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 manuscript by Gala et al, the authors perform meta-analysis of transcriptomic datasets from different studies focused on synaptically-associated mammalian glial cells. Based on a detailed analysis, the authors predict 1700 localized transcripts and attempted to identify the local transcriptome in glial cells conserved in Drosophila and mammals. This kind of data mining is informative, and can predict the top candidates relatively well, however, the study suffers from lack of depth and a careful assessment comparing across the different datasets. The main concerns have been listed below and suggestions for a more detailed an careful assessment of the data.<br /> 1. The authors have pooled data from different studies across different type of glial cells performed from in vitro to in vivo. While pooling datasets may reveal common transcripts enriched in processes, this may not be the best approach considering these are completely different types of glial cells with distinct function in neuronal physiology.<br /> 2. It is important to note the limitations of the study. For example, DeSeq2 is biased for highly expressed transcripts. How robust was the prediction for low abundance transcripts?<br /> 3. The authors identify 1,700 transcripts that they classify as "predicted to be present" in the projections of the Drosophila PNS glia. This was based on the comparison to all the mammalian glial transcripts. Since the authors have access to a transcriptomic study from Perisynaptic Schwann cells (PSCs), the nonmyelinating glia associated with the NMJ isolated from mice; it would be more convincing to then validate the extent of overlap between Drosophila peripheral glial with the mammalian PSCs. This may reveal conserved features of localized transcripts in the PNS, particularly associated with the NMJ function.<br /> 4. Fig 2: What is the extent of overlap between the translating fractions versus the localized fraction? It will be informative to perform the functional annotation of the translating glial transcripts as identified from Fig 1D.<br /> 5. "We conclude predicted group of 1,700 are highly likely to be peripherally localized in Drosophila cytoplasmic glial projections". To validate their predictions, the authors test some of these candidates in only one glial cell type. It might be worthy to extend this for other differentially expressed genes localized in another glial type as well.<br /> 6. Figure 5: The authors perform KD of candidate transcripts to test the effect on synapse formation. However, these are KD with RNAi that spans across the entire cell. To make the claim about the importance of "target" RNA localization in glia stronger, ideally, they should disrupt the enrichment specifically in the glial protusions and test the impact on bouton formation. Do these three RNAs have any putative localization elements?<br /> 7. RNA localization in oligodendrocytes has been well studied and characterized. The authors should cite and discuss those papers (PMID: 18442491; PMID: 9281585).

Significance

In the manuscript by Gala et al, the authors perform meta-analysis of transcriptomic datasets from different studies focused on synaptically-associated mammalian glial cells. Based on a detailed analysis, the authors predict 1700 localized transcripts and attempted to identify the local transcriptome in glial cells conserved in Drosophila and mammals. This kind of data mining is informative, and can predict the top candidates relatively well, however, the study suffers from lack of depth and a careful assessment comparing across the different datasets and the biological significance.

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:

The manuscript by Gala et al. describes efforts to systematically identify candidate localizing mRNAs within glial cells of the Drosophila larval nervous system. They address this issue with an interesting and novel approach that involves compiling candidate localized transcripts from available mammalian cell datasets and cross-referencing a list of mRNAs from homologous Drosophila genes with a small number of candidates for mRNA localization in Drosophila glia cells that this team identified in a previous study (which they have now validated further). The results provide compelling evidence that their pipeline has predictive power. The authors then go on to perform functional analysis of a small number of validated localizing transcripts and provide evidence that several of these genes are functionally important in glia, including in processes related to synaptic plasticity. These observations raise the possibility that the localization process is functionally important but this was not tested directly (doing so would require a separate long-term study).

The manuscript is well written and the figures are of high quality. Methodological details are provided, as is quantification where appropriate. However, some points are not yet sufficiently strongly supported by the data yet and the justification for some aspects of the workflow needs clarifying.

Major points:

1. The authors analyse the 1700 shortlisted genes for Gene Ontology and associations with austism spectrum disorder, leading to interesting results. However, it is not clear to what extent the enrichments they observe are driven by their presumptive localization or if the associations are driven to a significant extent by the presence of these genes in the selected cell types in the Fly Cell Atlas. One way to address this would be to perform the GO and SFARI analysis on genes that are expressed in the same cells in the Fly Cell Atlas but were not shortlisted from the mammalian cell datasets - the results could then be compared to those obtained with the 1700 localized transcripts.
2. Although the authors attempt to justify its inclusion, I'm not convinced why it was important to use the whole cell transcriptome of perisynaptic Schwann cells as part of the selection process for localizing transcripts. Including this dataset may reduce the power of the pipeline by including mRNAs that are not localized to protrusions. How many of the shortlisted 1700 genes, and how many of the 11 glial localized mRNAs in Table 5, would be lost if the whole cell transcriptome were excluded. More generally, what is the distribution of the 11 validated localizing transcripts in each dataset in Table 4? This information might be valuable for determining which dataset(s), if any, has the best predictive power in this context.
3. Did the authors check if any of the RNAi constructs are reducing levels of the target mRNA or protein? Doing so would strengthen the confidence in these important results significantly. In any case, the authors should also mention the caveat of potential off-target effects of RNAi.
4. Methods: what is the justification for assuming that if the RNAi cross caused embryonic or larval lethality then the 'next most suitable' RNAi line is reporting on a phenotype specific to the gene. If the authors want to claim the effect is associated with different degrees of knockdown they should show this experimentally. An alternative explanation is that the line used for phenotypic analysis in glia is associated with an off-target effect.

Minor points:

1. It would be helpful to have in the Introduction (rather than the Results, as is currently the case) an operational definition of mRNA localization in the context of the study. And is it known whether or not localization in protrusions is the norm in mammalian glia or the Drosophila larval glia? I ask because it may be that almost all mRNAs diffuse into the protrusion, so this is not a selective process. One interesting approach to test this idea might be to test if the 1700 shortlisted transcripts have a significant underrepresentation of 'housekeeping' functions.
2. Introduction: 'Asymmetric mRNA localization is likely to be as important in glia, as it is in neurons,...'. Remove commas

Significance

Whilst glial mRNA localization has been reported in other systems, this study is significant as it paves the way for functional and mechanistic dissection of this process in a genetically tractable model system that involves physiologically relevant neuron-glia interactions. The successful results of the data mining approach may also encourage others to address other problems in this manner. The study would be more impactful if additional hits from the 1700 transcripts were validated for glial localization but the findings in the current manuscript are still important. The advance is more methodological than conceptual. The work will appeal to cell biology and systems biology audiences.

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 study of mRNA localization in glia has been largely overlooked in comparison to neurons. However, glia are also polarized cells with long cytoplasmic processes that play important roles in neural function. In this manuscript, Gala and Lee et al attempt to fill in this missing void in the literature. They first use a meta-analysis of existing, cross-species, transcriptomic data to identify a set of 1,700 transcripts that are likely to be localized to the periphery of glia in Drosophila. They then used this list of transcripts to predict which mammalian glial transcripts are also likely to be localized. They're analysis suggests that a large proportion of the mammalian glial transcripts are predicted to be localized to the periphery of glia, and that these transcripts are enriched for functions involved in membrane trafficking, cytoskeleton regulation, local translation, and cell-cell communication. A connection with cross-cellular communication (ie glia-glia, glia-neuron, glia-muscle) at the neuromuscular junction, prompting the authors to assess if some of these transcripts play a role in plasticity of at the NMJ. Using siRNA driver lines, that loss of some of their assessed transcripts prevent new synapse formation- suggestive of a role for mRNA localization and local protein synthesis in driving synaptic plasticity. These findings suggest that mRNA localization is a widespread phenomenon in glia, and that it plays an equally important role as the more widely studied neuronal mRNA localization.

Key Points

• 1,700 transcripts are predicted to be localized to the periphery of PNS glia in Drosophila at the NMJ.
• A large proportion of the mammalian glial transcripts are predicted to be localized to the periphery of glia.
• Localized transcripts are enriched for functions involved in membrane trafficking, cytoskeleton regulation, local translation, and cell-cell communication.
• Localized glial transcripts may function in synaptic plasticity.

Major comments

Overall, the work presented within manuscript is well reasoned and supports the points the authors generally are trying to make. The work as is, does fall a little on the light side, resting on almost exclusive bioinformatic analysis with some limited validation. The presented work itself (Figures 1-4) does not need significant adjustments prior to publication, in my view, with only a few points to address. However, the work in Figure 5- doesn't really support the claims the authors make on its own, and would require some additional experiments or at the very least discussion of the caveats to its current form.

1. Localized glia transcripts, are they "glial/CNS/PNS" significant or are they similar to other known datasets of protrusion transcriptomes? The authors compared their 4801 "total" localized to a local transcriptome dataset from the Chekulaeva lab finding that a significant fraction are localized in both. As the authors note, this is in good agreement with a recent paper from the Talifarro lab showing conservation of localization of mRNAs across different cell types. What the authors haven't done here, is further test this by looking at other non-neuronal projection transcriptomic datasets (for example Mardakheh Developmental Cell 2015, among others). If the predicted glia-localized processes are similar to non-neuronal processes transcriptomes, this would further strengthen this claim and rule out some level of CNS/PNS derived linage driving the similarities between glia and neuronal localized transcripts.
2. The presentation/discussion around Figure 3 is a bit weaker than other parts of the manuscript, and it doesn't really contribute to the story in its current form. Notably there is no discussion about the significance of glia in neurological disorders until the very end of the manuscript (page 21), meaning when its first brought up.. it just sits there as a one off side point. The authors might consider strengthening/tightening up the discussion here, if they really want to keep it as a solo main figure rather than integrating it somewhere else/putting it into supplemental. In my view, Figures 2 & 3 should be merged into something a bit more streamlined.
3. Why aren't there more examples of different mRNAs in Figure 4? Seems a waste to kick them all to supplemental.
4. The plasticity experiments, while creative, I think need to be approached far more cautiously in their interpretation. Given that the siRNAs will completely deplete these mRNAs- it really needs to be stressed any/all of the effects seen could just be the result of "defective" or "altered" states in this glial population- which has spill over effects on plasticity in at the NMJ. Without directly visualizing if these mRNAs are locally translated in these processes and assessing if their translation is modulated by their plasticity paradigm, all these experiments can say is that these RNAs are needed in glia to modulate ghost bouton formation in axons. This represents the weakest part of this manuscript, and the part that I feel does not actually backup the claims currently being made. Without any experiments to A. quantify how much of these transcripts are localized vs in the cell body of these glia, B. visualize/quantify the translation of these mRNAs during baseline and during plasticity; the authors cannot use these data to claim that localized mRNAs are required for synaptic plasticity.

Minor points:

1. The use of blue/green or blue/green/magenta is difficult to resolve in some places. Swapping blue for cyan would greatly aid in visualizing their data.
2. Make the colouring/formatting of the tables more consistent, its distracting when its constantly changing (also there is no need for a blue background.. just use a basic white table).

Significance

This work would be well received within the field, being at a time where increasing focus on mRNA localization and local protein synthesis are major players in glia along with the more widely studied neurons. Additionally, that mRNAs localized in glia are relevant to neurological disorders also comes at a time where increasingly, especially in neurodegenerative disorders, glia are believed to be drivers of the diseases as well- suggesting that dysregulation of local protein synthesis in glia may play a role. The major weakness of this work is that being largely bioinformatic (from existing datasets), a lot of this is speculative and no conclusive data is shown to demonstrate how/that this glia population is utilizing mRNA localization in protrusions to fuel local protein synthesis for a specific purpose (ie synaptic plasticity).

This work would be of interested to those broadly interested in glial biology and mRNA localization.

The reviewers background is in RNA localization and local protein synthesis in neurons.

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 characterized 4 proteins from P. falciparum via cellular (co-)localization, endocytosis, parasite growth, and artemisinin resistance assays. These proteins have been identified as candidates for Kelch13 compartment and a possible role in endocytosis in their previously work with quantitative BioID for potential proximity to K13 and Eps15 (Birnbaum et al. 2020). In the current work, additional 6 proteins were not confirmed as being associated to the K13 compartment. This experimental work was complemented by an in-silico analysis of protein domains based on AlphaFold algorithm. For this protein structure evaluation all proteins were chosen, which were experimentally confirmed to be linked to the K13 compartment in the current publication and previous work. With the work 3 novel proteins linked to artemisinin resistance or endocytosis could be functionally described (KIC12, MCA2, and MyoF) and a number of hypotheses were generated.

Major comments:

The quality of the presented work is solid, the experimental design is adequate, and methods are presented clearly. The publication contains a lot of results both presented in text and in the figures and it is not always straight forward for the reader to follow the descriptions due to many details presented and a lack of context for some of these experiments.

Specific suggestions for consideration by the authors to improve the manuscript.

Abstract: - R 31: Mention how the 4 proteins were identified as candidates, you need to refer to previous work to clarify this - R38: "Second group of proteins" is confusing - different from the 4 mentioned above? Significance to endocytosis unclear. Please unify terminology in the manuscript, see also comment below on proxiome - Abstract can only be understood after reading the full publication

Results: Table 1 is missing from the submitted materials Consider to shorten and stratify the result section to focus on the significant data Unclear how the localization and functionalization assays might be impaired by the fusion proteins Significance of ART resistance assay is not clear, in presence of strong growth effects due to inactivation or truncation of genes/proteins

MCA Stratify results, order by significance of findings, it appears to be described in chronological order, improve readability/flow, eg ART resistance if mentioned in r138, but only reported in r183ff MyoF R195 to 197 - consider moving to discussion as it is distracting here Term proxiome is introduced above, but not used in result section - suggest to unify language, eg r195 uses "K13 compartment DiQ-BioIDs" instead, which is not very convenient for the reader What is the enrichment factor? Please provide for this and the following proteins, eg in Table 1 R225 to 243 - overall significance of the growth experiments with mislocaliser is not clear, consider removing from manuscript or explain relevance more clearly KIC11/KIC12 Enrichment factor?

Referees cross-commenting

I would like to applaud reviewer #1 for a great, very thorough review and lots of detailed suggestions. I agree with the conclusions mentioned in the significance evaluation from reviewer #1 and #2: the work presented does not contain novel methods and the scope is rather narrow with the current results. (I am working on clinical studies with novel antimalarial agents)

Significance

On the one hand side, the authors have wrapped up some of the remaining protein candidates of the K13 compartment and could verify 4 of 10 proteins. The work is of interest for the scientific community working on endocytosis and malaria drug resistance mechanisms. Overall, the conclusions and findings from the previous work, Birnbaum et al. 2020, could be confirmed and extended mainly using the methods previously described. On the other hand, the authors made use of progress in protein structure predictions and identified domains linking the K13 compartment proteins to putative functions. The overlaid protein folds of the newly identified domains in figure 5 look convincing, but I can't comment on the technical details or cut-off used for this in-silico analysis.

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 a previous publication the Spielmann lab identified the molecular mechanism of ART resistance in P. falciparum by connecting reduced levels of the protein K13 to decreased endocytosis (uptake of hemoglobin from the RBC cytosol), which results in reduced ART susceptibility. Using quantitative BioID the authors further identified proteins belonging to a K13 compartment, highlighting an unusual endocytosis mechanism.

In the present manuscript the authors follow up on this work and closely examine ten more proteins of the K13/Eps15-related "proxiome". They successfully link MCA2 to ART resistance in vitro, while the proteins MyoF and KIC12 are involved in endocytosis but do not confer in vitro ART resistance when impaired. They further characterize one candidate (KIC11) that partially colocalizes with K13 in trophozoites but to a lesser degree in schizonts. Growth assays suggest an important function for KIC11 in late stages of the intraerythrocytic developmental cycle. Five analyzed proteins however do not colocalize with the K13 compartment, while a sixth was refractory to endogenous tagging.

Using AlphaFold predictions of the KIC protein structures the author identify domains in most constituents of the K13 compartment, highlighting vesicle trafficking-related features that were not identified on primary sequence level before. The combination of functional data together with structure predictions leads them to propose a refinement of the K13 compartment as being divided into proteins participating in endocytosis and proteins that have an unknown function.

Major comments:

• Table 1 is missing
• Lines 117-123: Given the total list of uncharacterized candidates encompasses 13 proteins, can the author gives the reason why only the top 10 and not all 13 were characterized in this study?
• Line 174: 20% of observed MCA2 foci show no overlap with K13 and 21% only partially overlap, can the author confirm that the observed MCA2 foci in schizonts are the ones that co-localize with K13. (Addition of a schizont stage image in Fig 1C would be sufficient).
• The localization and observed phenotype of KIC11 is interesting but unfortunately the authors do not explore it further. Does KIC11 localize with markers of e.g. the secretory organelles (micronemes or rhoptries) in schizonts and could therefore be involved in RBC invasion? Can the author distinguish if KIC11 is involved in RBC invasion or in establishment of the ring-stage parasite?

Minor comments:

• Table S1: Please add the criterion for the order of proteins (abundance in "proxiome"?) in the table as a separate column. I would also suggest adding a new column that highlights the 10 proteins investigated in this study as I found the color-coding slightly confusing.
• 154-155: There is a discrepancy between the text and Fig1C regarding the % of partial overlapping and non overlapping foci.
• The y-axis label is missing in Fig 3E
• Fig 4I left graph, the superscript 2 is missing in μm2
• Did the author colocalize KIC11 in schizonts with other proteins found in the K13 compartment group of proteins not involved in endocytosis/ART resistance? This may help to further subgroup these proteins.
• As a general comment: to make the beautiful IFAs more accessible to a broader readership, I would encourage the authors to switch the color-coding to green/magenta/blue or an equivalent color system or add grayscale images.

Significance

Characterizing the molecular components involved in Plasmodium endocytosis will not only reveal interesting biology in these highly adapted parasites, but will more importantly lead to a better understanding and potentially open new avenues for intervention of ART resistance. The here presented manuscript is a carefully executed follow-up on previous work done in Dr. Spielmann's lab focusing on the K13 compartment. The authors use established assays to characterize novel components and reveal three new players in endocytosis with one mediating ART resistance in vitro. The proposition that parts of the K13 compartment have a function other than endocytosis is interesting, but will have to await more data from future studies. Taken together this manuscript adds significantly to our understanding of endocytosis in P. falciparum.

This work is of interest for cell and molecular biologists working on Apicomplexa, but especially for the Plasmodium community.

I am a cell and molecular biologist working on Toxoplasma gondii

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

With the emergence and spread of resistance to Artemisinin (ART), a key component of current frontline malaria combination therapies, there is a growing effort to understand the mechanisms that lead to ART resistance. Previous work has shown that ART resistant parasites harbour mutations in the Kelch13 protein, which in turn leads to reduced endocytosis of host haemoglobin. The digestion of haemoglobin is thought to be critical for the activation of the artemisinin endoperoxide bridge, leading to the production of free radicals and parasite death. However, the mechanisms by which the parasites endocytose host cell haemoglobin remain poorly understood.

Previous work by the authors identified several proteins in the proximity of K13 using proximity-based labelling (BioID) (Birnbaum et al. 2020). The authors then went on to characterise several of these proteins, showing that when proteins including EPS15, AP2mu, UBP1 and KIC7 are disrupted, this leads to ART resistance and defects in endocytosis leading to the hypothesis that these two processes are inextricably linked.

