where off-rates were quantified and Kaplan-Meier analysis was performed on the resulting lifetimes. We determined that for ternary Cse4-Ndc10 residences, Cse4 and Ndc10 co-occupied the CEN DNA an average of ~84% of the total Cse4 residence.

• Page 9, third sentence of third paragraph it is stated that the "results suggests that Scm3 helps promote more stable binding of Cse4 ...". This is indeed one possible explanation of the results, and this possibility is tested by overexpressing Psh1 or Scm3 by endpoint colocalization analysis. 1) Taking the concerns regarding the endpoint vs time-lapse results into account, wouldn't it be more accurate to compare either time-lapse results against each other or endpoint results? 2) Alternatively, more stable Cse4 particles are able to recruit Scm3, simply because of the increased binding opportunity of a more stable particle. In this scenario, just the residency time of Cse4 alone is the predicting factor in likelihood to associate with Scm3. To test the latter possibility, Cse4 stability would need to be altered. Please consider this experiment- should be relatively easy with the right mutant of either CSE4 or CDEII (see Luger or Wu papers).

We appreciate the points raised here and addressed both as follows. For point (1) we altered the text in the third paragraph of the section, The conserved chaperone Scm3HJURP is a limiting cofactor that promotes stable association Cse4CENP-A with the centromere, to make it more clear to the reader that in the experiments presented in Figure EV4, endpoint analysis results were only compared to each other, and likewise time-lapse experiments were only compared to each other for each genetic background. While the results were consistent between experiments, we did not directly compare results from one to the results of another, but instead we used both assays to characterize Cse4CENP-A behavior more completely in differing contexts.

To test the alternative hypothesis proposed in point (2), we sought to avoid potential selection bias by utilizing off-rate analysis, which enabled us to separate the portions of Cse4 residences that occurred prior to ternary association with either Ndc10 or Scm3. This unbiased approach allowed us to compare Cse4 residence lifetimes pre and post ternary association and we found that there were still significant differences in off-rates and median lifetimes of the associated ternary and non-ternary residences using this updated analysis. We thank the reviewer for helping to guide us towards this more robust analysis.

Based on the recommendation in point (2), we also sought to directly compare the behavior of Cse4 and Scm3 on the “Very Unstable” CDEII mutants described in the section, DNA-composition of centromeres contributes to genetic stability through Cse4CENP-A recruitment. In this case, equivalent extracts were used and Cse4 stability was altered directly via the DNA template. When the off rates of ternaryScm3 Cse4 residences were compared, we found a significant increase in off-rates of Cse4 on the “Very Unstable” CDEII mutant CEN DNA (Appendix Figure S3B) compared to WT CEN DNA. If the alternative hypothesis proposed in point (2) were true, we would expect this reduction in median lifetime to correlate with a lower proportion of Cse4-Scm3 ternary association but quantification yielded proportions that, while varied, were not on average lower than the proportion of Cse4-Scm4 association on CEN3 DNA (.23 vs .31, Appendix Figure S3A). This finding, taken together with the fact that it would be difficult for us to propose an alternative hypothesis that explains the results outlined in Figure EV4, supports our hypothesis that Scm3 helps promote more stable binding of Cse4 and that this stabilization is directly influenced by DNA sequence composition.

• In Figure 1C and Supplemental Figure 5B, there appears to be foci that CEN3-ATTO-647 positive, but Cse4-GFP negative and visa verse. It seems logical that there are DNA molecules that didn't reconstitute Cse4 nucleosomes. But how can there be Cse4-GFP positive foci without a positive DNA signal? Is it possible that unlabeled DNA make it onto the flow chamber? If so, can these unlabeled DNA be visualized by Sytox Orange for instance to confirm that no spurious Cse4 deposition occurred? Please clarify.

Because it is unlikely that random associations will colocalize with the labeled DNA based on control assays (Supp. Figure 1C) that show this occurs rarely (

• On page 10, at the end of the first paragraph, growth phenotype of pGAL-SCM3 and pGAL-PSH1 mutants were tested. On GAL plates, pGAL-PSH1 shows reduced growth, but not pGAL-SCM3. Wouldn't a more accurate conclusion be that Psh1 is limiting stable centromeric nucleosome formation, instead of Scm3?

The growth defects on galactose don’t necessarily mean that a factor is limiting in cells. Instead, they report on whether changing the relative amounts of the complex lead to phenotypes in cells that could be the result of many causes that would require characterization of the phenotypes to understand. In this case, we presume that Psh1 titrates Scm3 away from Cse4 to prevent nucleosome formation in vivo. However, we have not directly tested this so we just concluded that the relative levels of the complexes are important for cell growth.

• In the section where DNA was tethered at either one or both ends, an important control is missing. How does such a set-up impact nucleosome formation in general. Can H3 nucleosomes form on random DNA that is either tethered at one or both ends? Does this potentially affect the unwrapping potential/topology of AT-tract DNA? Please comment.

This is an interesting point and one that we hope to explore further in the future but was beyond the scope of this paper. We suspect that restrictions via tethering would also limit canonical nucleosome formation on random DNA. We envision that unwrapping may be affected as well and hope to explore this via other, potentially better suited techniques like optical tweezers.

Minor comments + Censored data points are not explained in the text.

A brief explanation of censorship was added to the figure legends and we have now provided a clear explanation of the term “censored” in the text on p. 28, lines 25-27. It has also been reported in the Statistical tests section of the methods section to address this point.

• Number of particles tested should be reported in the main and supplemental figures, not just the legends for those readers who first skim the manuscript before deciding to read it.

We add these values to all Kaplan-Meier plots in all figures (Main, Expanded View, and Appendix)

• Typo on page 5, first line: "nucleosom" should be "nucleosome".

We fixed this in the text.

• Typo on page 9, second line: sentence is missing something "... is required for Scm3-dependent ..."

We fixed this in the text.

• It is unclear how the difference in Supplemental figure 5D was calculated.

We included log-rank test generated P-values as well as description in the figure legend of EV4.

• Figure 4C: why are there more Ndc10-mCherry foci observed in double tethered constructs vs single tethered constructs?

There can be variances in DNA density between slides, particularly with non-dye labeled DNA template. We updated figure panel C to include a representative image with similar Ndc10 density.

• For the display of individual traces as shown in Fig 2B, 3A, 4E, and 5E, it might be more informative if the z-normalized signal and the binary read-out are shown in separate windows to better appreciate how the z-normalized signal was interpretated.

Due to spacing limits within figures we attempted to accommodate this by reducing the thickness of the binary read-out and ensured that the raw data traces were overlaid for easier interpretation by the reader.

• Page 17, fifth line of the second paragraph, it is stated that a conserved feature of centromeres is their AT-richness. This is most certainly true for the majority of species studied thus far, but bovine centromeres for instance are about 54% GC rich. Indeed, Melters et al 2013 Genome Biol showed that in certain clades centromeres can be comprised of GC-rich sequences. It might be worthwhile to nuance this statement.

We have updated the text to reflect that AT-rich DNA is widely conserved but not a universal feature of centromeres.

• Page 17, last paragraph. Work by Karolin Luger and Carl Wu is cited in relationship to AT-rich DNA being unfavorable for canonical nucleosome deposition. A citation is missing here: Stormberg & Lyubchenko 2022 IJMS 23(19): 11385. Also, the first person to show that AT-tracts affect nucleosome positioning are Andrew Travers and Drew. This landmark work should be cited.

We thank the Reviewer for noticing this and have added the appropriate citations.

• Page 18, 9th line from the top, it is stated that yeast centromeres are sensitive to negative genetic drift. This reviewer is of the understanding that genetic drift is a statistical fluctuation of allele frequency, which can result in either gain or loss of specific alleles. Population size is a major factor in the potential power of genetic drift. The smaller a population, the greater the effect. Budding yeast is found large numbers, which would mean that drift only has limited predicted impact. Maybe the authors meant to use the term purifying selection?

We appreciate this clarification and agree with the reviewer, we have updated the manuscript to cite purifying selection and not genetic drift as at centromeres.

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 #3

Evidence, reproducibility and clarity

The manuscript "Single molecule visualization of native centromeric nucleosome formation reveals coordinated deposition by kinetochore proteins and centromere DNA sequence" by Popchock and colleagues describes a new high-throughput single-molecule technique that combines both in vitro and in vivo sample sources. Budding yeast centromeres are genetically defined centromeres, which makes them ideal for studying short DNA segments at the single-molecule level. By flowing in whole cell lysates, Cse4 nucleosomes can form under near physiological conditions. Two analytical experiments were performed: endpoint and time-lapse. In the former case, nucleosomes were allowed to form within 90 minutes and the latter case, nucleosome formation was tracked for up to 45 minutes. In addition, well described genetic mutants were used to assess the stability of Cse4 nucleosomes, as well as different DNA sequences (this we particularly liked- well done). Overall, this is a very interesting technique with potential to be useful for studying any DNA-based effect, ranging from DNA repair to kinetochore assembly. This is strong and impactful work, and the potential this kind of microscopy has for solving kinetic problems in the field. We think it's worthy of publication after revising technical and experimental concerns that would elevate the ms significantly.

Major comments:

Statistics are highly recommended for all the data in the ms.

• At what rate is data collected in the TIRFM setup. For clarity for the reader, it is important to provide imaging details for time-lapse. What is the impact of photobleaching on the stability of the fluorophore signal? Please clarify.
• The power of single-molecule technique is precisely that such data can be made quantitative. Indeed, the Kaplan-Meyer graphs do show nice quantitative results. Unfortunately, in the text few quantitative measurements are reported. In fact, the Kaplan-Meyer graphs can be interpreted in a quantitative manner such as probability of residency time. Although in most cases the statistical significance between two conditions can be expected, this is not formally calculated and shown. What is the 50% survival time of Cse4 alone or with Ndc10, for instance? This manuscript would greatly benefit from a quantitative approach to the data, including a summary table of the results of the various conditions tested. Please add this important table.
• This reviewer would expect that the endpoint (90 min) would roughly reflect the occupancy results from time-lapse (45 min) experiments. Based on the data presented in Figures 1, 2, S1-3, this does not appear to be the case. 50% of Cse4-GFP and 70% Ndc10-mCherry colocalized with CEN3 DNA in the endpoint experiment, whereas in Fig 2C, ~30 and ~52 traces were positive for Cse4-GFP and Ndc10-mCherry, resp. with the former having drastically lower residency survival compared to Ndc10-mCherry. If indeed, 50% of Cse4-GFP makes it to the endpoint, about 50% of all traces should reach the end of the 45 minutes time-lapse window. Only about 1/3 of all positive Cse4-GFP traces can be seen at the end of the 45 min window. Could this be due to technical issues of photostability of GFP? Why does the colocalization of Cse4 signal with the DNA signal disappear so readily? Are Cse4 so unstable? Is the same true for canonical H3 nucleosomes? This unlikely true for nucleosomes in cells. Along the same lines, in Suppl Fig 3 there is a disconnect between residency survival and endpoint colocalization on either CEN3, CEN7, or CEN9. What could be the underlying mechanism between the discordance of endpoint results and time-lapse results? Could this be the result of experimental differences?
• What fraction of particles show colocalization between Cse4-GFP and Ndc10-mCherry? What fraction of occupancy time show colocalization between Cse4-GFP and Ndc10-mCherry? Altogether, understanding the limitation and benefits of endpoint analysis and time-lapse analysis in this particular experimental set-up is important to be able to interpret the results. Please clarify these points.
• Page 9, third sentence of third paragraph it is stated that the "results suggests that Scm3 helps promote more stable binding of Cse4 ...". This is indeed one possible explanation of the results, and this possibility is tested by overexpressing Psh1 or Scm3 by endpoint colocalization analysis. 1) Taking the concerns regarding the endpoint vs time-lapse results into account, wouldn't it be more accurate to compare either time-lapse results against each other or endpoint results? 2) Alternatively, more stable Cse4 particles are able to recruit Scm3, simply because of the increased binding opportunity of a more stable particle. In this scenario, just the residency time of Cse4 alone is the predicting factor in likelihood to associate with Scm3. To test the latter possibility, Cse4 stability would need to be altered. Please consider this experiment- should be relatively easy with the right mutant of either CSE4 or CDEII (see Luger or Wu papers).
• In Figure 1C and Supplemental Figure 5B, there appears to be foci that CEN3-ATTO-647 positive, but Cse4-GFP negative and visa verse. It seems logical that there are DNA molecules that didn't reconstitute Cse4 nucleosomes. But how can there be Cse4-GFP positive foci without a positive DNA signal? Is it possible that unlabeled DNA make it onto the flow chamber? If so, can these unlabeled DNA be visualized by Sytox Orange for instance to confirm that no spurious Cse4 deposition occurred? Please clarify.
• On page 10, at the end of the first paragraph, growth phenotype of pGAL-SCM3 and pGAL-PSH1 mutants were tested. On GAL plates, pGAL-PSH1 shows reduced growth, but not pGAL-SCM3. Wouldn't a more accurate conclusion be that Psh1 is limiting stable centromeric nucleosome formation, instead of Scm3?
• In the section where DNA was tethered at either one or both ends, an important control is missing. How does such a set-up impact nucleosome formation in general. Can H3 nucleosomes form on random DNA that is either tethered at one or both ends? Does this potentially affect the unwrapping potential/topology of AT-tract DNA? Please comment.

Minor comments

• Censored data points are not explained in the text.
• Number of particles tested should be reported in the main and supplemental figures, not just the legends for those readers who first skim the manuscript before deciding to read it.
• Typo on page 5, first line: "nucleosom" should be "nucleosome".
• Typo on page 9, second line: sentence is missing something "... is required for Scm3-dependent ..."
• It is unclear how the difference in Supplemental figure 5D was calculated.
• Figure 4C: why are there more Ndc10-mCherry foci observed in double tethered constructs vs single tethered constructs?
• For the display of individual traces as shown in Fig 2B, 3A, 4E, and 5E, it might be more informative if the z-normalized signal and the binary read-out are shown in separate windows to better appreciate how the z-normalized signal was interpretated.
• Page 17, fifth line of the second paragraph, it is stated that a conserved feature of centromeres is their AT-richness. This is most certainly true for the majority of species studied thus far, but bovine centromeres for instance are about 54% GC rich. Indeed, Melters et al 2013 Genome Biol showed that in certain clades centromeres can be comprised of GC-rich sequences. It might be worthwhile to nuance this statement.
• Page 17, last paragraph. Work by Karolin Luger and Carl Wu is cited in relationship to AT-rich DNA being unfavorable for canonical nucleosome deposition. A citation is missing here: Stormberg & Lyubchenko 2022 IJMS 23(19): 11385. Also, the first person to show that AT-tracts affect nucleosome positioning are Andrew Travers and Drew. This landmark work should be cited.
• Page 18, 9th line from the top, it is stated that yeast centromeres are sensitive to negative genetic drift. This reviewer is of the understanding that genetic drift is a statistical fluctuation of allele frequency, which can result in either gain or loss of specific alleles. Population size is a major factor in the potential power of genetic drift. The smaller a population, the greater the effect. Budding yeast is found large numbers, which would mean that drift only has limited predicted impact. Maybe the authors meant to use the term purifying selection?