In this manuscript, Schmidt et al. set themselves the task of characterising more K13 component candidates identified in their previous work (Birnbaum et al. 2020) that were not previously validated or characterised. They chose 10 candidates and investigated their localisations, and colocalisation with K13, and their involvement in endocytosis and in vitro ART resistance, 2 processes mediated by K13 and some members of the K13 compartments

The authors show that of their 10 candidates, only 4 can be co-localised with K13. Then, using a combination of targeted gene disruption (TGD) as well as knock sideways (KS), they characterised these 4 proteins found in the K13 compartment. They show that MyoF and KIC12 are involved in endocytosis and are important for parasite growth, however their disruption does not lead to a change in ART sensitivity. The authors also confirm the findings of their previous publication (Birnbaum et al. 2020), using a slightly different TGD, that MCA2 is involved in ART resistance, however they did not check whether its disruption impacts haemoglobin uptake. They also show that KIC11 is not involved in mediating haemoglobin uptake or ART resistance. To finish, the authors used AlphaFold to identify new domains in the proteins of the K13 compartment. This led them to the conclusion that vesicle trafficking domains are enriched in proteins of the K13 compartment involved in endocytosis and in vitro ART resistance.

The majority of the experiments conducted by the authors are performed to a good standard in biological and technical replicates, with the correct controls. Their findings provide confirmation that their 4 candidate genes seem to be important for parasite growth, and show that some of their candidates are involved in endocytosis. While the KD and KS approaches employed by the authors to study their candidate genes each have their own advantages and can be excellent tools for studying a large sets or genes, this manuscript highlights the many limitations of these approaches. For example, the large tag used for the KS approach can mislocalise proteins or disrupt their function (as is the case for MyoF), resulting in spurious results, or indeed the inability to generate the tagged line (as is the case for MCA2). The KS approach also makes the results of a protein with a dual localisation, like KIC12, extremely difficult to interpret.

Moreover, the manuscript is disjointed at times, with the authors choosing to conduct certain experiments for only a subset of genes, but not for others. For example, considering that the aim of this paper was to identify more proteins involved in ART resistance and endocytosis, it is confusing why the authors do not perform the endocytosis assays for all their selected proteins, and why they do not do this for the proteins they identify in their domain search. There is significant room for improvement for this manuscript, and a generally interesting question. But in it's current format, other than confirming that MCA2 is involved in ART resistance (which was already known from the Birnbaum paper), the authors do not further expand our understanding of the link between ART resistance and endocytosis in this manuscript.

Major Comments

line 31: please change defined to characterised - defined suggests that novel proteins were identified in this study, which is not the case.

line 37: please change 'second' to "another". As explained further below, the authors identified 3 classes of proteins (confer ART resistance + involved in HCCU, involved in HCCU only, or involved in neither).

Line 40: You define KIC11 as essential but according to your data some parasites are still alive and replicating 2 cycles after induction of the knock sideways. Please consider changing "essential" to "important for asexual parasite growth"

Line 40: please change 'second group' to 'this group'

line 41: state here that despite it being essential, it is unknown what it is involved in.

Line 50: the authors should state here that there is actually a reversal in this trend over the last few years.

Line 54: please separate out the references for each of the two statements made in this line (a: that ART resistance is widespread in SEA, and b: that ART resistance is now in Africa) Reference 14 also seems to reference ART resistance in Amazonia - which is not covered by the statement made by the authors (in which case the authors should state ART is now present in Africa and South America). The authors should also reference PMID: 34279219 for their statement that ART resistance is now found in Africa (albeit a different mutation to the one found in SEA).

Line 65: it is also worth mentioning here that there are other mutations in proteins other than K13, such as AP2mu and UBP1 (PMID: 24994911;24270944) that can lead to ART resistance.

Line 80, 86: ref 43 is misused. Reference 43 refers to Maurer's clefts trafficking which takes place in the erythrocyte cytosol and is not involved in haemoglobin uptake as far as I know. Please replace ref 43 with one showing the role of actin in haemoglobin uptake.

Line 98: the authors state here that they 'identified' further candidates from the K13 proxiome. This suggests that they identified new proteins in this paper, when in fact the list was already generated in ref 26. All they did was characterise proteins from that list that were not previously characterised. The authors should therefore remove identified from this statement.

Line 107-108: it is not clear from this sentence why these proteins were left out of the initial analysis in Ref 26. A sentence here explaining this would be valuable for the reader.

Line 117-123: The authors say that PF3D7_0204300, PF3D7_1117900 and PF3D7_1016200 were not studied because they were not in the top 10 hits. However, the current organisation of Supplementary Table 1 shows all 3 proteins among the top 10 hits (MyoF, KIC12, UIS14 and 0907200 being after them). I think the authors should reorganise their table. It is also unclear according to what the proteins in the table are ranked. Could the authors indicate the metric used for the ranking?

Line 129-141: Can the authors be clearer with their explanations of the identification of mutation Y1344Stop? One dataset (ref 61) shows that 52% of African parasites have a mutation in MCA2 in position 1344 leading to a STOP codon. But another dataset (ref 62) shows that the next base is also mutated, reverting the stop codon. That should have been seen in the first dataset as well. Could the authors please clarify.

Line 147: the authors say that MCA2 is expressed throughout the intraerythrocytic cycle as shown by live cell imaging. In Birnbaum et al 2020 fig 4I, the authors show that MCA2 is mainly expressed between 4 and 16hpi. But in Figure 1B of this manuscript there is a clear multiplication of MCA2 signal between trophozoite and schizont. How do the authors explain this discrepancy? Could expression of the truncated MCA2 be different than the full length? This cannot be assessed as expression and localisation of the full-length HA tag MCA2 is not shown in Schizonts. MCA2 expression seems also different for the MCA2TGD-GFP with no expression in rings.

Line 158: would it not have been more useful for the authors to have episomally expressed MCA2-3xHA in their MCA2Y1344STOP-GFPENDO line to make sure that the truncated protein is indeed going to the correct compartment? The experiments done by the authors suggests that the MCA2Y1344STOP goes to the right location but does not really confirm it.

Line 191: it is stated that MCA2 confers resistance independently of the MCA domain, however in both the MCA2-TGD and MCA2Y1344STOP-GFPENDO parasites, the MCA domain is deleted, and for both parasites, there is resistance (albeit to a lower level in the MCA2Y1344STOP-GFPENDO line). Therefore, how can the authors state that the ART resistance is independent of the MCA domain? This statement should be that resistance is dependent on the loss of the MCA domain.

Line 192: Why did the authors not check if MCA2 is involved in endocytosis? They state later on in the manuscript that they did not do endocytosis assays with TGD lines, however if the authors include the correct controls, this could be easily done. It would also be really interesting to see whether endocytosis gets progressively worse going from WT to MCA2Y1344STOP to MAC2TGD. This experiment (as well as doing endocytosis assays for KIC4 and KIC5 TGD lines) would drastically increase the impact of this study. These experiments would not take more than 3 weeks to perform, and would not require the generation of new lines.

The authors should consider re-organising the MCA2 section, first showing that the 3xHA tagged line colocalises with K13, then performing the new truncation.

Line 197: Once again ref 43 is not correct to illustrate that actin/myosin is involved in endocytosis

Line 202: the authors state that MyoF localises near the food vacuole from ring stage/trophs onwards. However, how can this statement be made in schizonts based on these images (Fig. 2A), where it doesn't look like MyoF is anywhere near the FV? This statement can only be made for schizonts if co-localised with a FV marker (which is done in Fig. 2B), however, based on the number of MyoF foci, it appears that this was not done for schizonts. Please either remove the statement that MyoF is near the food vacuole from trophs onwards (because it is only seen near the FV up until trophs) or show the data in Fig. 2B of schizonts to substantiate these claims.

Line 204-206: what does this statement bring to the paper? Is it to show that it is the real localisation of MyoF because 2 tag cell line show the same localisation? I don't think this is needed, especially as later in the manuscript an HA-tag MyoF line is used and show similar localisation.

Line 212: The overlap of K13 with MyoF in Fig 2C 3rd panel (1st trophozoite panel) is not obvious, especially as the MyoF signal seems inexistant. I would advise the authors to replace with a better image. Also, why are there no images of schizonts shown in Figure 2C?

Line 217: the spatial association of MyoF with K13 is very different when it is tagged with GFP and when it is tagged with 3xHA. The way the authors word it here, it seems that there is agreement with the two datasets, when this is not in fact the case (59% overlap for MyoF-GFP and only 16% overlap with MyoF-3xHA). These data suggest that the GFP and the multiple FKBP tags are doing something to the protein and therefore maybe the ensuing results using this line should not be trusted or be taken with a pinch of salt.

Line 219: the authors state here that they could not detect MyoF-GFP in rings, when in Figure 2C they show MyoF-GFP in rings, and also show that they could detect MyoF in Sup Fig. 3B with the 3xHA tagged line. Is this a labelling mistake in Figure 2C? If the authors could indeed not see MoyF-GFP in rings, this statement should have been made when Figure 2A was presented, and not so late in the manuscript, which causes confusion. Line 237: Showing a DNA marker (DAPI, Hoescht) for Figure 2E, and subsequent figures using mislocalisation to the nucleus, would help the reader assess efficiency of the mislocalisation.

Line 254-256: authors should show the results of the bloating assay for parental 3D7 parasites (+ and - rapalog) to see whether the MyoF line - rapalog has increased baseline bloating. This applies to all subsequent FV bloating assays.

Line 254-257: The authors say that because fewer parasites show a bloated food vacuole upon inactivation of MyoF it means that less hemoglobin reached the food vacuole. I understand the authors statement, however, shouldn't they look at the size of the food vacuole, instead of the number of parasites with bloated FV, to make such a statement? This has been done for KIC12 so why not doing it for MyoF?

Line 259-261: these results would be difficult to interpret namely because the authors have dying parasites, which is exacerbated with the protein being knocked sideways. The authors should mention the pitfalls their knock sideways and tagging design here.<br /> Line 260-261: RSA is an assay relying on measuring parasite growth 1 cycle after a challenge with ART for 6 hours.

Line 261-263: the authors sate that MyoF has a function in endocytosis but at a different step compared to K13 compartment proteins. I am not sure what they mean here. Can this be clarified? Do the authors mean that it is involved in endocytosis but not in ART resistance? If so, this is a very difficult statement to make since the parasites are dying. Is there any evidence of point mutations in MyoF in the field?

Line 298: the authors state that there is no growth defect in the first cycle when rapalog is added to the KIC11 line, however based on Figure 3D, there is evidently a 25% reduction in growth compared to - rapalog at day 1 post treatment, and a 60% reduction by day 2, which is still within the 1st growth cycle. The authors should either revise their statement or provide an explanation for these findings. The authors should also explain why their Giemsa data in Fig. 3E is not in accordance with their FACS data.

Line 301: KIC11 could also be important very early for establishment of the ring stage for example for establishment of the PV. Also, was mislocalisation assessed in rapalog-treated parasites at 72 hours or in cycle 3?

Line 311: the authors should change the sentence from 'not related to endocytosis' to 'not related to endocytosis or ART resistance'.

Line 323-325: Authors say that a nuclear GFP signal can be observed in early schizonts for KIC12. According to the pictures provided in Figure 4A and Figure S5A it is not very obvious. Also faint cytoplasmic GFP signal could only be background as we can see that exposure is higher for schizont pictures

Line 326-328: The authors say that kic12 transcriptional profile indicate mRNA levels peak (no s at peak) in merozoites. Should they show live cell imaging of merozoites then? Because from the Figure 4A schizont pictures where schizonts are almost fully segmented no signal can be observed. Line 347: The authors state that using the Lyn mislocaliser the nuclear pool of KIC12 is inactivated by mislocalisation to the PPM. This tends to suggest that only the nuclear pool of KIC12 is mislocalised. How is it possible that only the nuclear pool is mislocalised? Line 368-369: Effect was also only partial for MyoF. Why didn't you measure the same metrics for MyoF? Line 379: you don't know if all proteins acting later in endocytosis will have an increased number of vesicles as a phenotype

Line 413-414: The authors state that no growth defect was observed upon KS of 1365800. Is growth alone enough to say that there is no impact on endocytosis?

Line 432: in this section, the authors state that KIC4 and KIC5 seem to have domains that may suggest these proteins are involved in endocytosis, based on the alpha fold data that is publicly available. Considering the authors have TGD-SLI versions of these lines (Birnbaum et al. 2020) and have already confirmed in this previous publication that they confer resistance to ART; it would make sense to look at endocytosis for these genes. This would be a relatively simple and straightforward experiment, taking no longer than two to three weeks, and would require no additional reagents or line generation. Doing these experiments would add a lot more weight to this final section. The authors later state that KIC4 and 5 are TGD lines, so not the best for endocytosis assays. It is unclear why this would be difficult to do if an adequate control is contained in the experiment (such as parental 3D7). It explains why they did not perform the MCA2 endocytosis assays further up, but in my opinion, an attempt at doing these assays is important and would significantly increase the impact of this paper.

Line 490-493: the authors state that the K13 compartment proteins fall in two groups, some that are involved in ART resistance AND endocytosis, and some that have different functions. However, in this manuscript the authors have demonstrated 3 flavours that K13 compartment proteins can come in: • Some that confer ART resistance and are involved in HCCU (MCA2) • Some that are involved in HCCU but not ART resistance (MyoF & KIC12) • Some that are involved in neither (KIC11) The authors should therefore revise this statement.

Line 508: the authors state that they expanded the repertoire of K13 compartments, when in fact they functionally analysed them - they did not do another BioID to identify more candidates.

Line 570-572: has anyone ever tested whether CytoD or JAS treatment in rings, is sufficient to mediate ART resistance? Something similar to what was done in PMID 21709259 with protease inhibitors. If not this would be a pretty interesting experiment for the authors to do that could shed more light on the MyoF data. It would take maybe 2 weeks to do and not require the generation of any new lines. This would clarify whether other Myosins other than MyoF are involved in endocytosis, as is suggested by previous publications (PMID: 17944961).

Line 608: inhibitors targeting the metacaspase domain of MCA2 may inadvertently inactivate other essential parts of the protein. They authors should acknowledge this possibility in the text.

Line 624-625: the authors state that MyoF is 'lowly expressed in rings' - indeed this is the case in their MyoF-2xFKBP-GFP-2xFKBP line which the authors established has defects due to the tag, but it appears from their MyoF-3xHA tagged line that it is expressed in rings. The authors should therefore revise their statement, and be careful of making claims based on their defective line and using fluorescence imaging as their only metric. If they do want to make the statement that it is not there in rings, they should also do a western blot, which is much more sensitive since it amplifies the signal compared to an image of one parasite.

Line 635: arguably this is the 3rd variety and not the 2nd (the authors already mentioned 2 types - ones that are involved in HCCU AND ART and those involved in HCCU only). See comment for line 490-493 above.

Line 785: Bloated food vacuole assay/E64 hemoglobin uptake assay method specify that a concentration of 33mM E64protease inhibitor was used. However, in reference 44, cited in the manuscript, a concentration of 33µM E64 was used. Please confirmed if this is just a typo or if 1000x E64 concentration was used which renders the experiment invalid.

Line 788: it is unclear from this section what is considered a bloated food vacuole - is there an area above which the FV is considered bloated? Do the authors do these measurements manually or use an addon in FIJI/ImageJ? What is the cutoff for if a FV is bloated? Please clarify. Additionally, for the representative images + rapalog for Figures 2H and 4H, it would be useful to see where the authors delineate the FV (add a white circle showing what is actually measured).

Line 863-864: this sentence seems to be out of place.

Line 875: the authors state that there is a light blue wedge, when the circle consists of grey and black wedges. Please revise this.

Line 1059-1061: it is unclear whether the individual growth curves are different clones or whether they are just the same experiment repeated? If it is the latter, then why are they not combined, as is traditionally done?

Line 919-924: the authors mention a blue and red line, but there is only a black line in figure 3D. Moreover, the experiment of using the LYN mislocaliser was only done for KIC12 according to the manuscript. Additionally, the y axis of the figure states relative growth day 4[%] compared to rapalog, but then on the x axis there are several days. In the text it says there is no growth defect until the second cycle, but from this graph it appears the growth defect is evident as early as 1 day post rapalog treatment. Can the authors please clarify and correct the issues pointed out.

Figure 1 panel B & C: the label of the figure where the signal from MCA2Y1344STOP-GFP is shown with the DAPI signal overlayed is deceptive since it suggests that this is the signal of full length MCA2. Please change the label of this panel from MAC2/DAPI to MCA2Y1344STOP/DAPI. The same is true for Panel C for the image labeled MCA2/K13 - please change this to MCA2Y1344STOP/K13.

Figure 2B: what stages are these parasites? Please state this in the figure. Based on the MyoF pattern, it looks like rings in the upper panel and trophs in the bottom pannel. Why were schizonts not shown?

Figure 2D&F: it is not very meaningful when growth assays are shown as a final bar after 4 days of growth. It is much more useful and informative to see a growth curve instead (as is shown in the supplementary), since it shows if the defect is apparent in the first growth cycle or later. With the way the data is currently shown, this is not apparent. I would advise the authors to switch the graph in 2F out of a combined graph of all the biological replicates growth curves for S3D - showing error bars.

Figure 3: why were the calculation of FV area, parasite area and FV/parasite area only done for KIC12 and not done for MyoF? It would be interesting to see if any of these values are different for MyoF - whether the parasites are smaller in area and therefore FV smaller. Please present them Figure 2. Images should be already available and would not require further experiments to be done, only the analysis.

Figure 3B: why is there no spatial association assessment for KIC11 and K13 as was done for the MCA2 and MyoF? The authors should show a pie chart showing the degree of association here as was done for the other proteins.

Figure 3D: The y axis of the figure states relative growth day 4[%] compared to rapalog, but then on the x axis the experiment takes place over several days. Is this a typo in the y axis? Additionally, the authors state in line 287-290 that the growth defect upon addition of rapalog is only seen in the second cycle, but from this graph it appears the growth defect is already evident 1 day post rapalog addition. The figure legend also does not make sense for this figure since it mentions a blue and a red line, when there is only a black line present. The legend also mentions the LYN mislocaliser which was used for KIC12 not KIC 11 (see above).

Figure 3E: the colour for Control and Rapalog 4 hpi are very similar and very hard to discern. Please choose an alternative colour or add a pattern to one of the samples. The y axis is also missing a label. Is this supposed to be parasitemia (%)?

Figure 4A: the ring shown in this figure does not appear to be a ring (it is far too large and appears to have multiple nuclei?). Do the authors have any other representative images to show instead?

Figure 4B: why is there no spatial association assessment for KIC12 and K13 as was done for the MCA2 and MyoF? The authors should show a pie chart showing the degree of association here as was done for the other proteins. This should be done for the different life cycle stages considering the changing localisation of KIC12.

Figures 4C&E: it is extremely important to show the DNA stain in both these samples considering that a portion of KIC12 is in the nucleus! Please add the DAPI signal for these figures (as for all other figures!).

Figure 4E: this figure should be presented before 4D (considering the line being presented in 4E is used in an experiment in 4D). The authors should switch the order of these two.

It is unclear why in many of the fluorescence images the authors do not show the DAPI signal - particularly when colocalising with K13 and when doing the knock sideways experiments. Please add these images to the figures - I would assume they have already been taken, so would simply involved adding the images to the panel.