Significance

This study developed an in vitro imaging technique that allows native proteins from whole cell lysates to associate with a specific DNA sequence that is fixed to a surface. By labeling proteins with specific fluorophore-tags colocalization provides insightful proximity data. By creating mutants, the assembly or disassembly of protein complexes on native or mutated DNAs can therefore be tracked in real time. In a way, this is a huge leap forward from gel shift EMSA assays that have traditionally been used to do comparative kinetics in biochemistry.

This makes this technique ideal for studying DNA binding complexes, and potentially, even RNA-binding complexes. This study shows both the importance of using genetic mutants, as well as testing the effects of the fixed DNA sequence. As many individual fixed DNA molecules can be tracked at one, it allows for high-throughput analysis, similar to powerful DNA curtain work from Eric Greene's lab. Overall, this is a promising new single-molecule technique that combines in vitro and ex vivo sample sources, and will likely appeal to a broad range of molecular and biophysics scientists.

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

Centromeres drive chromosome inheritance from one cell generation to the next, so understanding their nature is of utmost importance in biology. The assembly of centromeric chromatin is of outstanding interest since defects in this step impair faithful genetic inheritance. Popchock et al investigate the molecular mechanisms of Cse4(CENPA) deposition and stabilization on a native budding yeast centromeric DNA. Using TIRMF (Total Internal Reflection Fluorescent Microscopy) enabled single molecule visualization of de novo Cse4(CENPA) nucleosome formation from a yeast cellular extract. The centromeric DNA used in this study was derived from PCR amplification of plasmids containing the native yeast point centromere from chromosome 3 (CEN3 DNA) sequences containing the CDEI, CDEII, CDEII, flanking pericentromeric DNA and linker DNA totaling ~750bp. In the PCR preparation, the CEN3 DNA contains a 5'-fluorescent dye and a biotinylated 3'-end, and was tethered to a functionalized (streptavidin) slide for TIRFM. The yeast extract was extracted from cells arrested in mitosis. This system is a novel application of a single molecule to study Cse4(CENPA) formation on centromeric DNA.

Using this system, the authors observed coordinated Cse4(CENPA) deposition on the CEN3 DNA, reporting inherent, transient colocalization of Cse4(CENPA) with CEN3 DNA. Stable Cse4(CENPA) colocalization on the CEN3 DNA is correlated with the Cse4(CENPA) chaperone Smc3(HJURP), and an ability for nucleosome formation on the CEN3 DNA. Further stabilization of Cse4(CENPA) was shown to depend on the DNA binding CCAN protein chl4(CENPN) and okp1(CENPQ) which dimerizes with the DNA/CENPA binding Ame1(CENPU).

Using this single molecule system they also demonstrated a role for the A/T run (>4) content in the CDEII as specifically important for Cse4(CENPA) deposition on CEN3 DNA. Cse4(CENPA) colocalization was preferred on the native CDEII sequence, relative to mutant CDEII sequences with similar A/T content but variable homopolymeric runs.

Major comments:

1. Figure S1A-D seem like some of the most compelling data in the paper to bolster the rigor of their experimental setup. There appears to be plenty of space to include these data in the main figure set in Figure 1 after panel D. The authors would be well served to consider moving S1A-D somewhere in the main figure set.
2. The authors conclude that Ndc10 recruits HJURP(Smc3) to the yeast point centromeres. If this is the case, can the TIRFM assay measure ternary residence lifetimes complexes between Ndc10/HJURP(Smc3)/CenDNA?
3. Throughout the manuscript short- and long- term residence lifespans are mentioned, referencing the figures containing lines with lengths depicting residence times. This is a qualitative reference to short and long residences. Can the authors provide a graph for short-term (<120 s) and long-term (> 300 s) residence life-spans for, CENPA alone, CENPA/Ndc10, and CENPA/HJURP on CEN3 DNA? Or some figure similar to Figure 3C, but reporting the proportion of short-term vs long-term residence?
4. The choice of CCAN components for analysis in Fig. 5 is interesting, but many readers may be curious why Mif2 wasn't selected for disruption, since it has such a cozy placement with CENP-A and CEN DNA. Can choice to not include Mif2 mutants/degrons be mentioned/justified in the text (unless, even better yet, they choose to address Mif2 role directly in new experimentation)?

Minor comments:

1. Are these whole cell extracts (WCE) DNA-free? I'm curious if there is any competition from endogenous DNA from the yeast cellular extract.
2. In relation to Mif2 and comment #4 above, do the authors make any connection to their results with synthetic nucleosome sequences not being conducive to yeast centromere formation with the prior observation (Allu et al 2019) using recombinant components that the human version of Mif2 more easily saturates its binding sites on CENP-A nucleosomes when they are assembled with natural centromere DNA rather than the Widom 601 sequence?
3. Providing a gel (or other measure) of the DNA templates (750, 250, and 80 bp) used in TIRFM assay would be nice to show to confirm the designed size of the pre-tethered DNA.
4. Some of the references to figures/figure panels in the main text do not match the figures. (discussion, pg 16, paragraph 1 & 2;pg 18, paragraph 1).

Significance

This work demonstrates the dynamics of Cse4(CENPA) coordination with Smc3(HJURP) to form nucleosomes on point centromeric DNA, and the necessity for homopolymeric A/T runs. It is a truly impressive piece of work that makes sense of findings from prior genetics experiments. Then it extends the understanding and clarifies the role of both centromere proteins and DNA sequence. The quantitative and powerful single molecule-based experimentation, the high importance of the subject matter, and its connection to studies using yeast genetics, will make this work, upon modest improvements (see section A of this review), of outstanding interest to an extremely broad audience of biologist.

My relevant expertise keywords: centromeres, nucleosomes, biochemical reconstitution, chromosome engineering

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

Summary

This work presents a novel experimental setup to explore native centromeric nucleosome formation at a single molecule level with a high degree of temporal resolution. It combines TIRF microscopy with an immobilized fluorescent yeast CEN3 DNA sequence. Binding of centromere proteins was probed by using whole-cell lysates produced from various mutant yeast strains expressing fluorescently tagged centromere proteins. This system was used to investigate the recruitment of Cse4 (yeast CENP-A), the histone H3 variant which defines centromeres, to the native yeast CDE3 sequence. Consistent with previous studies, the authors identified a requirement for Ndc10, which binds the CDE3 sequence with high efficiency, in Cse4 recruitment. Colocalization of Ndc10 with Cse4 significantly increases the stability of the Cse4-CDE3 interaction, and Ndc10 binding predominantly precedes Cse4 recruitment. Secondly, they investigated the impact of Scm3 (yeast equivalent of HJURP) on Cse4 stability and found that Scm3 is not required for transient short-term associations between the DNA and Cse4 but is needed for longer more stable interactions between Cse4 & the CDE3 sequence. Overexpression of Scm3 & its negative regulator Psh1 also showed that Scm3 availability is the rate-limiting factor in stable Cse4 associations with DNA. However, these interactions were abolished when the CDE3 DNA sequence was topologically restricted by tethering both ends to the streptavidin-coated coverslip suggesting that the formation of a nucleosome is essential for stable binding of Cse4 to CEN3. The sequence-specific DNA binding kinetochore proteins ChI4 & Okp1 were also found to improve the lifetime of Cse4-CDE3 interactions suggesting a stabilizing interaction. Finally, the impact of A/T runs on nucleosome formation was examined using a library of scrambled CDEII sequences, with approximately native levels As & Ts but shorter runs of the 2 nucleotides. Previous studies found a correlation between the length of AT runs and the incidence of chromosome missegregation. They now extend this correlation to a decreased ability of DNA to stably recruit Cse4. Finally, experiments with other sequences including the Wisdom601 sequence suggest that A/T runs along with CBF3 binding inhibit the recruitment of histone H3 to the CDEII sequence which is overcome by the interactions between Cse4 and Scm3.

Assessment

This is an insightful, well-executed study of a high technical standard. While several observations confirm or further validate previous findings in vivo (such as the requirement for Ndc10 and Scm3 for Cse4 assembly), this work adds a better understanding of the dynamics of Cse4 requirement and the role of Scm3 in assembly and the role of CCAN components in stabilizing Cse4. Figures are well laid out and methods are clearly described. In general, the claims are supported by the data. The development of their single-molecule setup to assay Cse4 nucleosome assembly is a promising tool for future work and for this reason work reporting. In short, I believe this paper contains exciting developments in the understanding of the specific mechanism and temporal dynamics of the formation of a Cse4 nucleosome on its native DNA and the interactions which underpin its stability in vitro and in vivo.

Major comments

I have several comments and suggestions for further analysis of the data. I also have suggestions for additional experiments, but I would stress that none of these are essential for publication.

Figure 2E. This shows the residence time of Cse4 without Ndc10 association. How does this compare to the residence time on mutant CEN3 (Supplemental Figure 1). It looks like Cse4 still binds to CEN3 with some specificity even in the absence of Ndc10. Does this suggest that Cse4 has some intrinsic ability to recognize CEN3? Alternatively, Ndc10 is still required for Cse4 binding but is below detection in the Cse4-alone residence lifetimes. Ideally, the authors would compare this with Cse4 binding in an Ndc10 mutant.

Figure 3 explores the very interesting relationship between Scm3 dynamics and Cse4 binding but I feel that the data is not fully flushed out. A key finding is that Cse4 can potentially bind to CEN DNA prior to engaging with Scm3 to be incorporated. This predicts that Cse4 should bind first and then colocalize with Scm3. Can this be detected in the timing of the traces? How often does Scm3 bind to already prebound Cse4 and does this correlate with Cse4 residing longer?

A related and perhaps even more interesting point is whether Scm3 is involved in "loading" of Cse4. If so, then one would expect that once Cse4 is assembled into nucleosomes it should be stable, even if Scm3 leaves. Can the authors extract this from the data? Alternatively, it is possible that Scm3 remains associated to Cse4 to maintain the nucleosome which would imply a more extended role for Scm3 apart from assembly alone. It would be interesting if information on this can be extracted from the data.

Even in the presence of Scm3 and CCAN components, Cse4 appears to have a limited lifetime in the in vitro assay compared to in vivo stability. The authors should speculate on whether activities exist in their extract that actively disassembles nucleosomes. Perhaps ATP could be depleted to inactivate remodellers?

For Figure 6, it is not clear why AT-track mutants of CDEII are labeled as genetically stable and genetically unstable. This is confusing as the "genetically stable" show a more than 10-fold increase in chromosome loss rates. Perhaps these can be renamed into "unstable" and "very unstable" or "weak" and "strong" mutants, just to make clear that these classes are both poorer than wild type.

Finally, it would be wonderful to include data to assess whether a full Cse4 nucleosome is assembled or a partial nucleosome e.g. just Cse4/H4 tetrasomes. This could be done by tracking the accumulation of H2A or H2B at the CEN3. This would give further insight into what step Scm3 catalyses.

Minor comments:

Typo on page 5, line 1 "nucleosom" missing an e.

Kaplan-Meyer should be spelled Kaplan-Meier

The term "censored" is mentioned across many figures but comes up just ones in the methods where it is not clearly explained. Perhaps this could be clarified in the legend.

The abstract states that Cse4 can arrive at the centromere without its chaperone. More likely, Cse4 is in complex with other chaperones that may allow it to bind. Perhaps the abstract can be modified to read "Cse4 can arrive at the centromere without its dedicated centromere-specific chaperone Scm3..."

Related to this point, the discussion states the possibility that Cse4 can initially bind to CEN3 via other more general chaperones. However, it should be acknowledged that transient Cse4 binding in their assay may simply occur through mass action due to high concentrations of CEN3 DNA. In vivo, this transient binding may not be that relevant.

Significance

This paper offers new insight into the assembly of yeast centromeres with a focus on the role of the Chaperone Scm3 in the assembly of the centromere-specific histone Cse4. This is still a poorly understood process and the authors offer an elegant in vitro system to study this and have presented new insights. For this reason, this is study is of interest to a broad readership in the area of mitosis and chromosome structure. The advance is strong at the technical level but also new insight is provided particularly in the role of Scm3 and the nature of centromeric DNA in centromeric chromatin assembly. Overall a strong, high-quality paper.

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

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

The authors do not wish to provide a response at this time.

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 #3

Evidence, reproducibility and clarity

Summary

In this manuscript, Wang and colleagues demonstrate that a single systemic injection of a high dose of Akkermansia muciniphila (A.m.) lysate drives a rapid pancytopenia followed by prolonged anaemia and hepatosplenomegaly with late-onset extramedullary hematopoiesis (EMH). The latter, as well as the splenomegaly, were likely mediated through activation of pattern recognition receptors and IL-1R signalling pathways. This was demonstrated through the partial and full phenotype reversal in Tlr2;4-/- and MyD88;Trif-/- mice, respectively. Moreover, the phenotype was partially reversed following IL-1R antagonism. After performing multiplex protein assays and flow cytometry, the authors conclude that EMH, in this model, is mediated by IL-1a produced within the spleen by local monocytes and DC.

Overall, the manuscript by Wang et al. is quite well presented, the experiments are mostly well controlled, the methods are well reported, and the data fit a clearly defined story with clinical relevance. Nevertheless, there are several major concerns that if addressed would greatly increase the strength of the authors conclusions.

Major comments

1. The "two wave" hypothesis of hematopoiesis - first in the bone marrow (BM) and then in the spleen - is interesting. However, although an early wave of BM hematopoiesis would make sense, under these circumstances, I don't think the data are strong enough to support this hypothesis as they stand. For example, although the frequency of LSK cells increase, the numbers of most LSK subsets decrease. Given the decrease in the absolute number of BM cells 1d after A.m. injection, isn't it possible that the LSK cells are only proportionally increased relative to the remaining Lin- cells? What happens to the absolute number of LSK cells following A.m. injection?

Also, describing "two distinct waves of HSPC increase in the A.m.-injected spleen" (Fig 2 & S2 titles) and describing a "first wave" of HSPC expansion in the BM (lines 396, 399, 402 etc.) is misleading for the following reasons: (i) the data strongly support a single wave of increasing HSPC in the spleen, peaking at d14, and (ii) there is no evidence HSPC are increased in the BM until d56, although there does appear to be an early increase in MPP. The language should be changed accordingly. 2. The flow cytometry panel is not comprehensive enough to fully characterize the mature hematopoietic cell populations to the levels that are claimed here. For example, it is a stretch to assume that all B220- CD3- CD11c- cells are DC (splenic NK cells, eosinophils, monocytes and red pulp macrophages, for example, can express CD11c, particularly following inflammatory insult), or that CD11b+ F4/80+ SSC-hi cells are eosinophils, especially when eosinophils should be F4/80-lo are not known to express Ly6C in the spleen (For reference, see Immgen). These gating issues may explain the conspicuous absence of macrophages (should be F4/80+CD11b+Ly6C- and would also have a higher SSC than monocytes) in the plots. The B cell gate will also contain PDC, which express B220 (but can be easily excluded using Ly6C and CD11c). With respect to assessing the mature leukocyte populations in the spleen, relabelling the gates (CD11c+ cells instead of DC, F4/80+ myeloid cells instead of eosinophils) would suffice, however, these issues become a problem when trying to identify which cell populations express IL-1a.