Throughout the manuscript, there is no western blot confirming the correct size of their modified proteins. This should be provided.

None of the figures are appropriate for individuals with colour blindness, limiting their accessibility to the paper. Please change the colour schemes for all fluorescent images using magenta/green or an alternative colour combination appropriate for colourblind individuals.

Minor Comments

line 29: remove 'are'.

Line 29: the text says "HCCU is critical for parasite survival but is poorly understood, with the K13 compartment proteins are among the few proteins so far functionally linked to this process." The sentence should be: 'HCCU is critical for parasite survival but is poorly understood, with the K13 compartment proteins among the few proteins so far functionally linked to this process."

line 44: remove 'the'

Line 48: consider mentioning here that malaria is caused by the parasite Plasmodium - otherwise the first mention of parasite in line 52 is confusing for the non-specialist reader.

Line 49: estimated malaria-related death and case numbers are from the 2021 WHO World malaria report. You cite the 2020 WHO World malaria report.

Line 53: please insert the word 'have' between now and also.

Line 54: please change 'was linked' to is linked

Line 72: I would specify that free heme is toxic to the parasite. Especially as you mention that hemozoin is nontoxic. Sentence would be "where digestion results in the generation of free heme, toxic to the parasite, which is further converted into nontoxic hemozoin"

Line 90: authors should either say "in previous works" or "in a previous work"

Line 91: "We designated these proteins as K13 interaction candidates (KICs)"

Line 95: please change 'rate' to number

Line 109: Please include a coma before (ii).

Line 112: as shown by Rudlaff et al in the paper you are citing, PPP8 is actually associated with the basal complex. You can say that "(ii) were either linked or had been shown to localise to the inner membrane complex (IMC) or the basal complex (PF3D7...).

Line 114: Protein PF3D7_1141300 is called APR1 in the manuscript but ARP1 in Supplementary Table 1. Please correct.

Line 131: please define SNP - this is the first use of the acronym.

Line 133-134: South-East Asia instead of "South Asia"

Line 135: please explain what TGD is - it is referred to over and over again in the manuscript without ever being explained.

Line 145: change 'Western blot' to western blot - only Southern blot is capitalised since it is named after an individual, while the other techniques are not.

Line 152: add "the" between 'and spatial'

Line 158: please define SLI as selected linked integration, since it is the first use of the acronym.

Line 178: introduce a coma after protein. Sentence should be "Proliferation assays with the MCAY1344STOP-GFPendo parasites which express a larger portion of this protein, yet still lacking the MCA domain (Figure 1), indicated no growth ...

Line 195: the authors could mention that MyoF was previously called MyoC in the Birnbaum 2020 paper. I wanted to check back in the Birnbaum 2020 paper and could not find MyoF

Line 200: "Expression and localisation of the fusion protein was analysed by fluorescent microscopy". Why expression was not analysed also by western Blot same as for MCA2?

Line 204: I could not find any mention of MyoF (Pf3D7_1329100) in reference 65. Please remove reference 65 if not correct. Also reference 66 looks at Plasmodium chabaudii transcriptomes so I would specify that "This expression pattern is in agreement with the transcriptional profile of its Plasmodium chabaudii orthologue"

Line 208: Please indicate a reference for P40 being a marker of the food vacuole

Line 220-224: The authors should consider changing to " Taken together these results show that MyoF is in foci that are mainly close to K13 and, at times, overlapping, indicating that MyoF is found in a regular close spatial association with the K13 compartment."

Line 255: In Figure 2H, and subsequent figures showing bloated FV assay, I would delineate the food vacuole with dashed line as in Birnbaum et al. 2020 to help the reader understanding where the food vacuole is.

Line 265-266: Here the title says that KIC11 is a K13 compartment associated protein, but the title of Figure 3 says KIC11 is a K13 compartment protein. I noticed that you make the difference between K13 compartment protein et K13 compartment associated protein for MyoF for example which is not clearly associated with the K13 compartment. Which one is it for KIC11?

Line 309-310: indicate a reference for your statement "which is in contrast to previously characterised essential K13 compartment proteins".

Line 377: Figure 4I, please correct 1st panel Y axis legend

Line 404: replace "dispensability" with dispensable

Line 416: can the authors provide any speculation as to why they observed these proteins as hits in the BioID experiments?

Line 451: Where the "97% of proteins containing these domains also contain an Adaptin_N domain and function in vesicle adaptor complexes as subunit " come from. Do you have a reference?

Line 465-467: the same could be said for KIC4 as it also has a VHS domain.

Line 477-479: Can be rephrased to "However, we found this protein as being likely dispensable for intra-erythrocytic parasite development and no colocalisation with K13 could be demonstrated, suggesting a limited role for PF3D7_1365800 in endocytosis. Or something like that. Makes it clearer.

Line 535: Have AP-2 or AP-2 been shown to be at the K13 compartment?

Line 569: reference 43 is wrong

Line 746: typo "ot" instead of or.

Line 801: method for Domain Identification using AlphaFold specify that RMSDs of under 5Å over more than 60 amino acids are listed in the results. However, there is a typo in Figure 5B for KIC5 where it says "RMSD 4.0 Å over 8 aa". Please correct.

Line 856: In Figure 1E, please use the same Y axis legend as in Figure 2D "relative growth at day 4 [%] compared with 3D7"

Figure S1: Some PCR gels check for integration are presented as 5', 3' and ori whereas other gels are presented as ori, 5' and 3'. This is confusing. Figure S1: Why was the expression of only MCA2 was verified by Western blot? What about the other proteins?

Line 493: Considering KIC11 was not involved in HCCU or ART resistance it might be worth mentioning in this section that it is of note that there are no domains detected that would be involved in endocytosis.

Line 503-506: is it wise to generate more drugs that target a pathway that is already highly susceptible to mutations? The authors should add a statement explaining how this might be avoided.

Throughout, scale bars are stated in the figure legends at the end of the legend. This is a slightly confusing format. The authors should consider stating the scale bar for each sub-legend where a fluorescence image is taken.

Referees cross-commenting

After reading reviewer 2 and 3's comments, I think there are significant overlaps in the key points raised in terms of questions about fusion proteins and their potential partial mis-localisation, better descripton of results and target selection. Overall I think we agree that the work has potential, but in its current form does not represent a major advance. It would be immensely helpful if the manuscript would be carefully edited for a better flow and linear description of results.

Significance

The authors set out to test whether other proteins that are in the vicinity of K13 are involved in mediating ART resistance and endocytosis. This is an interesting question. However, other than MCA2 which was already known to be involved in mediating ART resistance (and was not tested for its involvement in endocytosis), none of their candidate proteins seem to be involved in mediating both these functions. The authors show that the other proteins tested appear important for parasite growth, with KIC12 and MyoF involved in mediating endocytosis. While these findings are novel, the KS approach used by the authors casts some doubt over the findings, and would mean that these findings would have to be re-tested with a more reliable approach, such as the GlmS system or generating a conditional knockout using the DiCre system. Despite not advancing our understanding of ART resistance, or identifying further players involved in this process, this manuscripts provides two candidates that are involved in mediating endocytosis and a further candidate that appears to be important for parasite growth. Further work on these proteins will be required to understand their exact roles. As stated above, there is currently limited interest for these results (limited to researchers working on endocytosis in apicomplexan parasites and possibly the wider endocytosis field from an evolutionary perspective), however with further work, this could increase the impact and interest of this work substantially.

The authors do not describe any novel methods/approaches within this work.

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

In this manuscript Pho et al., characterize the effects of actin confinement versus acto-myosin contractility on nuclear deformation and nuclear envelope (NE) rupture/blebs. They use both genetic and pharmacological perturbations that impact nuclear shape and integrity such as chromatin decompaction and genetic deletion of nucleo-skeleton components (LaminA and LaminB). First, the authors show that modulation of acto-myosin contractility does not affect nuclear height (a proxy for the effects of actin confinement on the nucleus). This finding allowed them to study the effects of acto-myosin contractility on nuclear shape and integrity independently of actin-mediated nuclear confinement. Using fibroblasts expressing a NLS-GFP construct to assess nuclear shape and integrity, the authors show that increased acto-myosin contractility by treatment with Rho activator (CN03) leads to nuclear blebs and NE rupture events (curiously, they show that in LaminA-/- fibroblasts the NE rupture events were not bleb-based and were not affected by CN03). With these results, authors conclude that actin contraction is a major determinant of bleb-based NE rupture, independently of actin confinement. Next, since many studies have associated loss of nucleo-cytoplasmic compartmentalization with increased DNA damage, the authors addressed the occurrence of DNA damage foci in several contexts/perturbations. They find that increased NE rupture frequency by CN03 treatment does not lead to increased DNA damage. Based on the sole finding, authors then claim that DNA damage is mostly associated with abnormal shaped nuclei instead of ruptured nuclei. The data presented is overall convincing and well controlled but a more in-depth characterization of the DNA damage events in VPA-treated cells, LaminB-/- and LaminA-/- cells is needed to support the claim that DNA damage is not associated with NE events in their system. More details below.

We thank the reviewer for careful consideration of our work. We are happy that the reviewer found, “__The data presented is overall convincing and well controlled…”. __While the author appears positive on the paper as a whole, what follows is the Reviewer’s largest concern with the paper.

“ a more in-depth characterization of the DNA damage events…is needed to support the claim that DNA damage is not associated with NE events in their system.” We appreciate the Reviewer’s feedback and agree that this one statement was an overstatement. To address this concern,

1) We removed this disagreeable/unsupported conclusion,

2) We replaced it with more measured statement supported by the data (red text highlights changes to the original manuscript),

3) because we have removed the unsupported conclusion, we prefer not to carry out the suggested experiments because they constitute a completely new study beyond the scope of this manuscript.

Major points 1,2,3 all focus on the removed unsupported conclusion. They outline numerous new experiments which we feel constitute a new study and separate manuscript and thus beyond the scope of the current manuscript. Instead, we tone down or conclusions clearly labeled by red text.

__

__

__ __

__Major comments:____

1-In fig. 4, increased NE rupture frequency by CN03 treatment did not lead to an increase in gH2AX foci, so authors claim that DNA damage is not associated with NE events. However, the increase observed in NE rupture frequency is rather very subtle (from 1.5 to 3 in WT cells; from 2.5 to 4 in VPA treated ells and from 2 to 2.5 in LaminB-/- cells) so I wonder if such a minor increase might be sufficient to cause an increase in DNA damage foci. Moreover, DNA damage events are repaired within minutes and the analysis throughout this manuscript is made on snapshots. Therefore it becomes difficult to draw any conclusion regarding the absence of a link between NE rupture and DNA damage. To assess DNA damage events in a quantitative manner and show that abnormal shape (as opposed to NE rupture) is the main driver of DNA damage, authors should perform live cell imaging of cells co-expressing a DNA damage marker and a NE rupture marker to track DNA damage events following NE rupture.__

The reviewer raises concerns that we have too strongly stated a conclusion of nuclear rupture affects DNA damage less than shape. Overall, we understand this concern and have revised the text to remove this conclusion and in general tone down our conclusion from this data. We agree with the reviewers point that doubling the frequency either might not be sufficient to increase DNA damage OR a single rupture is sufficient to increase DNA damage. We now state: “ DNA damage measured by γH2AX foci did not significantly increase upon CN03 activation of actin contraction, suggesting that the slight yet significant increase in nuclear blebbing, rupture, and rupture frequency from CN03 did not have a significant impact on DNA damage (Figure 4, C and B).

We prefer not to. The reviewer’s request for live cell nuclear rupture and DNA damage data is now beyond the scope of the paper because we have revised the manuscript by removing this conclusion. A full analysis of live cell DNA damage analysis in conjunction with nuclear rupture for bleb-based, non-bleb-based normal circularity, and non-bleb-based normal circularity across 4 conditions (WT, VPA, LMNB1-/-, and LMNA-/-) is a paper in and of itself.

2-Fig. 4C: to show that frequency of NE rupture is less important than abnormal nuclear shape for DNA damage appearance, authors need to quantify gH2AX in normal x blebbed x abnormal nuclei in each one of the conditions (untreated, CN03 and Y27). But again, this analysis must be complemented by live cell imaging experiments to track DNA damage foci appearance (or not?) following NE rupture events.

We prefer not to Similar to point (1) we have removed this disagreeable conclusion and revised the manuscript. Further extensive studies of DNA damage for nuclear rupture and shape, while interesting, are beyond the scope of the current manuscript.

We now clearly state this in the Discussion section: “Our data in wild type, VPA, and LMNB1-/- cannot decouple the roles of nuclear shape and ruptures, which are intertwined, causing increased DNA damage (Figure 4). However, we provide novel data that actin contraction is necessary for the behaviors of nuclear blebbing, rupture, and increased DNA damage, independent of changes in actin confinement.

__ 3- Since VPA treatment increases both the percentage of bleb-based NE rupture and NE rupture frequency in LaminA-/- cells, authors should show that gH2AX foci in this sample remain unaltered to further support their claim that frequency of NE rupture is less important than abnormal nuclear shape.__

We prefer not to Similar to points (1) and (2) we have revised the manuscript to remove this disagreeable conclusion. Instead, Figure 6 is focused on the interesting phenomena where LMINA-/- nuclei display abnormal nuclear shape and majority non-bleb-based nuclear ruptures. To determine if LMNA-/- nuclei have the capacity to increase nuclear blebbing and bleb-based ruptures, we treated with VPA, which causes both. LMNA-/- with VPA shows that these calls lacking both lamin A and C can form blebs and increase in bleb-based ruptures.

__ 4-Does methylstat treatment rescue LaminB-/- phenotypes (blebs, ruptures, ...)?__

We do not see how this relates to a specific change to the manuscript but instead is a direct question of interest.

The reviewer would like us to clarify how our past work rescued Lamin B1 null (LMNB1-/-) phenotype which we cited in the original manuscript as a supporting point (top of page 16).

“This new data agrees with our previous data showing that increased heterochromatin levels via histone demethylase inhibition by methylstat treatment (Stephens et al., 2018) and mechanotransduction (Stephens et al., 2019a) rescues nuclear shape in LMNB1-/- nuclei.”

In our previous publication we showed that methylstat, which increases heterochromatin levels, can suppress nuclear blebbing in LMNB1-/- (Stephens 2018 MBoC). In another manuscript we increased heterochromatin levels via a mechantransduction pathway. This resulted in decreased nuclear blebbing, ruptures, and DNA damage (Stephens 2019 MBoC). However, this manuscript goes on to show that loss of facultative heterochromatin alone via GSK126 can recapitulate the LMNB1-/- which shows loss of facultative heterochromatin (Figure 5).

__ 5-Auhors mention throughout the manuscript the "actin confinement" component (actin cables localized at the top of the nucleus-not convincingly shown in fig. 2C). Some studies have reported the occurrence of perinuclear actin caps that surround the entire nucleus (DOI: 10.1038/s41467-018-04404-4; DOI: 10.1038/srep40953; DOI: 10.1038/ncb3387). Can the authors investigate the existence of such perinuclear actin ring on the LaminB-/-, LaminA-/- and VPA-treated cells? Could this ring affect NE rupture and shape? Also, if these perinuclear actin rings can be observed, what would they look like in CN03- and Y27-treated cells?__

The reviewer is requesting that we address possible changes to actin perinuclear cap and surrounding structure. We plan to address this concern by closer analysis of our current data and gather more data if needed.

__ Minor comments:__

We will address all the minor comments during revision.__

1-Please mention/discuss CytoD treatment in main text (Fig 2C).__

We plan to add text to address this concern.

__ 2-Can the authors comment on why CN03 treatment on VPA cells does not cause changes in blebs or NE rupture (fig. 3A, B)?__

We plan to comment on this point.

__ 3-I'd move fig. 2B to supplement.__

We prefer to keep this material in the main manuscript Figures as it supports a major support of the title and major conlcusion.

__ 4-In my opinion schematics on figs. 1A, 2A, 3E are confusing and do not add anything to the manuscript.__

We prefer to keep this material in the manuscript.

__ 5-There is something missing on the following sentence, please revise it: "We hypothesized that LMNA-/- nuclei do not show bleb-based behaviors because this perturbation cannot, due to reported disrupted nuclear-actin connections (Broers et al., 2004; Vahabikashi et al., 2022)."__

We have revised this sentence to read: “We hypothesized that LMNA-/- nuclei do not show bleb-based behaviors because of the reported disruption of nuclear-actin connections (Broers et al., 2004; Vahabikashi et al., 2022).”

__ 6-Can the authors explain in the Discussion why there is a decrease in gH2AX foci in VPA-treated cells and LaminA-/- cells upon CN03 treatment?__

We do not have an explanation at this time.

Significance

Nuclear deformation is a common event observed in homeostasis and disease and both extra-cellular physical cues and different cellular components play critical roles in nuclear morphology and integrity. It is well known that the actin cytoskeleton exerts a wide range of forces on the nucleus and causes nuclear deformation (via LINC complex) as cells migrate, grow or spread within complex microenvironments. The contribution of mutations of nucleo-skeleton components to nuclear abnormalities and rupture are well described. Additionally, more recently, the contribution of the actin cytoskeleton to nuclear integrity and morphology has also been characterized. However, the role played by actin contraction on nuclear shape and integrity, independently of actin confinement (exerted by the actin cables localized at the top of the nucleus), remain elusive. In this manuscript Pho et al., address this question using cells expressing NLS-GFP to detect nuclear rupture events and potential nuclear deformations. There is no real conceptual advance in this study but there is a novel finding as it shows that acto-myosin contraction affects nuclear integrity and morphology independently of dorsal actin cables (actin confinement). Moreover, the experiments were performed on flat 2D surfaces, a distant scenario from the 3D in vivo landscapes. As a classic cell biology study this manuscript has the potential to be of interest to basic researchers in the field of cell migration, crosstalk of nucleo-skeleton/cytoskeleton and nuclear mechano-sensing.

We appreciate that Reviewer states

• “role played by actin contraction on nuclear shape and integrity, independently of actin confinement (exerted by the actin cables localized at the top of the nucleus), remain elusive.”
• “In this manuscript Pho et al., address this question…” and “[our manuscript] is a novel finding as it shows that acto-myosin contraction affects nuclear integrity and morphology independently of dorsal actin cables (actin confinement).” Thus our manuscript provides a novel finding of interest to the cell biology community

While individually, chromatin decompaction via VPA, lamin B1 knockout (LMNB1-/-), and lamin A/C knockout (LMNA-/-) have been well-studied, there is no other publication that directly compares them and furthermore decouples the role of actin contraction from confinement.

__ __

Reviewer 2

The authentication of cell lines was not clear. The origin of the cell lines used was not clear. Were they immortalized? Did they examine primary MEFs? The authors did not seem to be aware of convincing data showing that Lamin B1 increases nuclear dispensability, whereas lamin B1 deficiency has the opposite effect.

Cell lines have been previously published multiple times, but originated from Shimi et al. 2011. We have revised the text to clarify that these are immortalized MEFs used in many previous studies cited in the main manuscript and the materials and methods, and not primary MEFs.

“ MEFs were immortalized with SV40 large T antigen by retroviral transduction of the gene encoding the SV40 large T antigen as previously described (Shimi et al., 2011, 2015).”

We also included citations to (Vahabikashi et al., PNAS 2022) which has recently compared many lamin knockouts and knockdowns.