Due to the limited antibody panel used here, there is not enough evidence to suggest that DC and monocytes are producing IL-1a. Moreover, the histograms showing the changes in expression of IL-1a on the "DC" and "Mo" are not very convincing. How does the IL-1a staining look on a dot plot? Is there good separation between positive and negative? These plots need to be included. What happens if you gate on the IL-1a+ cells first, then phenotype them?

Macrophages and splenic stromal cells are also likely candidates for IL-1a production. To assess which cell types are the true source of IL-1a, the authors need to repeat these experiments (namely, injecting A.m. and assessing IL-1a expression by leukocytes (and ideally also mesenchymal cells)) at d1 and d14, using a more comprehensive panel. Consider adding MHCII, CD64, Siglec F and CD24 to help differentiate between DC, MF, eosinophils and monocytes. CD45+ vs CD45- could be used as a minimum to assess the expression of IL-1a on leukocytes vs. stroma.

OPTIONAL: The mechanism could be better defined using bone marrow chimeras to assess the different contribution of TLR2/4 signalling and IL-1R signalling on the hematopoietic vs. mesenchymal cell compartments. 3. From these experiments, it is difficult to fully rule out a contribution from the adaptive immune system to the splenomegaly phenotype due to the marked difference in the size of BALB/c and MSTRG spleens at steady state. The authors should show the differences in spleen weight and total cell number as a % increase from control. The no of HSPC should also be normalized per gram of tissue weight or represented as a fold change compared to the relevant control groups. 4. When using fluorescent imaging to compare the abundance of HSPC and other cell populations in the spleen, the authors should provide absolute quantification from multiple FOV and multiple mice. 5. Finally, although the experiments are adequately replicated, the stats are not always appropriate. For example, a t-test shouldn't be used when there are >2 groups, or for a time course. This needs to be amended.

Minor comments

• Line 82-83: I'm fairly certain monocytes and inflammatory Ly6Chi cells are the same thing.
• Line 83-84: "IL-1a is crucial for sustaining inflammatory responses, recruiting myeloid cells to infected tissue and inducing hematopoietic stem and progenitor cell (HSPC) mobilization and expansion both in vitro and in vivo" - I don't believe IL-1a has been shown to be crucial for either, even if it has been shown to play a role. If I am mistaken, please reference with a manuscript showing relevant phenotypes using KO mice.
• Line 214: "Thus, we decided to use 200ug of lysate for the rest of all experiments." - is this what was usen for Figures 1A-C? This is not mentioned anywhere.
• Line 227: "containing both non-hematopoietic cells and immature HSPCs" Please reference Fig. 1H here. Otherwise, it is unclear how you identified the "HSPC and other cell types" in Fig. 1G.
• Figure S2A is described in text before Supp 1I-O and Fig S1H is not referenced in text at all.
• It would be interesting to include what happens to hepatomegaly in MSTRG, Tlr2;4-/- and MyD88;Trif-/- mice.
• Please define WBM. Presumably whole bone marrow?
• Notably, CCL2 is increased in spleen lysate, BM lysate and serum. Given is role in myeloid cell mobilization from the BM, I would expect its role in the phenotype described here to at least be discussed.
• HSPC LT gate includes MPP1, and should be labelled as such.

Significance

General assessment: The manuscript provided by Wang et al. describes, for the first time, a prolonged anaemia and hepatosplenomegaly with late-onset extramedullary hematopoiesis following a single systemic injection of A.m. lysate. The EMH phenotype appears robust and the data implicating TLR-signalling and IL-1a production are compelling. The work has clinical relevance as it increases our understanding of the factors driving EMH.

There are two key limitations that let this study down. Firstly, the lack of depth in the flow cytometry panel used for immunophenotyping means it is not at all clear which cell types are producing IL-1a. Secondly, the authors use an enormous dose of bacterial lysate that is well above physiological levels, even following a loss of barrier integrity (e.g., in patients with IBD). This makes me question the biological relevance of the study, particularly with respect to Akkermansia translocation.

Advance: With some improvement, this study will advance the field, in general. Previous work has looked at EMH following LPS injection, or live E. coli infection, however; the authors are able to demonstrate a distinct Akkermansia-specific effect that differs to that of LPS, membrane components of a different gram-negative bacteria, B. theta. The advancements implicate IL-1a in the modulation of EMH, for the first time, providing some mechanistic insight into this phenomenon.

Audience: This work will likely be of interest to basic researchers interested in EMH. It may also be of interest to clinical researchers of pathologies where EMH is a known complication, such as rheumatoid arthritis and cancer. The impact of the work will depend on whether or not EMH contributes to pathogenesis, or is an epiphenomenon. To my knowledge, this has not been fully established, although this is not my area of research.

I am a basic researcher with expertise in immunology focused on host-microbe interactions, both within the intestine and at distal tissues. I have knowledge of BM hematopoiesis and the microbial factors that influence if although my knowledge on extramedullary hematopoiesis is limited.

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 study, Wang et al reported single injection of Akkermansia muciniphila (A.m.) induces two waves of extramedullary hematopoiesis (EMH), and demonstrated the mechanism of "second wave" was contributed by IL-1α secretion in spleen DCs and Ly6C+ monocytes. It is an important work on understanding infection-induced EMH. However, several major concerns about claims in this manuscript need to be addressed.

1. The authors demonstrated that A.m.-induced EMH were alleviated by knockout of Tlr2;4 or Myd88-Trif, or even IL-1R inhibition. EMH in the spleen is a mechanism by which the hematopoietic system responds to stresses. Therefore, whether inhibition of EMH by these ways can affect normal hematopoiesis in mice? Do mice have pancytopenia? Will the function of HSC in bone marrow be affected?
2. In Fig.S1E, why did the WBC and PLT recover quickly after the first day, while the RBC took 14 days to recover? Are WBC and RBC regulated by two waves of EMH, respectively?
3. The authors should show absolute numbers of each cell type, not just the percentage of immunophenotypically defined cells. For example, in Fig.1G, I, 2B.
4. In transplantation assays, whole BM cells or splenocytes was use. However, the proportion of HSCs in BM and spleen were all changed post A.m injection, which could affect the outcome of chimerism rate after transplantation. Transplantations should be done on by transplanting sorted fresh immunophenotypic HSCs.
5. The concentration of IFN-γ in both spleen and serum were increased continuously from D1 to 14. Would IFN-γ cause the second wave of EMH? Relevant assays for exclusion are necessary.

Significance

It is an important work on understanding infection-induced EMH.

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

Wang et al. examined the mechanisms of splenic extramedullary hematopoiesis upon systemic injection of Akkermansia muciniphila in mice. They showed that components of this mucin-degrading bacterium mobilize bone marrow hematopoietic cells and induce splenomegaly by MYD88/TRIF-dependent innate immune signaling pathways. Activation of TLRs and release of interleukin 1-alpha from splenic cells were then responsible for the expansion and differentiation of functional hematopoietic progenitors in the spleen. Genetic deletion of TLR2 and 4 restrained splenomegaly, while, pharmacological inhibition of IL1 receptor abrogated splenomegaly and extramedullary hematopoiesis suggesting their cooperation in the observed phenotype. It has been widely accepted that splenomegaly arises as a consequence of inflammation and that TLRs are major drivers of this process in the context of bacterial or viral infections. Here, the novelty relies on the potential circuit with the IL1alpha-IL1R axis as an additional driver of splenic extramedullary hematopoiesis. Although the results summarized above indicate that both TLRs and IL1 individually participate to some extent in splenomegaly after Akkermansia muciniphila administration, they fail to demonstrate that they concertedly do so in the spleen. In fact, the blockade of IL1R has a more profound impact.