The reviewer makes an unclear statement about lamin B1 levels and “dispensability” but provides no citations. As cited in the original manuscript lamin B1 null nuclei are one of the most studied models of nuclear blebbing and rupture (Vagas et al., Nucleus 2012; Hatch et al., JCB 2016; Young et al., MBoC 2020). __

The paper reads like a rough draft. Nomenclature inappropriate.__

The nomenclature used in this manuscript follows previous publications in the field including for chromatin notation (VPA, Stephens et al., 2017 and all previous manuscripts using this drug), lamin KOs (LMNB1-/- and LMNA-/-, Shimi et al. 2011.), and actin contraction and confinement are field appropriate.

We will revise the text to clarify nomenclature.

Significance

The impact of actin on blebs and nuclear shape is well established. The implications of these findings for distinct roles of the nuclear lamin proteins were not clear. The impact of the interventions on nuclear stiffness was not measured.

We agree with the reviewer that the impact of actin on nuclear shape is well established. However, the differential roles of actin contraction vs. confinement are not clear, as we state in the intro of the original manuscript. We believe our paper shows for the first time the separate role of actin contraction from actin confinement, where actin contraction is modulated, and we find that actin confinement measured by nuclear height remains the same. Reviewer #1 agrees that our data separating the roles of actin contraction and confinement is a novel finding (See Reviewer #1 Significance above).

The reviewer states that the distinct roles of lamins are not clear. We use different lamin knockout mutants as phenotypes of nuclear blebbing (LMNB1-/-, along side VPA) and abnormal nuclear shape measured as decreased nuclear circularity with no change in nuclear blebbing (LMNA-/-). These two different phenotypes of nuclear blebbing and abnormal shape also coincide with bleb-based nuclear ruptures and non-bleb-based nuclear ruptures, respectively (Figure 1). We do not examine the full role of lamins in this manuscript, as it is well beyond the scope of this work.

As cited in the original manuscript, interventions on chromatin (VPA) and lamins (LMNB1-/- and LMNA-/-) have been previously published and thus were not the focus. Nuclear stiffness measurement of perturbation of chromatin compaction via VPA has been published numerous times by our lab (Stephens et al., 2017,2018, 2019 MBoC; Berg et al., 2022 biorxiv) and others (Hobson et al., 2020 MBoC, Shimanoto et al., 2017). Nuclear stiffness in LMNB1-/- and LMNA-/- has been published in (Vahabikashi et al., PNAS 2022). We also have previously shown that depolymerization of actin does not impact nuclear stiffness (Stephens et al., 2017 MBoC), which would suggest that changes in actin contraction would not influence nuclear stiffness. This is supported by the fact that changes in actin contraction to not alter actin confinement pushing down on the nucleus resisting it (force balance) in all conditions except VPA Y27632 (Figure 2).

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 #4

Evidence, reproducibility and clarity

This article examines the cellular processes that predispose cells to nuclear blebbing and DNA damage in response to lamin and chromatin perturbations. The authors show key differences in these two types of perturbation and demonstrate a role for actin contractility. The experiments are well controlled and the data analysis generally rigorous. However, prior to acceptance, a number of issues must be fixed to improve the manuscript. I do not know the field sufficiently well to judge the novelty of the data.

Major issues:

• page 7, bottom: The authors state that measuring nuclear height gives an indication of confinement and force balance. But, if the nuclear mechanical properties have changed, then the nuclear height could change without any change in contractility. So, the authors would need to also verify that the level of contractility hasn't changed and that the mechanical properties haven't changed to really confirm that the cell height is a good measure of confinement. The level of contractility can be assessed by staining for pMLC. The nuclear mechanical properties may have been measured by others.
• In general, are the changes in contractility resulting from drug treatments sufficiently large to deform the nucleus? Can the authors show a time course of nuclear height in response to a treatment for WT for example? This would allow to link contractility to nuclear height.
• Page 9: The authors do not find any change in nuclear shape. Can they measure shape pre/post treatment on the same cells? It could be that the effect is lost in variability unless you do paired measurements?
• Page 11: the authors find nuclear ruptures unchanged in LMA -/- even when there is no contractility. They then state: "We hypothesized that LMNA-/- nuclei do not show bleb-based behaviors because this perturbation cannot, due to reported disrupted nuclear-actin connections". I do not understand this sentence.
• To characterise actin contractility better, it would be good to present images of the actin cables in each condition and pre/post treatments. This would allow to visually assess whether the morphology of the F-actin cytoskeleton has changed. This is one of the main topics of the study and as such it should be examined.
• On all bar charts, the authors should indicate: the number of independent experiments, the number of cells examined.
• I find the diagrams on Fig 1A, 2A etc do not help to illustrate what the authors think is happening. Can they redraw them in a more informative way?
• The abstract, introduction, and discussion are overly long and lack focus. These should be rewritten succinctly.

Minor issues:

• page 4: inhibitors of Rho-kinase will also modulate actin polymerisation indirectly through the action on Lim-kinase and cofilin.
• page 5, second paragraph: the authors should state that they are measuring the frequency of ruptures. At first, I thought this might be a mechanical strain.
• Page 7: In general, it may be useful to discuss the temporal evolution of the c/n and the circularity side by side. The change in circularity over time could be an indicator of mechanical strain, while the c/n would report on any transient loss of integrity of the nuclear membrane.
• Fig 1B: it would be nice to present the time course of the c/n as well.
• Fig S1: it might be interesting to characterise the dynamics/amplitude of the c/n for the different conditions. There doesn't appear to be any difference between the nuclear blebbing rupture and the non blebbing rupture. This suggests that the two phenomena (nuclear blebbing and nuclear rupture) are independent: i.e. rupture is not causally linked to blebbing.

Significance

This article examines the cellular processes that predispose cells to nuclear blebbing and DNA damage in response to lamin and chromatin perturbations. The authors show key differences in these two types of perturbation and demonstrate a role for actin contractility. The experiments are well controlled and the data analysis generally rigorous. However, prior to acceptance, a number of issues must be fixed to improve the manuscript. I do not know the field sufficiently well to judge the novelty of the data.

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, the authors set out to compare the contributions of chromatin and both A- and B-type lamins to the maintenance of nuclear shape and NE integrity, while also observing how actin contraction or confinement individually contribute to changes in nuclear shape and integrity. To do this, they used mouse embryonic fibroblasts (MEFs) expressing an NLS-GFP as a readout of nuclear shape and rupture, while perturbing chromatin compaction (using the HDAC inhibitor VPA) or using MEF lines devoid of A-type (LMNA -/-) or B-type (LMNB1 -/-) lamins, while simultaneously decreasing or increasing actin-mediating tensile forces but maintaining actin compressive forces by observing nuclear height. They found that increased actin contraction causes a higher instance of bleb-based nuclear shape changes and ruptures in WT and LMB1-/- cells, as well as cells with chromatin decompaction, but actin contraction did not impact the nuclear morphology or prevalence of bleb-based ruptures with a loss of LMNA. The authors also show that loss of LmnB1 causes chromatin decompaction phenotype, and loss of Lamin A/C creates and increase in bleb-based nuclear ruptures, but only under conditions of chromatin decompaction with VPA.

Major critiques:

1. Fig 2B- The authors use γMLC2 as a readout for actin contractility when CN03 and Y27632 is applied. While this method has been used previously as a readout for actin contractility, it would be more convincing if the authors included at least another technique to verify and increase or decrease in actin contractility in some of their conditions (e.g. traction force microscopy).
2. Fig 2C- the authors suggest that the shape changes and blebbing are not due to actin confinement on the nucleus, because nuclear height does not change. This is an large assumption made from only a small amount of data, considering that there could be increased confinement on the nucleus during actin contractility, yet it is too small to be measured by the side imaging technique described in the paper. The manuscript would benefit greatly from more experiments that could show that the blebs are being made even when there is no confinement.

Minor critiques:

1. Fig 2B- the statistical significance asterisks above the MLC2 relative fluorescence graph do not seem to be aligned appropriately with the bars, and it is difficult to know which asterisk belongs to which bar. Same is true for the nuclear height graph. The "vs. WT" box above the nuclear height graph is also confusing in that it is hard to see which statistical

Significance

The manuscript creates value in the field, in that the major findings of the relationship between lamins, chromatin, and actin and their respective impacts on nuclear shape and rupture are in agreement with findings from previous studies and therefore bolsters the current model on NE mechanics. The novelty in the paper comes from the reported finding that actin contraction and not confinement is what controls nuclear blebbing and ruptures, although this evidence seems to be limited to a single method (nuclear height measurements) presented in Fig 2 and Fig S3A. In contrast to the author's suggestions, it is possible that during actin contractility, actin cables running over the nucleus could be confining the nucleus to height changes that are too subtle to be measured using a confocal microscope, and there are forces acting on to top of the nucleus that could contribute to the blebbing phenotype observed. In my opinion, the conclusions drawn by the authors in this matter rely too few evidence.

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 authentication of cell lines was not clear. The origin of the cell lines used was not clear. Were they immortalized? Did they examine primary MEFs? The authors did not seem to be aware of convincing data showing that Lamin B1 increases nuclear dispensability, whereas lamin B1 deficiency has the opposite effect.

The paper reads like a rough draft. Nomenclature inappropriate.

Significance

The impact of actin on blebs and nuclear shape is well established. The implications of these findings for distinct roles of the nuclear lamin proteins were not clear. The impact of the interventions on nuclear stiffness was not measured.

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 Pho et al., characterize the effects of actin confinement versus acto-myosin contractility on nuclear deformation and nuclear envelope (NE) rupture/blebs. They use both genetic and pharmacological perturbations that impact nuclear shape and integrity such as chromatin decompaction and genetic deletion of nucleo-skeleton components (LaminA and LaminB). First, the authors show that modulation of acto-myosin contractility does not affect nuclear height (a proxy for the effects of actin confinement on the nucleus). This finding allowed them to study the effects of acto-myosin contractility on nuclear shape and integrity independently of actin-mediated nuclear confinement. Using fibroblasts expressing a NLS-GFP construct to assess nuclear shape and integrity, the authors show that increased acto-myosin contractility by treatment with Rho activator (CN03) leads to nuclear blebs and NE rupture events (curiously, they show that in LaminA-/- fibroblasts the NE rupture events were not bleb-based and were not affected by CN03). With these results, authors conclude that actin contraction is a major determinant of bleb-based NE rupture, independently of actin confinement. Next, since many studies have associated loss of nucleo-cytoplasmic compartmentalization with increased DNA damage, the authors addressed the occurrence of DNA damage foci in several contexts/perturbations. They find that increased NE rupture frequency by CN03 treatment does not lead to increased DNA damage. Based on the sole finding, authors then claim that DNA damage is mostly associated with abnormal shaped nuclei instead of ruptured nuclei. The data presented is overall convincing and well controlled but a more in-depth characterization of the DNA damage events in VPA-treated cells, LaminB-/- and LaminA-/- cells is needed to support the claim that DNA damage is not associated with NE events in their system. More details below.

Major comments:

1. In fig. 4, increased NE rupture frequency by CN03 treatment did not lead to an increase in gH2AX foci, so authors claim that DNA damage is not associated with NE events. However, the increase observed in NE rupture frequency is rather very subtle (from 1.5 to 3 in WT cells; from 2.5 to 4 in VPA treated ells and from 2 to 2.5 in LaminB-/- cells) so I wonder if such a minor increase might be sufficient to cause an increase in DNA damage foci. Moreover, DNA damage events are repaired within minutes and the analysis throughout this manuscript is made on snapshots. Therefore it becomes difficult to draw any conclusion regarding the absence of a link between NE rupture and DNA damage. To assess DNA damage events in a quantitative manner and show that abnormal shape (as opposed to NE rupture) is the main driver of DNA damage, authors should perform live cell imaging of cells co-expressing a DNA damage marker and a NE rupture marker to track DNA damage events following NE rupture.
2. Fig. 4C: to show that frequency of NE rupture is less important than abnormal nuclear shape for DNA damage appearance, authors need to quantify gH2AX in normal x blebbed x abnormal nuclei in each one of the conditions (untreated, CN03 and Y27). But again, this analysis must be complemented by live cell imaging experiments to track DNA damage foci appearance (or not?) following NE rupture events.
3. Since VPA treatment increases both the percentage of bleb-based NE rupture and NE rupture frequency in LaminA-/- cells, authors should show that gH2AX foci in this sample remain unaltered to further support their claim that frequency of NE rupture is less important than abnormal nuclear shape.
4. Does methylstat treatment rescue LaminB-/- phenotypes (blebs, ruptures, ...)?
5. Auhors mention throughout the manuscript the "actin confinement" component (actin cables localized at the top of the nucleus-not convincingly shown in fig. 2C). Some studies have reported the occurrence of perinuclear actin caps that surround the entire nucleus (DOI: 10.1038/s41467-018-04404-4; DOI: 10.1038/srep40953; DOI: 10.1038/ncb3387). Can the authors investigate the existence of such perinuclear actin ring on the LaminB-/-, LaminA-/- and VPA-treated cells? Could this ring affect NE rupture and shape? Also, if these perinuclear actin rings can be observed, what would they look like in CN03- and Y27-treated cells?

Minor comments:

1. Please mention/discuss CytoD treatment in main text (Fig 2C).
2. Can the authors comment on why CN03 treatment on VPA cells does not cause changes in blebs or NE rupture (fig. 3A, B)?
3. I'd move fig. 2B to supplement.
4. In my opinion schematics on figs. 1A, 2A, 3E are confusing and do not add anything to the manuscript.
5. There is something missing on the following sentence, please revise it: "We hypothesized that LMNA-/- nuclei do not show bleb-based behaviors because this perturbation cannot, due to reported disrupted nuclear-actin connections (Broers et al., 2004; Vahabikashi et al., 2022)."
6. Can the authors explain in the Discussion why there is a decrease in gH2AX foci in VPA-treated cells and LaminA-/- cells upon CN03 treatment?

Significance

Nuclear deformation is a common event observed in homeostasis and disease and both extra-cellular physical cues and different cellular components play critical roles in nuclear morphology and integrity. It is well known that the actin cytoskeleton exerts a wide range of forces on the nucleus and causes nuclear deformation (via LINC complex) as cells migrate, grow or spread within complex microenvironments. The contribution of mutations of nucleo-skeleton components to nuclear abnormalities and rupture are well described. Additionally, more recently, the contribution of the actin cytoskeleton to nuclear integrity and morphology has also been characterized. However, the role played by actin contraction on nuclear shape and integrity, independently of actin confinement (exerted by the actin cables localized at the top of the nucleus), remain elusive. In this manuscript Pho et al., address this question using cells expressing NLS-GFP to detect nuclear rupture events and potential nuclear deformations. There is no real conceptual advance in this study but there is a novel finding as it shows that acto-myosin contraction affects nuclear integrity and morphology independently of dorsal actin cables (actin confinement). Moreover, the experiments were performed on flat 2D surfaces, a distant scenario from the 3D in vivo landscapes. As a classic cell biology study this manuscript has the potential to be of interest to basic researchers in the field of cell migration, crosstalk of nucleo-skeleton/cytoskeleton and nuclear mechano-sensing.

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

We would like to first thank the reviewers for their patience with us given the delays in the generation of our revised manuscript. In addition to a maternity leave taken by Dr. Goyon, we took the reviewers comments very seriously and generated significant amounts of new data to address the insightful comments and suggestions of all three reviewers. The manuscript now has 9 figures, with 7 supplemental figures and 3 tables.

Our point-by-point rebuttal is below, but the major new additions are:

• Full analysis of bile acid species in the feces (to complement the liver and serum analysis provided in first submission). We also performed FDG-glucose PET analysis of the mice, which revealed significant alterations in proliferation in the gut in young MAPL KO mice. We did this in response to the concerns raised by reviewers 1 and 3 about the effects of the gut on bile acid regulation, and we discuss our findings below in response to reviewers, and within the revised manuscript.
• Our initial submission reported an Illumina approach for transcriptomics in livers of male wt and KO mice. We now completed RNAseq analysis on both male and female (littermate) control and MAPL KO mice at 3 months old. This validated what we had seen in the Illumina arrays, but allowed us a deeper look into transcriptional and sex specific changes that we present in response to review and within the revised manuscript. We have also expanded our feed/fast cycle analysis of the dynamic changes in gene expression of bile acid related pathways to further document the disruption in the feedback cycles regulating bile acid synthesis in liver.
• We have developed assays using primary hepatocytes from +/-MAPL mice for a better analysis of cell autonomous functions and bile acid secretion. This complements the tail-vein rescue experiments we had presented in initial submission.

---

---

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

In this study, the authors have investigated the impact of permanently silencing the expression of the mitochondrial anchored protein ligase (MAPL) in mice on bile acid (BA) metabolism through the alteration of ABCD3 SUMOylation. This ABC pump mediates the uptake of C27-BAs by peroxisomes and hence determines the shortening of the BA sidechain. In addition, other aspects of general metabolism have also been investigated. The study is highly relevant and contains valuable information__.

AUTHORS: We sincerely thank the reviewer for their supportive comments on the value of our work.

Major points

R1: 1) Sidechain shortening is essential for the synthesis of primary C24 BAs. This study suggests that the entrance of C27 BAs in peroxisomes, which depends on ABCD3 activity, is reduced by MAPL-dependent ABCD3 SUMOylation. Thus, knocking out MAPL in mice results in enhanced BA accumulation in serum and liver, presumably by facilitated uptake of C27 by the peroxisomes and stimulation of de novo synthesis of primary BA. Indeed, a decreased C27/C24 BA ratio was found. However, the results suggest that Cyp7a1 is not the main checkpoint for the control of BA synthesis or that Fxr/Fgf15/Cyp7a1 pathway is also affected by MAPL manipulation because Cyp7a1 expression, which could be expected to be downregulated in response to enhanced BA levels, is not affected in MAPL knockout. Moreover, no change in Fgf15 was found (Suppl. Fig. 2C, 2D, 2G). The authors must discuss these surprising findings.

AUTHORS: We entirely agree with the reviewer that the lack of feedback to downregulate bile acid synthesis through the canonical pathways was very unusual and is one of the novel aspects of our study. While the circulating bile was significantly elevated, this was not sensed by the pathways in the gut or liver to downregulate CYP7A1 through the activation of bile receptors FXR/FGF15. We have rewritten a great deal of the manuscript to be clearer about this, as we also feel the MAPL KO presents us a very unique model of bile acid dysregulation with many unexpected observations. In revision we completed an analysis of ~40 bile acid species within the feces to understand what may be happening in the gut. Interestingly, bile acid levels were decreased in feces, in contrast to the elevations seen in gut and liver (New Fig3). This prompted us to consider a block in bile acid delivery to the gut, or cholestasis. However, such a pathology is generally lethal, yet MAPL KO mice live past 2 years. Histological analysis (presented in New Fig 3) did not reveal any obstructions, necrosis or pathology in the bile canaliculi. Bomb calorimetry of the feces showed equivalent calories in KO mice (New Fig3), suggesting that digestion of food was fully intact, something that would have been altered if bile was absent. Lastly, the secondary bile acids that are generated by the microbiome were elevated in serum and liver, indicative of the successful transit of these bile species through the gut. Therefore, we conclude that the bile does reach the gut, yet appears to be significantly reabsorbed back into circulation without alerting the bile sensing pathways to secrete more FGF15. Ultimately, we have not answered the initial question, since we do not know yet how MAPL is required, directly or indirectly, for bile acid feedback loops. But we realize now that this will take significant effort to resolve mechanistically, something we will continue to work on in the next stage of our project.