Significance

My concerns are the following: - The authors mentioned that a specific lipid from Akkermansia muciniphila is able to trigger a non-canonical TLR2-TLR1 heterodimer to release inflammatory cytokines and regulate human immune response. Why TLR1 was not considered in the experimental strategy? - IL1alpha is up-regulated in several splenic cells (in particular in macrophages, Fig. S4F). To demonstrate a critical involvement of dendritic cells or monocytes, depletion studies or conditional mice models should be evaluated. What about megakaryocytes? Why were excluded from the analyses? - Interleukin 1alpha KO mice model should be also evaluated. - The paper is very dense and not easy to read and follow. English editing is required.

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

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

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

The manuscript presents a detailed numerical model of blood flow in a region of the zebrafish vasculature.

The results section is quite intense and detailed. it is difficult to understand what the authors are after. I think a rewrite would beneficial. The authors present simulations for a wild type and a couple of phenotypes. For each of these they speculate on the possible adaptation mechanism leading to the discussed phenotype, as preservation of constant wall shear stress. However, the comparison between experiments and numerical simulations is really elusive as the conclusions on those mechanisms. Overall we suggest a rewrite with clearer organisation in a way that the reader is not overflown with useless details.

We thank the reviewer for the advice on the general writing standard and data organization. We have reanalyzed experiment data and interpreted the findings more conservatively for application into the simulation models. As a result, some conclusions to the results sections have changed. Accordingly, we have done a major revision of the entire Results, Discussion and Models and Methods sections in the paper to articulate these reinterpretations while removing superfluous details that may obfuscate the data.

It is not always clear what info of the experiments are used in the simulations on top of the anatomy. Our understanding is that the pressure boundary conditions are set to match the red blood cel velocity observed in experiments. Is this always the case for the three phenotypes and which vessels ?

We thank the reviewer for the question. Only WT and Marcksl1 KO have been matched for peak velocities in the CA, CV and ISVs between experiments and simulations. WT results were compared to both the experimental reference of 27 embryos in Table 3 and also to the current experiment pool of WT (5 embryos) in Table 6. Marcksl1 KO simulation models 1, 2 and 3 were compared against the average level seen in the low and moderate perfusion Marcksl1 KO phenotypes (8 embryos) from the experiment (Table 5 and Table 6). Additionally, we also have represented the similar level of RBC hematocrit in the CA for WT model to WT experiment data from the reference cited in Table 3.

In addition to the velocity comparisons, we now use the experimentally observed trend of decreased flow rate in the CA of Marcksl1 KO experiment data to assess the model boundary conditions amongst Marcksl1 KO models 1, 2 and 3 that best reflect the experimental observations:

Page 11, lines 1 to 20

The Marcksl1 OE cannot be matched because we do not have the experiment data for that, the same goes for PlxnD1 where we have no experiment flow data. These two networks represent more conceptual discussions, particularly in PlxnD1 case where we have explicitly stated in the new discussion section:

Page 15, lines 24 to 34

There are about 7 inlets and outlets where to impose pressure boundary conditions. Can the author comment on the uniqueness of this problem?

Can different combination of pressure boundary condition leading to the same result ? In how many points/vessels is the measured velocity matched ?

We thank the reviewer on this insightful concern. Indeed, the uniqueness of flow and pressure field can be a problem without careful consideration. We have tried to address this to some extent, because CA, CV are connected by the ISV and DLAV network, to match flow velocity in all regions, the pressure distribution ought to be unique to the particular setting we employed.

As shown in table 3, average systolic peak flow velocities in the entire CA and CV encompassing the 5 ISV segment domain is matched between the simulation and the population-averaged experimental data from the experimental reference (27 fish sampled in the cited reference) for the same regions in WT network. Average systolic peak flow velocities for the 10 ISVs in the simulation were matched against WT experiment population-averaged systolic peak flow velocities in arterial and venous ISVs in the same caudal region.

Additionally, we also compared the flow velocities to the experiment conducted within this study (5 WT, and embryos). This comparison data is shown in Table 6. Admittedly the discrepancy was large for CV and ISVs regions likely due to a smaller data set sampled in this study and biological variations that happen from one experiment to another. We have acknowledged this deficiency in the revised discussion section:

Page 15, lines 3 to 9

The argument that similar beating frequency in the WT and GATA1 MO suggest pressure does not change is not clear. If the heart was a volumetric pump it would impose the same flow rate, not the same pressure. It would be more useful to measure the cardiac output in terms of flow rate in the Dorsal Aorta. Previous measurements by Vermot suggested the latter would not change much in gata1 MO. It could be that the cardiac output is the same but the vasculature network is different in a way that the shear stress remain the same. It does not look like this was checked by the authors.

We thank the reviewer for this insight. In accordance with the reviewer’s suspicion, we have estimated the flow rates in the CA of gata1 MO injected embryos and found the level to be similar to WT. This supports the reviewer’s opinion that the heart rate similarity indicates cardiac output similarity and not arterial pressure similarity as we previously put forward. Furthermore, we have checked that the gata1 morphants do in fact present reduced ISV diameters. In light of this reinterpretation, we performed an additional zero hematocrit model (model 3 in section 2.1). We have consequently rewritten the entire section on how RBC hematocrit modulates hemodynamics in a microvascular network:

Page 6, line 18 to page 8 line 10.

Additionaly, it would be useful to provide an effective viscosity for the different vessels, and an effective hydraulic impedance relating DP and Q to interpret the results.

We have followed the reviewer’s advice and have analyzed for vessel hydraulic impedance and effective viscosity in all the network models presented. This is included in the main figures and discussion. The vessel impedances are discussed for the various models in these following parts of the manuscript:

Page 9, lines 20 to 29

Page 11, lines 28 to 30

Page 12, line 1 to page 13 line 10

Is the hydraulic impedance of the vessels kept constant in the smooth-geometry model? This needs clarification

The SGM diameters have been determined based on geometric averages and not impedance equivalency. The reason why we did this is because the impedance will not be known until the CFD is performed for the WT network. This is because without a pressure distribution (which cannot be determined experimentally) we cannot calculate vessel impedance since only flow can be measured and both flow and pressure are requirements to impedance calculation. Our intention with the SGM is to highlight how geometric averaging of morphological characteristics lead to incorrect flow and stress predictions. However, we understand the reviewer’s sensibility and have revised the entire section of the SGM results. We have now discussed three SGM models with varying degrees of geometry simplification. The SGM1 in the revised manuscript matches WT network impedance in the ISVs by including both axial variation in lumen diameter of the WT network and the elliptical fit representation of cross-sectional skewness seen in WT ISV lumens. SGM 2 has representation of axial variation but not luminal skewness and SGM3 has only geometric average similarity to WT ISVs. The new findings and discussion can be found in the revised manuscript here:

Page 8, line 19 to page 9 line 36.

As mentioned by the authors they propose a very complex and time expensive simulation. However the results they report are kind of intuitive. Given the availability of the experimental results, would it be useful to use a simpler red blood cell model in the future, to make their simulation more practical? Or clarify when such demanding simulations can add something new?

We agree that the intuition feedback depends on the expertise of the investigator. The boundary condition selection is intuitive from the experimental findings and key data like pressures in the network cannot be measured. Furthermore, population-averaged flow data does not always match the flow-to-geometry situations that vary from sample to sample, thus demonstrated by the high margin of prediction discrepancy for flow velocities in table 6. We have discussed these challenges and our recommendations for improvement in the Discussion section:

Page 15, lines 3 to 9

Page 15, lines 35 to 40

Page 16, lines 12 to 15

On the topic of RBC model simplification, we agree with the reviewer that our work suggests the methodology would benefit from a further coarse-graining approach to the RBC phase. Accordingly, we discussed the possibility of using a low-dimensional RBC model already published in literature:

Page 14, lines 13 to 17

The authors should check their references as this is not the first time work has been done on the topic. Would be good to have a check in the work of Freund JB and colleagues, as well as Dickinson and colleagues and Franco and colleagues to discuss how the work compares. There may be interesting work in modelling cardiac flow forces in the embryo too.