R1: 2) The authors discussed that the alternative acidic pathway is responsible for these changes, but Cyp27a1 was, in fact, moderately downregulated in MAPL knockout mice.

AUTHORS: We apologize for the confusion. Yes, Cyp27A1 is moderately downregulated in MAPL KO mice, seen now within RNAseq analysis (New Fig 2C) and the western blots (Fig 3C) which we meant to say could reflect a specific feedback loop to inhibit bile acid synthesis from the acidic pathway, rather than through canonical, FXR/FGF15 mediated changes in Cyp7A1. We considered that very little is known about the regulation of the acidic pathway, and perhaps MAPL effects on bile acid metabolism may be more dominant in this loop. However, clearly it remains unresolved how the elevated bile, even in liver, goes undetected by FXR to downregulate Cyp7A1. We have tried to approach our results and discussion in a more systematic way to make these points clearer.

R1: 3) Serum BAs may reflect a higher BA pool. Nevertheless, this has not been assayed. Enhanced flow of C27-BA precursors into peroxisomes is consistent with increased C24-BA production and reduced intrahepatic concentration of C27-BA in MAPL knockout mice (Suppl. Table 2). However, it is not explained why C27-BA serum concentrations were increased in these animals (Suppl. Table 2 and Suppl. Fig. 2B).

AUTHORS: We thank the reviewer for pointing that out. We added quantification in the feces of the bile acid species (New Fig2C). Surprisingly we found decreased levels of C27 and C24 bile acids. We are speculating that some of the increased bile acid levels in the serum are due to increased synthesis/flux through the hepatocyte peroxisomes and some due to reabsorption.

R1: 4) C27-BAs have been described as more toxic species than most C24-BAs. In the liver of MAPL knockout mice, C27-BAs levels were decreased (Suppl. Table 2). Other toxic species such as DCA and CDCA were not markedly changed. Muricholic acids and ursodeoxycholic acid, which were increased, are believed to be non-toxic or even hepatoprotective. Therefore, the relationship between changes in BA homeostasis and liver carcinogenesis should be better justified.

AUTHORS: We apologize for generalizing too much in assigning elevated bile acid species as potential drivers/contributors of tumorogenesis. We have made note in the text of the reviewers points, including references to the protective nature of some bile species. We cannot yet pinpoint the precise cause of cellular transformation but have tried to balance the discussion around potential changes in 1) proliferative signaling cascades potentially linked to bile signaling, 2) ER stress, which has been linked to tumorogenesis, and 3) the protection against cell death pathways seen in MAPL KO cells.

R1: 5) SUMOylation may affect transporters which may simulate certain cholestasis with retention in serum of BAs. Expression levels of basolateral Ntcp, Oatps, and canalicular Bsep are required to better understand BA homeostasis. Besides, biliary secretion in MAPL knockout mice would give relevant information on what is actually happening in the biliary function of these animals.

AUTHORS: We thank the reviewer for an excellent point. To get a better view of the transcriptional changes of the transporters highlighted here, but also of all genes in liver, we completed RNAseq analysis. This showed no change in the mRNA levels of the transporters highlighted, so we performed qRT-PCR analysis from livers during a feed/fast experiment to determine whether the dynamic behavior of expression may be altered upon MAPL loss (New Fig 4). Importantly, we found that the transporters expression was unchanged (except for one of the Oatp which would limit hepatocyte reabsorption). We also added bile secretion from primary hepatocytes reproducing the phenotype. This reinforces our point that MAPL loss affect primarily bile acid flux through the peroxisome and that is enough to have increased bile acid in the serum. Lastly, to test whether bile was successfully transiting into the gut we completed the bile acid analysis of feces, along with bomb calorimetry, PET analysis and histology of the gut, all of which indicate that bile flux to the gut is intact.

__Reviewer #1 (Significance (Required)):

The study is relevant and original.__

AUTHORS: We thank the reviewer for appreciating the strengths of our study.

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

The present manuscript described novel interacting partners of a mitochondrial/peroxisomal Sumoylation ligase MAPL and describes the phenotype of a newly generated MAPL KO mouse model.

Major comments:

R2: The authors describe in the introduction that MAPL has multiple functions, including a role in mitophagy,mitochondrial division, inflammation and cell death. New is a role in regulation of peroxisomal bile salt handling. Also the role in hepatic cell proliferation in vivo has not been demonstrated before. The individual findings are generally convincing. However, the relation between the large number of observations is not clear. The authors postulate that multiple aspects of the MAPL KO mice are related to direct effects on PMP70/ABCD3 sumoylation and/or to effects on bile salts. This connection is highly speculative and mechanistically underexplored. As MAPL function was already implicated in many processes unrelated to bile salts/ABCD3, alternative explanations are likely.

AUTHORS: We appreciate these critical comments, and agree that MAPL has distinct substrates that play important roles in multiple aspects of mitochondrial (and now peroxisomal) signaling, survival and metabolism. This certainly makes it difficult to assign a given substrate to a broad set of pathologies in a KO mouse model. Our lab has been working on MAPL for over 15 years, and a major question has been the physiological function of MAPL in vivo. Our report that the primary phenotypes appear in liver, that effects global metabolism (lean phenotype, insulin sensitivity), proliferation and cancer provide the first evidence of the importance of MAPL in metabolism. Our BioID approach to identify MAPL partners led us to ABCD3, the peroxisomal bile acid transporter, in addition to the identification of established proteins of the mitochondrial and peroxisomal fission machineries. Furthermore, we provide critical new evidence that MAPLs primary role is not to regulate the degradation of its substrate partners, since they were not stabilized upon inhibition of the proteasome. We confirmed the interaction and SUMOylation of ABCD3 from liver tissue using multiple approaches (BioID, co-IP, SIM beads, glycerol gradients), therefore we do not consider the interaction between the two proteins as speculative. We agree that more will need to be done to develop structure/functional analysis of the SUMOylated bile acid transporter. Important to this study, functional data in vivo demonstrates major increases in bile acid production in liver and (new to the revision) primary hepatocytes, consistent with a role for MAPL mediated SUMOylation to gate ABCD3, the only primary bile acid transporter in peroxisomes. We have attempted to position our findings within the context of MAPL function in other pathways and broadened our discussion in terms of all mechanisms and phenotypes.

In this revised manuscript we expanded our analysis into the gut, given the important role of the gut in bile acid homeostasis. In searching for an explanation for the disruption in FXR/FGF15 responsiveness, we observed a striking proliferative phenotype in the duodenum. The limited proliferation in duodenum is consistent with previous work showing that bile acids can promote proliferation through a number of mechanisms, from the signaling of TGR5 bile receptors to YAP activation. It may also reflect cell autonomous functions of MAPL in enterocytes responsible for the suppression of proliferation, however the limitation of the proliferative phenotype to the top section of the gut does suggest a link to bile. We have included these data, along with a full bile acid analysis of feces, in the revised manuscript given the essential role of the gut as a driver of the feedback loop for bile acid homeostasis. We hope the reviewer will now be convinced that our work places MAPL as a key metabolic regulator offering a new animal model that highlights some very unusual bile acid phenotypes, and a model of spontaneous hepatocellular carcinoma. These are unexpected phenoptypes for a MAPL KO animal given all previous work into MAPL/MUL1.

R2: Similarly, the metabolic consequences of bile salt signalling (the authors postulate this may occur via TGR5) versus effects of the ER-stress/FGF21 pathway remain unclear.

AUTHORS: It is not entirely clear to us which metabolic consequences the reviewer refers to in this concern, so hopefully we will answer the question as intended. Assuming the reviewer refers to the lean, insulin sensitive phenotype of MAPL KO mice, we understand there remains some speculation in the relationship between the bile acids or FGF21 as drivers of the insulin sensitivity. The phenotype mimics FGF21 overexpression, as this hepatokine has been long linked to leanness and insulin sensitivity in mice. The curious finding was in our tail vein rescue experiments, even empty adenovirus induced ER stress, so we could not test if MAPL expression would rescue CHOP expression. Yet we had a full restoration of both FGF21 and circulating bile. This indicates that the ER stress is not the main driver of FGF21 expression (or bile secretion) in this system. Given the direct interactions we observed between MAPL and the bile acid transporter, we hypothesized that the rescue of MAPL would return the gating function of ABCD3, and perhaps that the bile acids themselves were driving FGF21 expression. This is consistent with a 2018 study demonstrating that FGF21 signaling resulted in the downregulation of bile acid synthesis (PMID: 29615519), a potential feedback loop to explain the downregulation of many bile acid enzymes seen in our RNAseq analysis. We have been more careful to state the limitations and outstanding questions in the study. It will take many additional mouse crosses, knock-in models carrying mutations in ABCD3, and many other experiments to resolve this question fully. We sincerely hope the reviewer can agree that the observations are robust and well controlled, and will open new avenues of research in the future.

R2: The title and discussion is too speculative in my opinion, in particular the claim linking ABCD3 activity to all the metabolic effects observed in the MAPL KO.

AUTHORS: We have changed the title of the manuscript to better reflect the global consequences of MAPL loss and the novelty of our findings overall.

- 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.

R2: Experimental support for an altered role of ABCD3 activity as CAUSAL for the observed phenotype is essential

AUTHORS: We argue that our data has established the MAPL-dependent SUMOylation of ABCD3 in vivo using the SIM-bead pull down, our BioID identified this transporter as the top partner of MAPL, validated with immunoprecipitation from liver in floxed and MAPLKO mice, and we observe biochemical alterations in the oligomeric state of ABCD3. Identifying the SUMO sites and generating CRISPR KI mice to confirm effects on bile acid flux would represent another year (at least) of work. We strongly believe that our study provides a number of very important new insights into bile acid metabolism with phenotypes that have not been seen before (as explained by Rev1). We understand that this will be a final documentation of the structure/function relationship between MAPL and ABCD3, but we have established this interaction and substrate/enzyme pairing between a newly identified bile acid transporter (for which very little work has been done anywhere), and an evolutionarily conserved SUMO E3 ligase.

- 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.

R2: As Sumoylation sites can be predicted to some level, an MAPL-insensitive ABCD3 protein could be made and used to link effects of ABCD3 sumoylation to MAPL and consequences of MAPL deficiency. Minimally, data linking the modest effects on ABCD3 activity (for example by PMP70 knockdown in vivo) on the observed phenotype of MAPL KO is required to support the currents aims.

AUTHORS: Knocking down ABCD3/PMP70 would be possible with tail vein injection. However, the loss of ABCD3 would give the opposite phenotype, where bile acid production would be lost, as already documented in Ferdinandusse et al 2015 (PMID: __25168382). ABCD3 is the only known bile acid transporter in peroxisomes so we do not agree that our phenotypes are so obviously explained by another mechanism, nor why the reviewer considers the effects on bile acid to be “modest”? We suggest, based on established paradigms for the SUMOylation of ion transporters, that the SUMOylation would gate the transporter to inhibit it. We agree this is a next step, but it is beyond the scope of this manuscript. __

• 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? Yes.

R2: However, the initial and main finding of the manuscript, the identification of ABCD3 as MAPL interacting partner is plotted somewhat vague. Seems like data is from a single experiment, while the method section suggests otherwise

AUTHORS: We have repeated these experiments (BioID, co-IP, SIM bead experiments) numerous times from multiple mouse livers and cell lines. We have ensured the numbers are in the methods and legends. We would also submit that the interaction with ABCD3 is not, in fact, the main finding of the manuscript. The entire characterization of the mouse pathology, ending in the spontaneous development of hepatocellular carcinoma, will be of high interest to researchers in the fields of metabolism, liver biology and cancer, MAPL/MUL1 function, and insulin sensitivity. The BioID and interactomics was (to us) also very important for the things it did NOT find, for example, the substrates others consider to be regulated by MUL1 ubiquitination. Our experiments with MG132 clearly show that the candidate substrates identified in BioID are not targeted for degradation, including Mfn2 and others. MAPL loss did not lead to changes in mitochondrial or peroxisome mass, nor did it significantly alter the gene expression of these proteins (new RNAseq analysis). While some of these aspects represent negative data, it is an important, in vivo demonstration that MAPL is not a key player in mitochondrial quality control. In this sense, the entire study is highly unexpected, with such clear phenotypes in global metabolism, bile acid and liver specific effects, proliferation and cancer.

Minor comments:

R2: Abstract states: "BioID revealed the peroxisomal bile acid transporter ABCD3 as a primary MAPL interacting partner, which we show is SUMOylated in a MAPL-dependent manner." The method aspect of this sentence is too unclear as it assumes all readers know what BioID entails.

AUTHORS: We apologize for the confusion and have clarified that sentence.

R2: The abstract also states that increased bile salt secretion is occurring. No experimental data supporting increased hepatocellular bile salt secretion is provided, only increased serum levels, which is not the same.

AUTHORS: We thank the reviewer for pushing us to examine bile acid secretion from primary hepatocytes isolated from MAPL KO or littermate control mice. This took us some time since the MAPL KO hepatocytes didn’t seed as well as control cells, but we adapted our protocols and the cells adhered well. These experiments showed that MAPL KO hepatocytes produce 2-3X more bile acids than wild type hepatocytes over a 48 hours culture period. Therefore we have confirmed that the hepatocytes are producing more bile.

R2: How was FGF15 measured? The methods section is unclear about this, and the legends indicates this was measured by ELISA. Earlier paper suggests that FGF15 is not easily detectable and controls for the elisa should thus be included (PMID: 26039452).

AUTHORS: We quantified FGF15 by ELISA using established protocols without any difficulty, and have included all standards and controls in the excel sheets. The paper cited is from 2015, which is nearly 10 years ago so it appears that these tools have improved.

R2: Figure 1D; last lane with the duplo of the rescue with the mutant MAPL seems missing, only single value is plotted.

AUTHORS: The quantification presents n=3 biological replicates, the figure is a representative immunoblot of the 3 replicates, where only one mutant MAPL rescue is loaded. We clarified that in the legend.

Reviewer #2 (Significance (Required)):

- Describe the nature and significance of the advance (e.g. conceptual, technical, clinical) for the field.

R2: The individual findings/observation are very interesting. However, as the causal link and relative contribution of the multitude of processes affected in the MAPL KO remains unclear current impact is limited.

AUTHORS: We respectfully disagree with the reviewer that the impact is limited by the fact that we are studying an E3 ligase with multiple substrates. Clearly we understand that MAPL has functions at mitochondria regulating DRP1 and other processes, as we have worked on MAPL for over years. The novelty and importance of this study is that it is the first full characterization of a MAPL KO mouse that presents with very unexpected phenotypes that will be used to advance the field in multiple ways. The identification of ABCD3 as a substrate represents the first in peroxisomes to be examined, and very little is known about the regulation of these transporters. Even if there are additional functions of MAPL at play in liver (which I’m sure there are), linking it to bile acid flux, is a novel finding.

- Place the work in the context of the existing literature (provide references, where appropriate).

R2: A highly related manuscript recently appeared: mitochondrial ubiquitin ligase MARCH5 is a dual-organelle locating protein that interacts with several peroxisomal proteins. Peroxisomal MARCH5 is required for mTOR inhibition-induced pexophagy by binding and ubiquitinating PMP70 (J Cell Biol. 2022 Jan 3; 221(1): e202103156.) This is not discussed at all. However, it supports that better scientific insight into regulation of peroxisomal processes, including the activity of ABCD3/PMP70 is very relevant to the field.

AUTHORS: We apologize for omitting the study of MARCH5 in our manuscript. The reviewer is correct that this highlights the unique function of MAPL in the regulation of the transporter through SUMOylation. MAPL loss does not alter the turnover or expression of ABCD3/PMP70 (or peroxisomes for that matter), which is the opposite of MARCH5.

• State what audience might be interested in and influenced by the reported findings. An audience interesting in peroxisomal function and/or bile salt signalling/toxicity
• 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. R2: bile salt signalling; transport

AUTHORS: We understand that the reviewer focused on the first figure (of 9), and perhaps did not appreciate the unique findings we have made in characterizing a very novel bile acid phenotype where the feedback loops are interrupted. The links to cancer were also not mentioned by the reviewer, something we also feel strongly about since it is consistent with the roles of MAPL in cell death pathways. The establishment of a novel animal model of HCC is of value to the community.

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

R3: In the manuscript „SUMOylation of ABCD3 restricts bile acid synthesis and regulates metabolic homeostasis", the authors showed that MAPL has a critical role in regulation of bile acid synthesis and loss of MAPL leads to changes in metabolism and development of liver cancer. The findings are of certain interest, However, before publication of the manuscript several shortcomings have to be clarified.

AUTHORS: We thank the reviewer for recognizing that the identification of MAPL as having a critical role in metabolism, bile acid signaling and cancer is of importance to the field.

Major Comments: R3: Since levels of muricholic acid is drastically increased, the authors should investigate Cyp2c70 expression since this enzyme is responsible for muricholic acid synthesis. Is there are direct regulatory effect of MAPL on Cyp2c70?

AUTHORS: We thank the reviewer for the suggestion, we measured Cyp2c70 and found no change in MAPL KO livers, as seen in our new RNAseq analysis and qRT-PCR (New Fig4).

R3: Does MAPL directly regulate the changes on cyp enzymes or is the regulation indirect via acting on the nuclear receptors known to regulate bile acid synthesis such as FXR, CAR, PXR. Please provide data on that.

AUTHORS: We completed RNAseq analysis from livers of control and MAPLKO mice to generate a more complete picture of precise transcriptional changes in all genes. We also looked specifically at FXR, LXR, PXR and PPARa target genes using qRT-PCR approaches from livers in feed/fasting experiments to examine dynamic changes in expression. Most were unchanged and responded normally, with the exception of some PPARs target genes that were increased, supporting the metabolism necessary to handle high levels of bile acid flux.

R3: The authors show that Cyp4a14 is increased due to loss of MAPL. Cyp4a14 is also a downstream target of PPARa. The authors should provide data on PPARa signalling in MAPLKO mice, especially on beta oxidation. This may explain why MAPL KO mice are lean.

AUTHORS: See previous response

R3: The authors did not investigate bile duct proliferation and activation of cholangiocytes, features which often occur in the context of changes in bile acid homeostasis. Do MAPL. KO mice show increased ductular proiliferation and reactive cholangiocyte phenotype? Please provide data such as expression and staining of CK19, KI67, OPN, VCAM • EGR1 and EGFR are key regulators in HCC and are known to be regulated by bile acids. The authors should investigate whether these key regulators may play a role in development of HCC in MAPL KO mice.

AUTHORS: We thank the reviewer for these suggestions. We have the Ki67 data included in Figure 8. It shows increased hepatocyte proliferation but not cholangiocytes. Moreover our histology stainings do not support any change in canicular structure. We also measured the EGF receptor activation in our model and found no change (Supplemental Figure 3). We also tried to find other indication of inflammation (as suggested), in our RNAseq dataset, we can find some known inflammatory signals like SPP1/Osteopontin, VCAM1, CCL2, CD68 being increased. However, the pathway analyses did not reveal any increased inflammatory status, which is also supported by absence of immune cell infiltration. It is possible that some immune or inflammation remodeling is happening but not at a large scale and not following the canonical inflammatory liver diseases.

Reviewer #3 (Significance (Required)):

R3: The finding that loss of MAPL is involved in regulation of bile acid synthesis is of certain interest for the field of cholestatic liver and bile duct injuries. MAPL KO mice might be an interesting model to study potential therapeutics for these diseases. Furthermore, the fact that MAPL KO mice develop spontaneous HCC is also of particular interest, since such models are quite rare.

AUTHORS: We thank the reviewer for finding our work ‘of particular interest’

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 manuscript „SUMOylation of ABCD3 restricts bile acid synthesis and regulates metabolic homeostasis", the authors showed that MAPL has a critical role in regulation of bile acid synthesis and loss of MAPL leads to changes in metabolism and development of liver cancer. The findings are of certain interest, However, before publication of the manuscript several short cummings have to be claryfied.

Major Comments:

• Since levels of muricholic acid is drastically increased, the authors should investigate Cyp2c70 expression since this enzyme is responsible for muricholic acid synthesis. Is zhere are direct regulatory effect of MAPL on Cyp2c70?
• Does MAPL directly regulate the changes on cyp enzymes or is the regulation indirect via acting on the nuclear receptors known to regulate bile acid synthesis such as FXR, CAR, PXR. Please provide data on that.
• The authors show that Cyp4a14 is increased due to loss of MAPL. Cyp4a14 is also a downstream target of PPARa. The authors should provide data on PPARa signalling in MAPL KO mice, especially on beta oxidation. This may explain why MAPL KO mice are lean.
• The authors did not investigate bile duct proliferation and activation of cholangiocytes, features which often occur in the context of changes in bile acid homeostasis. Do MAPL. KO mice show increased ductular proiliferation and reactive cholangiocyte phenotype? Please provide data such as expression and staining of CK19, KI67, OPN, VCAM
• EGR1 and EGFR are key regulators in HCC and are known to be regulated by bile acids. The authors should investigate whether these key regulators may play a role in development of HCC in MAPL KO mice.

Significance

The finding that loss of MAPL is involved in regulation of bile acid synthsesis is of certain interest for the field of cholestatic liver and bile duct injuries. MAPL KO mice might be an interesting model to study potential therapeutics for these diseases. Furthermore, the fact that MAPL KO mice develop spontaneous HCC is also of particular interest, since such models are quite rare.

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:

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

The present manuscript described novel interacting partners of a mitochondrial/peroxisomal Sumoylation ligase MAPL and describes the phenotype of a newly generated MAPL KO mouse model.

Major comments:

• Are the key conclusions convincing?

The authors describe in the introduction that MAPL has multiple functions, including a role in mitophagy,mitochondrial division, inflammation and cell death. New is a role in regulation of peroxisomal bile salt handling. Also the role in hepatic cell proliferation in vivo has not been demonstrated before. The individual findings are generally convincing. However, the relation between the large number of observations is not clear. The authors postulate that multiple aspects of the MAPL KO mice are related to direct effects on PMP70/ABCD3 sumoylation and/or to effects on bile salts. This connection is highly speculative and mechanistically underexplored. As MAPL function was already implicated in many processes unrelated to bile salts/ABCD3, alternative explanations are likely. Similarly, the metabolic consequences of bile salt signalling (the authors postulate this may occur via TGR5) versus effects of the ER-stress/FGF21 pathway remain unclear. - Should the authors qualify some of their claims as preliminary or speculative, or remove them altogether?

Yes, the title and discussion is too speculative in my opinion, in particular the claim linking ABCD3 activity to all the metabolic effects observed in the MAPL KO. - 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.

Experimental support for an altered role of ABCD3 activity as CAUSAL for the observed phenotype is essential - 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.

As Sumoylation sites can be predicted to some level, an MAPL-insensitive ABCD3 protein could be made and used to link effects of ABCD3 sumoylation to MAPL and consequences of MAPL deficiency. Minimally, data linking the modest effects on ABCD3 activity (for example by PMP70 knockdown in vivo) on the observed phenotype of MAPL KO is required to support the currents aims. - 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?

Yes. However, the initial and main finding of the manuscript, the identification of ABCD3 as MAPL interacting partner is plotted somewhat vague. Seems like data is from a single experiment, while the method section suggests otherwise

Minor comments:

• Specific experimental issues that are easily addressable.
• Are prior studies referenced appropriately?
• Are the text and figures clear and accurate?

Abstract states: "BioID revealed the peroxisomal bile acid transporter ABCD3 as a primary MAPL interacting partner, which we show is SUMOylated in a MAPL-dependent manner." The method aspect of this sentence is too unclear as it assumes all readers know what BioID entails. The abstract also states that increased bile salt secretion is occurring. No experimental data supporting increased hepatocellular bile salt secretion is provided, only increased serum levels, which is not the same.

How was FGF15 measured? The methods section is unclear about this, and the legends indicates this was measured by ELISA. Earlier paper suggests that FGF15 is not easily detectable and controls for the elisa should thus be included (PMID: 26039452). Figure 1D; last lane with the duplo of the rescue with the mutant MAPL seems missing, only single value is plotted. - Do you have suggestions that would help the authors improve the presentation of their data and conclusions?

no

Significance

• Describe the nature and significance of the advance (e.g. conceptual, technical, clinical) for the field.

The individual findings/observation are very interesting. However, as the causal link and relative contribution of the multitude of processes affected in the MAPL KO remains unclear current impact is limited. - Place the work in the context of the existing literature (provide references, where appropriate).

A highly related manuscript recently appeared: mitochondrial ubiquitin ligase MARCH5 is a dual-organelle locating protein that interacts with several peroxisomal proteins. Peroxisomal MARCH5 is required for mTOR inhibition-induced pexophagy by binding and ubiquitinating PMP70 (J Cell Biol. 2022 Jan 3; 221(1): e202103156.) This is not discussed at all. However, it supports that better scientific insight into regulation of peroxisomal processes, including the activity of ABCD3/PMP70 is very relevant to the field. - State what audience might be interested in and influenced by the reported findings.

An audience interesting in peroxisomal function and/or bile salt signalling/toxicity - 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.

bile salt signalling; transport

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 study, the authors have investigated the impact of permanently silencing the expression of the mitochondrial anchored protein ligase (MAPL) in mice on bile acid (BA) metabolism through the alteration of ABCD3 SUMOylation. This ABC pump mediates the uptake of C27-BAs by peroxisomes and hence determines the shortening of the BA sidechain. In addition, other aspects of general metabolism have also been investigated. The study is highly relevant and contains valuable information.

Major points

1. Sidechain shortening is essential for the synthesis of primary C24 BAs. This study suggests that the entrance of C27 BAs in peroxisomes, which depends on ABCD3 activity, is reduced by MAPL-dependent ABCD3 SUMOylation. Thus, knocking out MAPL in mice results in enhanced BA accumulation in serum and liver, presumably by facilitated uptake of C27 by the peroxisomes and stimulation of de novo synthesis of primary BA. Indeed, a decreased C27/C24 BA ratio was found. However, the results suggest that Cyp7a1 is not the main checkpoint for the control of BA synthesis or that Fxr/Fgf15/Cyp7a1 pathway is also affected by MAPL manipulation because Cyp7a1 expression, which could be expected to be downregulated in response to enhanced BA levels, is not affected in MAPL knockout. Moreover, no change in Fgf15 was found (Suppl. Fig. 2C, 2D, 2G). The authors must discuss these surprising findings.
2. The authors discussed that the alternative acidic pathway is responsible for these changes, but Cyp27a1 was, in fact, moderately downregulated in MAPL knockout mice.
3. Serum BAs may reflect a higher BA pool. Nevertheless, this has not been assayed. Enhanced flow of C27-BA precursors into peroxisomes is consistent with increased C24-BA production and reduced intrahepatic concentration of C27-BA in MAPL knockout mice (Suppl. Table 2). However, it is not explained why C27-BA serum concentrations were increased in these animals (Suppl. Table 2 and Suppl. Fig. 2B).
4. C27-BAs have been described as more toxic species than most C24-BAs. In the liver of MAPL knockout mice, C27-BAs levels were decreased (Suppl. Table 2). Other toxic species such as DCA and CDCA were not markedly changed. Muricholic acids and ursodeoxycholic acid, which were increased, are believed to be non-toxic or even hepatoprotective. Therefore, the relationship between changes in BA homeostasis and liver carcinogenesis should be better justified.
5. SUMOylation may affect transporters which may simulate certain cholestasis with retention in serum of BAs. Expression levels of basolateral Ntcp, Oatps, and canalicular Bsep are required to better understand BA homeostasis. Besides, biliary secretion in MAPL knockout mice would give relevant information on what is actually happening in the biliary function of these animals.

Significance

The study is relevant and original.

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

If you wish to submit a preliminary revision with a revision plan, please use our "Revision Plan" template. It is important to use the appropriate template to clearly inform the editors of your intentions.]

1. General Statements

We were naturally pleased to read the enthusiasm coming from both reviewers. Both mentioned that an extension to experimentation in cells would increase the impact of the study, even though both recognize that the biophysical and biochemical experiments constitute a study that is significant and interesting to a broad readership.

2. Point-by-point description of the revisions

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

This manuscript by Bryan et al., describes the use of Hydrogen/Deuterium-exchange Mass Spectrometry (HXMS) as a powerful tool to identify key amino acid residues and associated interactions driving liquid-liquid demixing. They have particularly focused on the Chromosomal Passenger Complex (CPC), an important regulator of chromosome segregation, which has recently been shown to undergo liquid-liquid demixing in vitro. Their work presented here allowed them to identify a few key electrostatic interactions as molecular determinants driving the liquid-liquid demixing of the CPC. Their work also shows that crystal packing information of protein molecules, where available, can provide valuable insight into likely factors driving liquid-liquid demixing.

Major comments:

[#1] A previous study by Trivedi et al., NCB 2019 identified an unstructured region in Borealin (aa residues 139-160) as the main region driving the phase separation of CPC. Interestingly, this region only shows a moderate reduction in HX upon liquid-liquid demixing. But no experiments or discussions related to this observation are presented in the current version of the manuscript.

In the Trivedi et al. paper, the authors were careful to state that the region of borealin between 139-160 contributed to phase separation, but there was clearly a remaining propensity to phase separate in vitro in the mutant. Thus, it is fully expected that there should be other regions in the complex that contribute to phase separation. It was satisfying that this region was independently identified in the hydrogen-deuterium exchange experiments and we suggest that a “moderate” reduction is consistent with a protein condensate having liquid properties. Since this region was already characterized we have focused our work in this paper to the new region identified by the hydrogen-deuterium exchange experiments.

[#2] In the absence of cellular data on if and how these mutations (within the triple-helical bundle region) affect CPC's ability to phase separate in cells, the implication of this work is very limited - One can't say for sure these are interactions driving phase separation of CPC in a cellular environment. In the absence of any cellular data with the mutants described here, much of the discussion on the possible roles of CPC phase separation in cells does not appear relevant to this manuscript. I would suggest that the authors focus mainly on highlighting the power of using HXMS as a tool to characterise the molecular determinants of liquid-liquid demixing at a relatively high resolution.

We have now added cellular data in the form of one of the key experiments used to explore CPC liquid-liquid demixing utilizing the Cry2 optogenetic system for inducible dimerization. The results of testing WT Borealin versus the mutant we identified is defective in droplet formation are shown in the all new Fig. 6. Some relation of our overall findings, encompassing observations made with purified components and now in cells, to the cellular function of the CPC is pertinent. In light of the reviewer comments, we have also reduced this aspect in the discussion (see the substantial edits on pg. 12).

Minor comments:

[#3] The authors should ensure that the introduction cites relevant literature thoroughly. For example, where the potential role of Borealin residues 139-160 in conferring phase separation properties to the CPC is mentioned, the authors failed to cite Abad et al., 2019, which showed the contribution of the same Borealin region in conferring nucleosome binding ability to the CPC.

We have made this particular change on pg. 4 and also have gone through to ensure we are appropriately citing relevant literature.

Reviewer #1 (Significance (Required)):

This is a highly relevant and significant work, particularly considering the rapidly growing list of examples for Phase separation of proteins/protein assemblies and their potential biological roles (in spite of ongoing debates in the field about the cellular relevance of several phase separation claims). The data presented in this manuscript are solid and convincingly establish HXMS as a useful tool to characterise molecular interactions driving liquid-liquid demixing. Considering its applicability to characterise wide-ranging protein assemblies implicated in phase separation, this work will be of interest to a broad readership.

We thank the reviewer for the strong praise of the significance of our study.

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

In this manuscript, using the technique of hydrogen/deuterium-exchange mass spectrometry (HXMS), the authors have tried to gain insights into the structure of the chromosomal passenger complex (CPC) within the phase separated chromatin body, known to regulate chromosome segregation in mitosis. The CPC phase separated compartment comprises three regulatory and targeting subunits, INCENP, Survivin, and Borealin, forming a three-helix bundle hetero-trimer. By measuring changes in the polypeptide backbone dynamics of this trimeric INCENP/Survivin/Borealin complex, in the liquid-liquid de-mixed state in comparison to its soluble state, using HXMS measurements, the paper puts forward high-resolution structural details of the phase separated CPC. Using a step-wise mutagenesis approach in conjunction with the information from HXMS measurements and previous crystallographic data, this work also identifies distinct regions/interfaces within this complex harboring crucial salt bridges, which directly contribute toward the liquid-liquid demixing of the CPC. Comments: 1) "The three non-catalytic subunits of the CPC (INCENP1-58, Borealin, and Survivin) form soluble homotrimers that have a propensity to undergo liquid-liquid phase separation.8 " Do the authors mean the hetero-trimeric CPC?

Yes, we meant heterotrimers. It is now corrected.

2) For better clarity, the authors can indicate the residue numbers of each of the components INCENP, Survivin, and Borealin in the CPC trimeric helix-bundle crystallographic structure in Fig 1.

These are included on the revised Figure 1A.

3) "In the condition we identified, 90% +/- 5% of the ISB protein was found within the rapidly sedimenting droplet population (Fig. 1C)." The authors should include the time-point corresponding to the gel shown in Fig 1C.

This information is now directly labeled in Fig. 1C.

4) Prior to the HXMS experiments on the phase-separated ISB protein complex, were the samples subjected to sedimentation to separate the dispersed from the condensed droplet phase? Since several time points after formation of phase-separated ISB complex have been characterized to compare and contrast between the dispersed and the droplet phase, the authors can consider performing a time-dependent sedimentation assay to ascertain the fraction of the ISB complex in the droplet phase.

The HXMS experiments were not performed on sedimented samples, so this complication in our HX workflow is not necessary. We note that the sedimentation that we include in our study (Figs. 1C, 5E, and S6), involves centrifugation for 10 minutes, and that length of time presents a substantial design challenge to our HX experimentation. We considered it at the outset of our study, but, in the end, our study was facilitated by our finding early on that this separation step was unnecessary. Further, we note that we report statistically significant differences at the earliest HX timepoints in the areas prominently protected from HX upon droplet formation (10 and 100 s; see Fig. 1C for an example). Indeed, we do not observe broadening of our HXMS spectra (examples shown for all timepoints, Fig. 2B,F) that would be expected if there were a large degree of mixed states (i.e. a large population of molecules in the free protein state and a large population of molecules in the droplet state) each having different HXMS rates. One can imagine that this sort of envelope broadening behavior (“EX1-like”) could be observed in other samples where there are multiple substantially populated states of a protein present at a particular timepoint, but this is not what we observe in the experiments we performed in this study.

5) "At the 100 s timepoint, the most prominent differences between the soluble and droplet state were located within the three-helix bundle of the ISB, with long stretches in two subunits (INCENP and Borealin) and a small region at the N-terminal portion of the impacted a-helix in Survivin (Fig. 1F)" According to Fig 1F, at the 100 s time-point, there is also another small region in Survivin (approximately residues 12-20) that exhibits slower exchange rates in the droplet state. Can the authors comment on whether this region undergoes any conformational change or if it exhibits homotypic interactions retarding the hydrogen/deuterium exchange rates in the droplet phase?

Our general approach in the Black lab over the past decade-plus of HXMS has been to restrict our conclusions whenever practical to do so to the consensus behavior. This permits multiple partially overlapping peptides to be used to generate confidence in the changes that drive our conclusions. The reviewer carefully recognizes the behavior of a single peptide (in 2 different charge states) that might have actual changes relative to some of the longer peptides that it partially overlaps with, and smaller changes can yield larger percentage changes on small peptides. We have chosen to not include this single peptide in the text describing our main conclusions from the work to be consistent with our longstanding strategy for rigorous interpretation of HXMS data. Our conclusion is that this region of not substantially changed upon droplet formation.

6) The authors mention that: "By the latest timepoint, 3000 s, there was some diminution in the number of droplets which may indicate the start of a transition of the droplets to a more solid state (i.e., gel-like)." As a result of this time points beyond 3000 s have not been used for comparing Hydrogen/Deuterium exchange rates in the condensed droplet phase with the soluble state. Can the authors comment on what happens to the nature of these specific interactions between the components of the CPC in the 'gel-like state'? A combination of both non-specific weak interactions as well as strong site-specific interactions between macromolecular components has been widely known to contribute towards the formation of several phase-separated compartments. It will be interesting to know the perspective of the authors on what sort of interactions get populated within these compartments to give rise to a more solid gel-like state. At this later time points, do the droplets exhibit reversibility under higher ionic strength conditions? Do the authors have some data to show how the material property of these droplets evolve as a function of time?

We offered the idea of a transition to a more solid state to the reader because it was a reasonable conclusion, although challenging to prove (something the Stukenberg lab is actively working on, though, see our response to point #9, below). The vast majority of our conclusions in the paper, and essentially all of what we emphasize are the important ones, are based on earlier timepoints where this is not an issue. Thus, we find an extended study of the late-developing features in our droplets something more appropriate for separate studies outside the scope of the current one.

7) "Examination of the entire time course shows that during intermediate levels of HX (i.e., between 100-1000 s), this region takes about three times as long to undergo the same amount of exchange when the ISB is in the droplet state relative to when it's in the free protein state (Figs. 2B, C and Supplemental Fig. 2). Upon droplet formation, HX protection within Borealin is primarily located in the interacting a-helix and is less pronounced at any given peptide when compared to INCENP peptides (Fig. 2E). Nonetheless, similar to INCENP peptides, it still takes about twice as long to achieve the same level of deuteration for this region of Borealin in the droplet state as compared to the free state." How do the hydrogen/deuterium exchange rates and extent of deuteration in the N-terminal part (residues 98-142) of the Survivin polypeptide chain, constituting the three-helix bundle core, evolve as a function of time? Also, how do the exchange rates for peptides in this region compare with those of the other protein subunits Borealin and INCENP and what inference can be drawn from these differences?

The peptides from a.a. 98-142 of Survivin exhibit HX protection through the timecourse (and before and after droplet formation) consistent with a folded a-helix (and comparable to the overall HX behavior of the other helices in the 3-helix bundle of the ISB)(Fig. S2). There is subtly slower HX in the droplet state for this region at later timepoints for this portion of Survivin (Fig. S4), and this is explicitly highlighted in the Results section on pg. 6.