Thank you for referring us to other publications that are related to our study. To our knowledge and after performing publication search on these authors, we find that although Dickinson and colleagues performed experiments to examine the effects of perturbed blood flow on vessel remodelling (Udan et al., 2013), they did not perform any numerical modelling to calculate hemodynamic forces such as WSS and luminal pressure. Instead, changes in vessel morphogenetic process were only correlated with blood flow velocity. In our study, we attempt to quantitatively correlate WSS and pressure distributions within a vascular network. Franco and colleagues (Bernabeu et al., 2014) developed PoINet to model haemodynamic forces in mouse retina model of angiogenesis. From what we understand, PoINet is different from our 3D CFD model by 1) not having red blood cells incorporated in their model and as such, the blood viscosity prediction is modelled using shear-rate dependent formulation and not through red blood cell hematocrit, 2) cross sections of blood vessels are assumed to be circular and therefore have no irregularity and 3) live imaging of blood flow is difficult in mouse retina therefore preventing accurate boundary conditions for the model.

We have included the reference to work of Franco and colleagues:

Page 14, line 28 to line 31

Page 9, lines 12 to 14

Freund JB indeed has had extensive work on RBC and cellular flow in microvessels. We have included a reference of his work in:

Page 14, lines 22 to 25.

Reviewer #1 (Significance (Required)):

The authors discuss the applicability of a detailed numerical model of blood flow in a region of the zebrafish vasculature.

We are not expert in the lattice boltzmann method used here, but the results are what it would be expected from a physical stand point, and together with the information from the method section, we do not have major concerns about the numerics.

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

Summary: The authors report corroborating numerical-experimental studies on the relationship between morphological alterations (e.g. vessel lumen dilation/constriction, network mispatterning) and hemodynamical changes (e.g. variation in flow rate, pressure, wall shear stress) in the vascular network of zebrafish trunk circulation. Various physiological or pathological adaptation scenarios were proposed and tested, with a range of simulation and experiment models. Where I found it a solid piece of work supported by abundant data, certain aspects need to be clarified/enhanced to improve the scientific rigor and potential impact of the manuscript. Below are my detailed comments in the hope of helping the authors improve the manuscript's quality.

Major comments:

1. Cellular blood flow in vascular networks has been extensively studied in recent years by existing computational models (some of which were published open-source) with similar methods and features to the one proposed by the present work. Can the authors be more explicit about the original contributions of the current model, and provide evidence accordingly (e.g. Github repository or code resources)

The RBC model is essentially the model developed by Fedosov and colleagues (Fedosov, et al., 2010). Likewise, the LBM solver for fluid flow calculation is not. Following the reviewer’s advice, we have removed the details of these non-novel aspects of the methodology and placed them in sections E and F of supplementary material instead. The new Models and methods now show condensed descriptions of the three numerical solvers used and the addition of a grid independence matrix discussion section:

Page 17, line 8 to page 20, line 33.

Crucial details for the simulation setup and model configuration are missing. What were the exact boundary conditions (e.g. inlet and outlet pressures) and initial conditions (e.g. feeding hematocrit of RBCs), and how the numerical-experimental validation process of "to match the velocities of various segments of the network by iteratively altering the pressure inputs ..." as stated on page 13 (lines 1-2) was performed for simulations in this work?

We apologize for the vagueness of our description on how numerical to experimental validations were performed. As replied to reviewer 1 for a similar clarification, we have indicated in Table 3 how average systolic peak flow velocities in the entire CA and CV encompassing the 5 ISV segment domain were matched between the simulation and the population-averaged experimental data for the same regions in WT network. Average systolic peak flow velocities for the 10 ISVs in the simulation were matched against WT experiment population-averaged systolic peak flow velocities in arterial and venous ISVs in the same caudal region.

With regards to what iterative alterations of pressure inputs mean, we monitored the average systolic peak velocities and hematocrit levels in CA, CV and ISVs in intervals of 5 cardiac cycle intervals before manually correcting the pressure input levels to better match average systolic peak velocities in these vessels from the experiment averages. Since we are using population averaged flow data, we do not expect their levels to match the levels in a particular fish-specific geometry, the degree of discrepancy between experiment averages and the model predictions of systolic velocities can be large (Table 6). Admittedly, this is one of the weaknesses of our approach and this limitation is stated in the Discussion section:.

Page 15, lines 3 to 9

As RBC flow typically requires roughly 5 cardiac cycles of flow to reach flow development this process of iterative correction typically takes place over 10 to 20 cardiac cycles. We understand that validation may be a subject of keen interest to readers, hence we have now briefly described the solution initialization and flow development protocol in our modeling approach here:

Page 6, lines 5 to 8

What lattice resolution was used for the flow solver and was the RBC membrane mesh chosen accordingly? Were there any sensitivity analysis (regarding pressure input) or grid-independence study (regarding lattice resolution)

We originally decided on the grid (∆X) and time (∆T) discretization resolutions (0.5 µm and 0.5 µs) based on the acceptable computing turnaround time for each model within our scale of resources. We have now included a section on the grid independence matrix in Models and Methods:

Page 19, line 20 to page 20, line 33

Details of the statistical tests (type of tests used, assessment of data normality, sample size etc.) should be given in the figure caption where applicable (e.g. Fig. 3C, Figs. 7-9).

We apologize for the lack of clarity. All statistical tests used have now been mentioned at least once in each section of results and also in Figure captions wherever significance bars are displayed.

The regression models should also be used with caution, e.g. in Fig. 4B, why should data from two different fish types, namely Gata1 MO and WT, be grouped to fit a linear regression model?

We understand the reviewer’s concern that two population data sets should not be carelessly pooled together for regression analysis without adequate justification. In this case we are utilizing gata1 morpholino injection as a means to alter hematocrit level. There is no reported side-effect as to the best of our knowledge, only hematocrit and possibly hemodynamics and morphological response related to hematocrit level should be affected. Moreover, we have mislabelled the companion set to the gata1 morpholino as WT, the data is in fact data from control morphants and not WT. This change has been applied to Fig. 3 graphs and Table 4 and results section:

Page 7, lines 3 to 16

Finally, as we want to generate a continuum range of varying hematocrit for embryos of the same developmental age. In this regard, we think that within the scope of our intentions and well-accepted usage of gata1 morpholino as a hematocrit reduction protocol it is reasonable to pool the two data sets together for regression analysis.

4.I found the data presented in Fig. 7 insufficient to confidently exclude the numerical models 2, 3 but favor model 1 as the adaptation scenario for the Marcksl1KO case. The first question is, how are the threshold RBC perfusion levels determined to categorize the experimented Marcksl1KO fishes into four groups, i.e. "high", "moderate", "low", "zero"? The authors also need to justify why the "high", "moderate", "low" groups can be mapped to the three modelling scenarios (namely models 1, 2, 3) is it just because "a qualitative match with the experimental trend of ascending CA blood velocity" (Fig. 7F)?

We thank the reviewer for his interpretation of our results. Firstly, we apologize for generating the confusion but we are not trying to map simulation models 1, 2 and 3 to high moderate and low groups respectively in Fig. 7. The high, moderate and low categorizations of experimental Marcksl1 KO phenotypes are based on RBC flux levels observed experimentally. We are trying to ascertain which Marcksl KO phenotype the models 1, 2 and 3 fit, if they do fit the experiment trend at all.

Second, in Fig. 7C, it is shown that no significant difference exists between the "high" group and the WT in their average ISV diameter, then what is defining that group as Marcksl1KO type ?

We apologize for the confusion generated. High flow phenotype is similar to WT flow, the diameter is also similar to WT. In Marcksl1 KO mutants we don’t always see clear phenotyping and often a range is presented from mutant to mutant. Hence the high group is essentially morphometrically and hemodynamically similar to WT, the only reason we know it is a mutant because we have genotyped the zebrafish (marcksl1a-/-;marcksl1b1-/-).

Third, a central assumption here is using heart rate as a measure of the pressure drop in different fish individuals (Fig. 7D). Can't two fishes with similar heart rate have distinct pressure drops in the trunk due to difference in network architecture and topology, vice versa?

We agree with the reviewer’s opinion and now feel that our initial proposition was naïve. After addressing the interpretation of heart rate similarity in the gata1 morphants with more convincing CA flow rate estimations, we now believe that heart rates might not be useful indicators of flow or pressure levels in the network. Instead, cardiac output in the form of CA flow rate as reviewer 1 has suggested might be a better indicator. As the reanalysis has dismantled the earlier interpretation, and found that based on the flow rate estimation for the CA, Marcksl1 KO networks have reduced blood flow rates in the CA.

Page 11, lines 9 to 20

This finding has been incorporated into the consideration of flow adaptation scenarios predicted by the simulation models accordingly in the revised manuscript:

Page 12, line 1 to page 13, line 10

Fourth, the authors should explain why a power-law fit (note that it is not "exponential" as stated on page 10, line 3) should be adopted for the regression analysis in Figs. 7E-v,vi (a useful reference may be Joseph et al. eLife 2019: 10.7554/eLife.45077).

We thank the reviewer for the useful reference and the careless mislabeling of regression curve used. This figure has been redone and a linear regression is instead used that does not attempt to imply any physical law for a power or exponential fitting.

Change made: Fig. 7C

Minor comments:

1. The state of art of cell-resolved blood flow models employed to simulate microcirculatory hemodynamics is not accurately described in the introduction (page 4). More recent works should be cited and critically reviewed to present a fair view on the novelty of the computational model developed herein.

We apologize that the models were mentioned in a passing manner. However ,the need for brevity in introduction somewhat limits their expansion. We have instead gave more direct discussion on similar studies and their relevance to our present work in the Discussion section:

Page 14, lines 13 to 31

It is unclear what "realistic representation of local topologies in the network" (page 7, lines 28-31) means as a claim of novelty. If it means vessel "diameter variation", this geometric feature has been modeled by the works the author referenced (namely Roustaei et al. 2022, Zhou et al. 2021). If it means something else, for example, unsmooth or non-circular vessel surface (or "irregularity of the local endothelium surface" as mentioned on page 5, line 2), then strangely the effects of such features are actually not described in the manuscript.

We apologize for not meeting the expectation of novelty as claimed. We see value in the SGM study matrix have now generated data on three SGM scenarios. The SGM1 in the revised manuscript matches WT network impedance in the ISVs by including both axial variation in lumen diameter of the WT network and the elliptical fit representation of cross-sectional skewness seen in WT ISV lumens. SGM 2 has representation of axial variation but not luminal skewness and SGM3 has only geometric average similarity to WT ISVs. Essentially the comparison between SGM1 and SGM2 highlights the role of luminal cross-sectional shape skewness while SGM2 to SGM3 highlights the role of axial variation in luminal diameter. With this new SGM data set, we think we can better qualify the aspiration of demonstrating how vessel shape “irregularities” can alter network hemodynamics. The new findings and discussion can be found in the revised manuscript here:

Page 8, line 19 to page 9 line 36.

Why should Fig. 8 contain data from Marcksl1KO model 2? The scenario underlying model 2 was rejected earlier in the manuscript (see point 6 above), and the Marcksl1KO model 2 data are not mentioned in the text when describing the results of Fig. 8, either.

We have reanalyzed the experiment trend and rewritten the outcome of this results section. In summary, both models 1 and model 2 meet the trend of flow rate reduction (with respect to WT levels) in the CA observed in the experiment. Hence, model 2 inclusion is relevant to the WSS analysis. The changes pertaining to this can be found here:

Page 11, line 9 to page 13 line 10.

It is a dense article with loads of data, which is an advantage but only if appropriately streamlined. More subheadings should be considered, especially for section 2.3 (for which the current subsections appear mistaken, 2.3.1 followed by 2.4.2) The manuscript could also benefit from restructuring through optimal combination of simulation visualizations and quantitative analyses. For example, in Fig. 6, not all simulation snapshots are needed here (it is difficult to visually compare the changes between different cases), whereas some quantification in the form of histograms or boxplots will be handy for the readers to note the variation of WSS magnitudes and ranges.

Thank you for the advice, we removed the unnecessary graphical plots and refer to simulation videos in supplementary data instead for such cases. The bad indexing of results subsections has been fixed, while new subsections have been made for better directional narrative to the paper. These changes are colored in red throughout the revised results section:

Page 4, line 37 to page 13 line 39

Related to point 8, the authors could also consider integrating or synthesizing the analyses for individual aISVs and vISVs presented in various figures. Current descriptions for the ISV data appear scattered with frequent exceptions to the summarized trends or relationships. Some minor formatting issues should also be addressed, e.g. the confusing color codes in Figs. 9D-i, E-i.

Thank you for the advice, we have now pooled aISVs together into one group and vISVs into another, instead of discussing data trends on each of the 10 ISVs.

The mispattening case presented in the end of the results section (section "2.4.2") is interesting but appears loosely connected to the preceding contents. Also, it seems not even mentioned in the discussion section.

We agree that the mispatterning case has been only tangentially relevant to the rest of the manuscript. We have linked the topic thematically by network alterations transforming network flows. It is also now included in the discussion section here:

Page 15, lines 30 to 34

Finally, apart from the effect of topological features on local blood flow, the authors should consider the global flow redistribution arising from the network structure (useful refs. Include Chang et al. PLOS Computational Biology 2017: 10.1371/journal.pcbi.1005892; Meigel et al. Physical Review Letters 2019: 10.1103/PhysRevLett.123.228103; Schmid et al. eLife 2021: 10.7554/eLife.60208).

Thank you for the additional references. These are solid pieces of work that have been added to the discussion here:

Page 16, lines 3 to 10

**Referees cross-commenting**

This review report resonates with mine from an experimental perspective and I agree with all points made regarding issues of the current manuscript that the authors need to address with a revised version.

Reviewer #2 (Significance (Required)):

Significance: The particular merit of the work lies in its comprehensiveness of design and abundance of data, which will be of great interest to both the computational and experimental communities in this research field. However, some crucial details (especially with respect to the modelling aspects) are missing, thus hampering the scientific rigor and potential impact of the work. Furthermore, certain justifying statements appear speculative and inconclusive to explain the obtained data, especially regarding the effect of boundary conditions and systemic parameters. The citation of references (some not cited, some cited already but not properly discussed) also needs to be enhanced with engaging discussions to better bridge the findings of the current work (e.g. RBC partitioning in vascular network, effect of WSS on vasculature morphogenesis) with recent works on this research topic.

References

Fedosov DA, Caswell B, Karniadakis GE. 2010. A Multiscale Red Blood Cell Model with Accurate Mechanics, Rheology, and Dynamics. Biophys J 98:2215–2225. doi:10.1016/j.bpj.2010.02.002

Freund JB, Goetz JG, Hill KL, Vermot J. 2012. Fluid flows and forces in development: functions, features and biophysical principles. Dev Camb Engl 139:1229–45. doi:10.1242/dev.073593