8) The authors mention that mutating either all the glutamate residues or combinations of these residues on the acidic patch on the INCENP subunit, to positively charged residues, causes a decrease in the propensity of phase separation, as formation of salt bridges with Borealin subunit from adjacent hetero-trimeric complexes appears to be the major driving force for phase separation. Can the authors elaborate on how the reduction in the phase separation propensity of these salt-bridge inhibiting mutants might be directly affecting the subsequent localization of the CPC to the inner centromeres? Can the authors supplement their existing in vitro data with further in vivo characterization of CPC recruitment or localization to the centromeres, for each of the constructs exhibiting reduced propensity of phase separation?

As we state in the introduction, the recruitment to centromeres requires established ‘conventional’ targeting via the specific histone marks to which we refer. We also cite the correlations demonstrated between prior mutations in Borealin (impacting aa 139-160) that both disrupt phase separation in vitro and reduce CPC levels at the centromere. In our revision, we have added what we feel are the most critical cell-based experiments to relate to our HX studies in the new Fig. 6. We are preparing for future studies to study mutants arising from our HX studies, and our plans are to pursue gene replacement approaches that will rigorously test the impact on the mitotic function of the CPC. In the process of these future studies, the impact on localization will be measured, too. As others in the field are investigating the correlations between observations made with purified components and those made in the cell, and where there are nuances at play in how the actual experiments are conducted, we are certain our cell-based studies will extend far beyond the timeframe appropriate for our HX-focused study. Rigorous cell-based studies of mitotic functions are what is needed, however, and we have made our plans with that in mind.

9) It might be really interesting for the authors to look at the recent preprint from Hedtfeld et al. 2023 Molecular Cell, (https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4472737). In this preprint they have recombinantly purified a stoichiometric trimer (referred to as CPC-TARGWT) comprising full length survivin, borealin, and a 1-350 residue fragment of INCENP (instead of 1-58 used in this study) and have tried to assess if any correlation exists between the in-vitro phase behaviour of CPC-TARGWT mutants and their corresponding recruitment to the inner centromere, to form a phase separated compartment. Targeting residues in the BIR domain of Survivin involved in interactions with the N-terminus of the Histone H3, Shugosin 1 or in the recognition of H3T3phos, and substituting them with Alanine or completely deleting C-terminal domain of Borealin (a region implicated in CPC dimerization and centromere recruitment), was found to result in poor centromere localization, although the in vitro phase separation properties of these constructs were found to be indistinguishable, suggesting no evident correlation between the two phenomena. Thus it might be a useful piece of data to correlate the phase separation propensities of the ISB complex variants used in this current study with the extents of their in vivo recruitment to the inner centromere. This maybe beyond the scope of the paper, but it would be good to comment on this.

For the correlation studies, please refer to our response to point #8, above. From our reading of the June 2023 preprint that the reviewer mentions, the main concern raised by the authors is questioning whether the region first identified in the Trivedi et al paper in Borealin (aa 139-160) has a role in phase separation. As the reviewer noted, Hedtfeld et al report using a complex that includes more of the INCENP protein than used in the Trivedi et al study, complicating the direct comparison between studies. Using the data in figure 5E of the Hedtfeld et al preprint, the authors suggest that the condensate formation of their version of the Borealin mutant D139-160 in vitro complex has similar phase separation properties as the wild type. However, we note that in our inspection of these data we see numerous differences. The mutant forms rounder, and larger condensates than WT and have reduced concentration of protein (less bright intensity). Finally only the WT protein has a “grape bunch” morphology. We note that unpublished data in the Stukenberg lab show these same differences can represent a defect in liquid demixing properties of a version of the purified CPC. While it is intuitive that larger condensates represent more phase separation, the unpublished data mentioned above suggests the opposite is true for the CPC. In particular, the data from the Stukenberg lab suggest the size of a droplet is mostly governed by the amount of droplet fusion in the first minutes after dilution and thus is limited by relatively rapid hardening of the complex. We note that in the course of discussions with the corresponding author of the preprint mentioned by the reviewer we did apprise them of the unpublished observations mentioned, above, in case they saw fit to include in their ongoing studies what would seem to be critical measurements (e.g. measuring circularity, droplet size, droplet intensity, and FRAP) to assess our suspicion that their construct contains a portion of INCENP that can accelerate condensate formation. If true, the Hedtfeld et al data are fully consistent with the Borealin mutant D139-160 having a significant condensate formation potential than the WT protein.

10[A]) "Our data also provide an important clue about the previously identified region on Borealin that is required for liquid-demixing in vitro and proper CPC assembly in cells 8. Specifically, our data (Fig. 1F, Supplementary Figs. 2, 4A) suggest this region of Borealin adopts secondary structure that undergoes additional HX protection in the liquid-liquid demixed state" This data fits perfectly with previous studies from Trivedi et al. (2019), which states that deletion of the Borealin 139-160 fragment obliterates its phase separation in vitro and also reduces the accumulation of CPC at the centromere. On the contrary, in the recent preprint from Hedtfeld et al. 2023 Molecular Cell, they have shown that the phase separation behaviour of their reconstituted CPC-TARGWT harboring the Borealin 139-160 deletion mutant was found to be indistinguishable from the WT. Can the authors comment on what might be the reason for this difference? Is it possible that this central Borealin region is involved in interactions with the additional fragment of INCENP subunit used in the helical bundle reconstitution, or with other centromere component proteins, whereby the deletion of region is causing inefficient recruitment to the inner centromere? This can be elaborated in the discussion section of the manuscript.

This is discussed in the response to #9, above. Through this format (the Review Commons procedure for public posting of author responses before submission of the study to a journal), our comments herein will be made public for those with the most interest in comparing our data to what is has been posted on preprint servers. We think that is the most appropriate for now, with more to surely come when the aforementioned results from the Stukenberg lab are posted/published and, hopefully when there is more information about the nature of the droplets reported in the Hedtfeld et al., study.

10 [B]) It is also well known that in addition to these electrostatic interactions, the core of the ISB helical bundle is formed by an extensive network of hydrophobic interactions. Have the authors ever looked into how perturbing any of these intra-trimeric complex hydrophobic interactions affect their ability to phase separate and perform their subsequent function?

We think there is some confusion, here. The electrostatics we focus on are between heterotrimers rather than within them. We certainly would predict that disrupting the hydrophobic surface that generates a stable heterotrimer would, in turn, disrupt individual heterotrimers. Our study assumes a stable heterotrimer as a starting point, so we view this type of perturbation as unrelated to our conclusions.

11) The phase separated CPC compartment is known to enrich several other inner centromere proteins such as the Histone H3, Sgo1, the histone H3T3phos, among others. Have the authors tried to increase the complexity of the reconstituted CPC scaffold by incorporating more components to look into whether that changes any of the interaction interfaces between the ISB trimeric complexes within the condensed phase? Can this CPC compartment be reconstituted using a bottom-up approach?

We are glad that our studies with a reductionist biochemical reconstitution approach have inspired the questions that require increased complexity. They are now warranted based on the advance we have made in the present study, and hopefully will form the basis for future, separate studies.

Overall, this paper brings forward a useful technique to probe the conformational landscape of proteins in the condensed droplet phase and compare it with its dispersed phase. This paper serves as an interesting read showing how specific salt-bridge interactions between multiple stoichiometric protein complexes can be the driving force for phase separation.

Reviewer #2 (Significance (Required)):

In this manuscript, using the technique of hydrogen/deuterium-exchange mass spectrometry (HXMS), the authors have tried to gain insights into the structure of the chromosomal passenger complex (CPC) within the phase separated chromatin body, known to regulate chromosome segregation in mitosis. The CPC phase separated compartment comprises three regulatory and targeting subunits, INCENP, Survivin, and Borealin, forming a three-helix bundle hetero-trimer. By measuring changes in the polypeptide backbone dynamics of this trimeric INCENP/Survivin/Borealin complex, in the liquid-liquid de-mixed state in comparison to its soluble state, using HXMS measurements, the paper puts forward high-resolution structural details of the phase separated CPC. Using a step-wise mutagenesis approach in conjunction with the information from HXMS measurements and previous crystallographic data, this work also identifies distinct regions/interfaces within this complex harboring crucial salt bridges, which directly contribute toward the liquid-liquid demixing of the CPC.

Overall, this paper brings forward a useful technique to probe the conformational landscape of proteins in the condensed droplet phase and compare it with its dispersed phase. This paper serves as an interesting read showing how specific salt-bridge interactions between multiple stoichiometric protein complexes can be the driving force for phase separation

We thank the reviewer for the positive comments on the significance of our 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 #2

Evidence, reproducibility and clarity

Structural Basis for the Phase Separation of the Chromosome Passenger Complex Nikaela W. Bryan, Ewa Niedzialkowska, Leland Mayne, P. Todd Stukenberg, and Ben E. Black# Reviewer Comments Manuscript Number: RC-2023-02017

In this manuscript, using the technique of hydrogen/deuterium-exchange mass spectrometry (HXMS), the authors have tried to gain insights into the structure of the chromosomal passenger complex (CPC) within the phase separated chromatin body, known to regulate chromosome segregation in mitosis. The CPC phase separated compartment comprises three regulatory and targeting subunits, INCENP, Survivin, and Borealin, forming a three-helix bundle hetero-trimer. By measuring changes in the polypeptide backbone dynamics of this trimeric INCENP/Survivin/Borealin complex, in the liquid-liquid de-mixed state in comparison to its soluble state, using HXMS measurements, the paper puts forward high-resolution structural details of the phase separated CPC. Using a step-wise mutagenesis approach in conjunction with the information from HXMS measurements and previous crystallographic data, this work also identifies distinct regions/interfaces within this complex harboring crucial salt bridges, which directly contribute toward the liquid-liquid demixing of the CPC.

Comments: 1. "The three non-catalytic subunits of the CPC (INCENP1-58, Borealin, and Survivin) form soluble homotrimers that have a propensity to undergo liquid-liquid phase separation.8 " Do the authors mean the hetero-trimeric CPC? 2. For better clarity, the authors can indicate the residue numbers of each of the components INCENP, Survivin, and Borealin in the CPC trimeric helix-bundle crystallographic structure in Fig 1. 3. "In the condition we identified, 90% +/- 5% of the ISB protein was found within the rapidly sedimenting droplet population (Fig. 1C)." The authors should include the time-point corresponding to the gel shown in Fig 1C. 4. Prior to the HXMS experiments on the phase-separated ISB protein complex, were the samples subjected to sedimentation to separate the dispersed from the condensed droplet phase? Since several time points after formation of phase-separated ISB complex have been characterized to compare and contrast between the dispersed and the droplet phase, the authors can consider performing a time-dependent sedimentation assay to ascertain the fraction of the ISB complex in the droplet phase. 5. "At the 100 s timepoint, the most prominent differences between the soluble and droplet state were located within the three-helix bundle of the ISB, with long stretches in two subunits (INCENP and Borealin) and a small region at the N-terminal portion of the impacted a-helix in Survivin (Fig. 1F)" According to Fig 1F, at the 100 s time-point, there is also another small region in Survivin (approximately residues 12-20) that exhibits slower exchange rates in the droplet state. Can the authors comment on whether this region undergoes any conformational change or if it exhibits homotypic interactions retarding the hydrogen/deuterium exchange rates in the droplet phase? 6. The authors mention that: "By the latest timepoint, 3000 s, there was some diminution in the number of droplets which may indicate the start of a transition of the droplets to a more solid state (i.e., gel-like)." As a result of this time points beyond 3000 s have not been used for comparing Hydrogen/Deuterium exchange rates in the condensed droplet phase with the soluble state. Can the authors comment on what happens to the nature of these specific interactions between the components of the CPC in the 'gel-like state'? A combination of both non-specific weak interactions as well as strong site-specific interactions between macromolecular components has been widely known to contribute towards the formation of several phase-separated compartments. It will be interesting to know the perspective of the authors on what sort of interactions get populated within these compartments to give rise to a more solid gel-like state. At this later time points, do the droplets exhibit reversibility under higher ionic strength conditions? Do the authors have some data to show how the material property of these droplets evolve as a function of time? 7. "Examination of the entire time course shows that during intermediate levels of HX (i.e., between 100-1000 s), this region takes about three times as long to undergo the same amount of exchange when the ISB is in the droplet state relative to when it's in the free protein state (Figs. 2B, C and Supplemental Fig. 2). Upon droplet formation, HX protection within Borealin is primarily located in the interacting a-helix and is less pronounced at any given peptide when compared to INCENP peptides (Fig. 2E). Nonetheless, similar to INCENP peptides, it still takes about twice as long to achieve the same level of deuteration for this region of Borealin in the droplet state as compared to the free state." How do the hydrogen/deuterium exchange rates and extent of deuteration in the N-terminal part (residues 98-142) of the Survivin polypeptide chain, constituting the three-helix bundle core, evolve as a function of time? Also, how do the exchange rates for peptides in this region compare with those of the other protein subunits Borealin and INCENP and what inference can be drawn from these differences? 8. The authors mention that mutating either all the glutamate residues or combinations of these residues on the acidic patch on the INCENP subunit, to positively charged residues, causes a decrease in the propensity of phase separation, as formation of salt bridges with Borealin subunit from adjacent hetero-trimeric complexes appears to be the major driving force for phase separation. Can the authors elaborate on how the reduction in the phase separation propensity of these salt-bridge inhibiting mutants might be directly affecting the subsequent localization of the CPC to the inner centromeres? Can the authors supplement their existing in vitro data with further in vivo characterization of CPC recruitment or localization to the centromeres, for each of the constructs exhibiting reduced propensity of phase separation? 9. It might be really interesting for the authors to look at the recent preprint from Hedtfeld et al. 2023 Molecular Cell, (https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4472737). In this preprint they have recombinantly purified a stoichiometric trimer (referred to as CPC-TARGWT) comprising full length survivin, borealin, and a 1-350 residue fragment of INCENP (instead of 1-58 used in this study) and have tried to assess if any correlation exists between the in-vitro phase behaviour of CPC-TARGWT mutants and their corresponding recruitment to the inner centromere, to form a phase separated compartment. Targeting residues in the BIR domain of Survivin involved in interactions with the N-terminus of the Histone H3, Shugosin 1 or in the recognition of H3T3phos, and substituting them with Alanine or completely deleting C-terminal domain of Borealin (a region implicated in CPC dimerization and centromere recruitment), was found to result in poor centromere localization, although the in vitro phase separation properties of these constructs were found to be indistinguishable, suggesting no evident correlation between the two phenomena. Thus it might be a useful piece of data to correlate the phase separation propensities of the ISB complex variants used in this current study with the extents of their in vivo recruitment to the inner centromere. This maybe beyond the scope of the paper, but it would be good to comment on this. 10. "Our data also provide an important clue about the previously identified region on Borealin that is required for liquid-demixing in vitro and proper CPC assembly in cells 8. Specifically, our data (Fig. 1F, Supplementary Figs. 2, 4A) suggest this region of Borealin adopts secondary structure that undergoes additional HX protection in the liquid-liquid demixed state" This data fits perfectly with previous studies from Trivedi et al. (2019), which states that deletion of the Borealin 139-160 fragment obliterates its phase separation in vitro and also reduces the accumulation of CPC at the centromere. On the contrary, in the recent preprint from Hedtfeld et al. 2023 Molecular Cell, they have shown that the phase separation behaviour of their reconstituted CPC-TARGWT harboring the Borealin 139-160 deletion mutant was found to be indistinguishable from the WT. Can the authors comment on what might be the reason for this difference? Is it possible that this central Borealin region is involved in interactions with the additional fragment of INCENP subunit used in the helical bundle reconstitution, or with other centromere component proteins, whereby the deletion of region is causing inefficient recruitment to the inner centromere? This can be elaborated in the discussion section of the manuscript. 10. It is also well known that in addition to these electrostatic interactions, the core of the ISB helical bundle is formed by an extensive network of hydrophobic interactions. Have the authors ever looked into how perturbing any of these intra-trimeric complex hydrophobic interactions affect their ability to phase separate and perform their subsequent function? 11. The phase separated CPC compartment is known to enrich several other inner centromere proteins such as the Histone H3, Sgo1, the histone H3T3phos, among others. Have the authors tried to increase the complexity of the reconstituted CPC scaffold by incorporating more components to look into whether that changes any of the interaction interfaces between the ISB trimeric complexes within the condensed phase? Can this CPC compartment be reconstituted using a bottom-up approach?

Overall, this paper brings forward a useful technique to probe the conformational landscape of proteins in the condensed droplet phase and compare it with its dispersed phase. This paper serves as an interesting read showing how specific salt-bridge interactions between multiple stoichiometric protein complexes can be the driving force for phase separation.

Significance

In this manuscript, using the technique of hydrogen/deuterium-exchange mass spectrometry (HXMS), the authors have tried to gain insights into the structure of the chromosomal passenger complex (CPC) within the phase separated chromatin body, known to regulate chromosome segregation in mitosis. The CPC phase separated compartment comprises three regulatory and targeting subunits, INCENP, Survivin, and Borealin, forming a three-helix bundle hetero-trimer. By measuring changes in the polypeptide backbone dynamics of this trimeric INCENP/Survivin/Borealin complex, in the liquid-liquid de-mixed state in comparison to its soluble state, using HXMS measurements, the paper puts forward high-resolution structural details of the phase separated CPC. Using a step-wise mutagenesis approach in conjunction with the information from HXMS measurements and previous crystallographic data, this work also identifies distinct regions/interfaces within this complex harboring crucial salt bridges, which directly contribute toward the liquid-liquid demixing of the CPC.

Overall, this paper brings forward a useful technique to probe the conformational landscape of proteins in the condensed droplet phase and compare it with its dispersed phase. This paper serves as an interesting read showing how specific salt-bridge interactions between multiple stoichiometric protein complexes can be the driving force for phase separation

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 Bryan et al., describes the use of Hydrogen/Deuterium-exchange Mass Spectrometry (HXMS) as a powerful tool to identify key amino acid residues and associated interactions driving liquid-liquid demixing. They have particularly focused on the Chromosomal Passenger Complex (CPC), an important regulator of chromosome segregation, which has recently been shown to undergo liquid-liquid demixing in vitro. Their work presented here allowed them to identify a few key electrostatic interactions as molecular determinants driving the liquid-liquid demixing of the CPC. Their work also shows that crystal packing information of protein molecules, where available, can provide valuable insight into likely factors driving liquid-liquid demixing.

Major comments:

A previous study by Trivedi et al., NCB 2019 identified an unstructured region in Borealin (aa residues 139-160) as the main region driving the phase separation of CPC. Interestingly, this region only shows a moderate reduction in HX upon liquid-liquid demixing. But no experiments or discussions related to this observation are presented in the current version of the manuscript.

In the absence of cellular data on if and how these mutations (within the triple-helical bundle region) affect CPC's ability to phase separate in cells, the implication of this work is very limited - One can't say for sure these are interactions driving phase separation of CPC in a cellular environment.

In the absence of any cellular data with the mutants described here, much of the discussion on the possible roles of CPC phase separation in cells does not appear relevant to this manuscript. I would suggest that the authors focus mainly on highlighting the power of using HXMS as a tool to characterise the molecular determinants of liquid-liquid demixing at a relatively high resolution.

Minor comments:

The authors should ensure that the introduction cites relevant literature thoroughly. For example, where the potential role of Borealin residues 139-160 in conferring phase separation properties to the CPC is mentioned, the authors failed to cite Abad et al., 2019, which showed the contribution of the same Borealin region in conferring nucleosome binding ability to the CPC.

Significance

This is a highly relevant and significant work, particularly considering the rapidly growing list of examples for Phase separation of proteins/protein assemblies and their potential biological roles (in spite of ongoing debates in the field about the cellular relevance of several phase separation claims). The data presented in this manuscript are solid and convincingly establish HXMS as a useful tool to characterise molecular interactions driving liquid-liquid demixing. Considering its applicability to characterise wide-ranging protein assemblies implicated in phase separation, this work will be of interest to a broad readership.

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)):

Centrioles are small cylindrical structures with roles in cell division, motility, and signaling. Typically, centrioles are highly stable structures which can persist for many cell generations. However, in some cells, such as the female germ line of many species, centrioles are programmed for elimination. This process is essential for maintaining centriole number from one generation to the next in sexually reproducing organisms, yet in nearly all species the molecular mechanisms underlying how centrioles are eliminated is unknown. The current study utilizes the nematode C. elegans to explore how centriole architecture changes during the elimination program in the female germ line. Using a suite of light microscopy techniques, the authors provide a stunning visual perspective of how centrioles are disassembled during oogenesis and show that removal of the central tube component SAS-1, a key regulator of centriole stability, is an early event in elimination. I have no major objections to the work and enthusiastically endorse its publication with the following minor revisions.

Page 9 line 200: In the pcmd-1 mutant, the authors state that centriolar foci devoid of nuclei are present in rachis, but they do not mention in the text that there are also nuclei that lack centriole foci in early pachytene. This is mentioned in the figure legend, but I felt it was important enough to mention in the text.

As per the reviewer’s suggestion, we will provide this information in the main text as well.

Page 9 line 211. The authors found that in the absence of dynein heavy or light chain that centrioles remain associated with the nuclear envelope (rather than moving to the periphery). To me this was striking as dynein depletion in the embryo results in the opposite phenotype with centrioles losing attachment to the nuclear envelope and moving to the cell periphery (Gonczy et al. 1999 JCB 147:135). It might be worth pointing this out somewhere in the manuscript and speculating about the reasons for this difference.

We will expand the Discussion section to better explain the difference of dynein’s involvement in the oocyte versus the embryo.

Page 11 line 277: The authors state that elimination timing is not affected by the loss of SPD-5. This is a small but important point. It really is the absence of PCMD-1 and not SPD-5, as SPD-5 is still present in the cell. An alternative would be to say "in the absence of PCM" or "in absence of a pericentriolar accumulation of SAS-5".

Fully agreed, we will modify the text accordingly.

Figure 4D: Why does loss of PCMD-1 result in a delay in oocyte maturation as judged by RME-2 accumulation? This is not mentioned in the paper. Is this a general response to a loss of PCM or is this specific to a loss of PCMD-1?

We realize that we were not sufficiently clear in explaining that RME-2 accumulation reflects the maturation state of oocytes. In the revised manuscript, we will clarify this point further and mention that a mild developmental delay (such as in pcmd-1(t3421ts) mutant animals) can impact the number of maturing oocytes present in the proximal gonad, and thereby lead to a slight shift in RME‑2::GFP distribution. See also related minor comment 2 of reviewer 2, and major comment 1 of reviewer 3.

Figure 7 E and F. The authors measure the tubulin and SAS-4 intensity in wild-type and sas-1(t1521) embryos and conclude that microtubules and SAS-4 signals decay faster in the sas-1 mutant than in the control. To me, this is convinceingly the case with microtubules in panel E but I am not so sure this is the case with SAS-4 as shown in panel F. The differences in SAS-4 levels are much smaller between mutant and control. Could the authors provide statistical analysis to show how significant the differences are?

We will provide the requested statistical analysis (which indeed shows significance).

Page 15 line 363. I think this sentence should be reworded to: "Finally, we demonstrate that the central tube protein SAS-1 is the first of the factors analyzed here to leave centrioles..."

In response to this suggestion and to the related comment of reviewer 2 (see below), we will rephrase this sentence to read “among the centriolar components analyzed to date, SAS-1 is the first to depart”.

Reviewer #1 (Significance (Required)):

The work contained in this manuscript represents a fundemental step forward in understanding the process of centriole elimination. The authors have carefully described the stepwise disassembly of the centriole including changes in the architechure during oogenesis. They have identified loss of the centriole stability factor SAS-1, as an early event in the elimination program and have found that in a sas-1 mutant, the centriole disassembles prematurely. They have also shown that loss of SAS-1 is followed by expansion of the centriole and ultimately loss of structural integrity. This work should be of interest to a broad range of scientists including those interested in centrosome dynamics, germ line development, and more generally cell biologists.

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

Summary In this manuscript Pierron et al. explore the mechanisms of centriole elimination during oogenesis in C. elegans. Centriole elimination is a common feature of oogenesis in many species, but it is relatively poorly understood and understudied. Here, the authors characterise the kinetics with which several key centriole and centrosome proteins are lost during this process in living worms, and they correlate this with an EM and expansion microscopy (U-Ex-STED) analyses of fixed tissues. They conclude that centriole elimination begins with the loss of SAS-1 from the central region of the centrioles, which correlates with the widening of the structure and the loss of the centriole MTs. A remnant structure containing several core centriole proteins remains, however, and this often ultimately detaches from the nuclear envelope and moves towards the plasma membrane in a MT-motor-dependent fashion before it dissipates (although detachment from the nucleus does not seem to be required for the eventual elimination of this residual structure). Intriguingly, centriole loss in this system does not appear to require the down-regulation of PLK activity, which is in contrast to the situation in Drosophila oogenesis.

The manuscript is generally well written and the data is of a high quality and is logically and clearly presented. Although the ultimate mechanisms regulating centriole elimination remain obscure (i.e. what triggers the loss of SAS-1, and how is this regulated?), the data presented here will be of significant interest to the centriole/centrosome field and I am supportive of publication. I have a few points that the authors should consider prior to publication.

Major comment:

In the EM shown in Figure 5F the authors claim that the central tube of the centriole is disrupted, but the other elements (inner tube, MTs and paddlewheel) are not. I don't think this is as clear cut as the authors claim-at least from comparing the images of the one normal centriole (5E) and one centriole that is starting to be eliminated (5F). It seems much harder to distinguish the MTs and the inner tube in the image in 5F. Perhaps this is obvious to the authors as they have compared many more images, but I think they need to find some way of showing this more convincingly (a montage of multiple centrioles)?

We understand that Figure 5F alone may have left the reviewer wondering whether the central tube is truly the first element to be disrupted during centriole elimination. We plan on strengthening this point by providing additional EM images as a Supplemental Figure.

This same issue is compounded in Figure 6D where, using a different technique (U-Ex-STED), the authors claim that the centriolar distribution of SAS-1 is gradually disrupted as centriole elimination proceeds. It does look like the amount of SAS-1 has decreased from early prophase to late pachytene, but the central tube it stains doesn't look particularly disrupted and, if anything, the MTs look more disrupted (and also possibly of lower intensity, perhaps explaining why the ratio of SAS-1/tubulin doesn't change very much over these stages, as shown in Figure 6G).

As the reviewer correctly noticed, there is some variability in central tube removal during oogenesis. In some cases, such as in the centriole on the right of the late pachytene panel in Fig. 6D, SAS-1 signal intensity diminishes uniformly, without apparent holes in the central tube. By contrast, in other cases, such as in the centriole on the left of the late pachytene panel, SAS-1 signal intensity diminution is accompanied by a loss of central tube continuity. We will clarify the writing and qualify our findings on this important point in the revised manuscript.

These points are important, as throughout the manuscript the authors assume it as a fact that SAS-1 leaves the centriole early (which is clear), and that this leads to the specific loss of the central tube (which, at least on the basis of this data, is not so clear).

As mentioned above, we will make certain that the results linking SAS-1 departure and central tube loss are explained in a clear and balanced manner in the revised manuscript.

Minor comments:

1. The authors state that the kinetics of GFP-SAS-7 or SAS-4 loss were not altered in pcmd-1 mutants (Figure 4A-C; Figure S3E,F). This doesn't look correct to me, as both proteins seem to stay brighter for longer in the mutant embryos (and this is quite easy to see on the quantification graph for SAS-7 in Figure 4C). It looks similar for SAS-4 from the pictures shown in Figure S3E,F, although this data is not quantified (and is there any reason why this data is not quantified?).

As mentioned in response to reviewers 1 and 3, we will mention in the revised manuscript that a mild developmental delay can impact the number of maturing oocytes present in the proximal gonad, thereby leading to this slight shift in GFP::SAS-7 and GFP::SAS-4 persistence.

1. The authors state that they demonstrate that SAS-1 is the first component to leave the disassembling centrioles. I would rephrase as they can't know this for sure (i.e. there could be some untested component that leaves earlier).

In response to this suggestion and to the related comment of reviewer 1 (see above), we will rephrase this sentence to read “among the centriolar components analyzed to date, SAS-1 is the first to depart”.

In the latter part of the Discussion the authors state that SAS-1 is critical for centriole elimination. I would rephrase, as this seems to suggest it is required for centriole elimination, which is not the case. It might also be worth discussing that the elimination machinery clearly seems to target SAS-1 early on, but we don't yet know what this machinery is or how it is regulated.

We thank the reviewer for raising this important point, which we will implement in the Discussion accordingly.

Reviewer #2 (Significance (Required)):

The manuscript is generally well written and the data is of a high quality and is logically and clearly presented. Although the ultimate mechanisms regulating centriole elimination remain obscure (i.e. what triggers the loss of SAS-1, and how is this regulated?), the data presented here will be of significant interest to the centriole/centrosome field and I am supportive of publication. I have a few points that the authors should consider prior to publication.

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

Pierron et al. uses C. elegans oocytes to tackle a fundamental, yet heavily under-studied question in developmental biology: how are centrioles are eliminated during gamete formation/maturation? The paper's main conclusion is that SAS-1 (a key protein that make up the central tube in C. elegans centrioles) plays a critical part to regulate the timing of centriole elimination. I congratulate the authors on all the experiments related to SAS-1 part of their story, as they are done meticulously and in unprecedented detail (particularly all the fascinating EM and expansion microscopy data!).

The paper also concludes that the Polo-like kinase family does not have a central role in this process, in stark contrast to a previous report demonstrating their importance for centriole elimination in Drosophila oogenesis (Pimenta-Marques et al. 2016 Science). Unfortunately, I am less convinced about this part of the paper, and half of my major comments below relate to the experiments/analyses in this regard. I was similarly not very enthusiastic about a part of story that I didn't find very relevant to the main point of the paper: half of the centrioles detach from the nucleus and translocate to plasma membrane prior to their elimination. I find the observations here quite epiphenomenal and lacking a direct/mechanistic relevance to either the PLK or SAS-1 part of the story. In my view, the authors should consider taking this part out.

Regarding this last suggestion: we think that even if the movement of centrioles remnant is not essential for final removal, an account of this process provides important information about cellular dynamics during oocyte maturation. We note also that the two other reviewers did not raise this point, but leave the final decision to the editor.

Overall, the piece is well written and organized, however it suffers from several shortcomings that preclude it from publication in its current form. I list my criticisms and suggestions below.

Major comments:

1. The authors state firmly at several places in the text that PCM components do not contribute to the timing of centriole elimination (e.g., lines 420-421), particularly given their experiments with Polo kinase paralogs. In my view, the data speaks otherwise. The centriole elimination process appears strikingly premature when SPD5__1__ (another PCM component) is overexpressed with the fluorescent transgene (Figure 1I). The opposite is also true - when another PCM component, PCMD-1, is knockdown by a temperature sensitive allele, the centriole elimination process is severely delayed 2 (Figure 4C). Even more extremely in the epistatic Polo mutant conditions (Fig. S3B), the centrioles do not appear to be eliminated at all__3__ (though the authors prefer to interpret this result differently in line 260-263, which could be flawed per my second comment below). How do the authors explain all these intriguing results? (underlining and numbering added above to clarify our responses point by point hereafter)

1 > We respectfully disagree, since our quantifications show clearly that the SAS-7 signal disappears with an analogous timing in the line expressing RFP::SPD-5 (Fig. 1J) when compared to the other lines (Fig. 1D, 1F and 1H). The image shown currently for RFP::SPD-5 (Fig. 1I) is somewhat of an outlier compared to the others (Fig. 1C, 1E and 1G), and we will therefore provide a more representative specimen in the revised manuscript to avoid confusion.

2 > As mentioned also in response to reviewers 1 and 2, we realize that we were not sufficiently clear in explaining that RME-2 accumulation reflects the maturation state of oocytes. In the revised manuscript, we will clarify this point and mention that a mild developmental delay (such as in pcmd-1(t3421ts) mutant animals) can impact the number of maturing oocytes present in the proximal gonad, and thereby lead to a slight shift in RME‑2::GFP distribution (as opposed to representing a delay in centriole elimination in pcmd-1(t3421ts) mutant animals).

3 > We used plk-1(or683ts); plk-2(ok1936) double mutants to further test whether there might be premature elimination in this strong reduction-of-function condition compared to RNAi-mediated depletion. Although centriolar foci appear to remain for a longer time, these gonads are extremely disorganized, so that our conclusion regarding PLK-1 and PLK-2 are based primarily on the combined data shown in Fig. 3 and Fig. S3, which do not exhibit premature centriole elimination. We will rectify the writing to clarify these points.

Also, I believe these claims (on the PCM components and their role in centriole elimination) will benefit from more nuanced statements. For instance, although Plk paralogs may not be necessary for the centriole elimination process, some other centrosome components clearly are. Paradoxically, the effects observed here (when disrupting or promoting PCM formation) has the totally opposite effects observed in Pimenta-Marques et al. 2016 Science. The 2016 piece claimed that the loss of PCM renders centrioles more vulnerable to losing their stability (which makes sense). How do the authors interpret their own results (i.e. that a disturbed PCM leads to slower centriole elimination, and vice versa)?

As suggested by the reviewer, we will consider toning down claims regarding the role of PCM components in centriole elimination. Moreover, we will expand the section in the Discussion comparing our results with the published work of Pimenta-Marques et al. in Drosophila. This being written, as mentioned above, our findings do not suggest that removing the PCM (in pcmd-1(t3421ts) mutant animals) alters centriole elimination timing in C. elegans.

I invite the authors to more carefully tread these nuances throughout their manuscript, which otherwise may cast major doubt on their claims.

See point above.

1. When investigating the role of Polo-like kinases, the authors assume that centriole elimination must follow (or correlate with) the dynamics of RME-2 (as a proxy for oocyte maturation). What guarantees that the centriole elimination process has to follow oocyte maturation? As far as I could tell, there is no direct evidence presented in the paper about this point. Do the authors have direct data (or reference to another work) that this trend must hold true at all times? I can readily see several places in the paper where this correlation doesn't appear to hold (e.g., in Fig. 4D the centriole elimination precedes the oocyte maturation under pcmd-1 condition).

   We will provide further data supporting the view that oocyte maturation and centriole elimination are correlated, whereby premature oocyte maturation mutants, such as let-60(ga89ts) and kin-18(ok395), exhibit precocious elimination.

To correctly interpret their results on the epistatic Polo mutants, the authors could examine centriole elimination timing with mutants that can pre-maturely trigger or delay oocyte maturation (and do so without affecting the centriole biology itself).

See above point.

1. Lines 155-159 on the dimness of the SAS-6 signal make me worried about how successfully the transgenes were generated. Could the authors comment on, or perhaps extend in detail in the Methods section, through what assays the transgenes were validated? For example, did the authors try to rescue a SAS-6-/- with a SAS-6::GFP transgene? I would like to see further support for their validities.

We will explicitly explain in the Material and Methods section that the SAS-6::GFP transgene indeed rescues the sas-6 null phenotype.

If the authors can demonstrate the validity of their transgenes more reliably, could they possibly comment on the bunch of seemingly random SAS-6::GFP foci in Fig. 1G?

We will comment on the presence of small SAS-6::GFP foci in the most mature oocytes, which correspond to potential precursors of centriolar elements later assembled in the embryo.

1. Starting from line 204, the authors use the percentage of oocytes with detached centrioles (from the nucleus) as a proxy for movement to plasma membrane. This can be very confounding in my view (due to erroneous detachments etc.). As the authors explicitly state that the detachment is a process followed by a directed movement (with a defined velocity) towards the plasma membrane, this calls for a much better measurement in general. The authors should directly measure how far the centrioles are from the closest plasma membrane region in each condition they are examining (and should do this as a function of the "time progression" in different oocytes as they get closer to fertilization).

As mentioned above, we think that an account of the movement of centriole remnants provides important information about cellular dynamics during oocyte maturation. However, given that this movement is not essential for the elimination of such remnants, it appears that providing additional complex 3D analysis as suggested by the reviewer will not benefit the present manuscript.

Do the authors observe any propensity in sas1(t1521ts) oocytes as to where the centrioles are being degraded more prominently in the cytoplasm (i.e., when attached to the nucleus vs. when near the plasma membrane)? They could perform analyses à la their assessments in Fig. S2 and see whether they can extract some more information about this. In other words, I am wondering whether SAS-1 regulates the centriole elimination process more prominently at near the nucleus or near the plasma membrane.

Centriole elimination occurs during pachytene in sas-1(t1521) mutant animals, when nuclei are packed in the gonad and surrounded by little cytoplasm. Therefore, even if foci were to detach from nuclei at this stage, we would not be able to quantify it with certainty. We will discuss these points in the revised manuscript.

I ask this because the section about "centrioles moving to plasma membrane" appears epiphenomenal and rather random (i.e., the chances of a centriole moving to plasma membrane appears 50-50 under some control conditions - see control RNAi in Fig. 2G for example). Could the authors explore their existing data more closely (like suggested above), to see whether they could find intriguing correlations that tells us a little more about whether the centriole elimination at these two places are achieved differently? Otherwise, I frankly do not think this section contributes significantly to the essence of the story.

We apologize for the confusion our writing seems to have generated. The chances of moving to the plasma membrane are not 50-50. The actual figure is 78.7% (reported as ~80% in the manuscript, line 187), and stems from the live imaging experiments where every travelling event can be monitored. By contrast, the analysis of fixed specimens is an underestimate as it provides only a snapshot of a dynamic process. We will expand the writing in the revised manuscript to clarify this point.

Finally, the statements about a deterministic function for the plasma membrane re-localization should be toned down, because unlike what the authors claim in the paper (that ~80% of the centrioles move to plasma membrane), the control data (in Fig. 2B) clearly demonstrates that this number is more like ~60% (hence close to its chances being 50-50).

Please see response just above.

The paper carefully quantifies most of the data (for which I sincerely congratulate the authors!), however the experiments in Fig. S3 fall short of this. It would be nice if the authors could do the same here for completion.

We will provide quantifications for Fig. S3E and S3F. However, due to the high disorganization of plk-1(or683ts); plk-2(ok1936) gonads, the presence of centriolar foci relative to oocyte position cannot be quantified accurately in this case.

Minor comments:

1. Sentence in lines 110-113 is too long and perturbs the flow. This should be shortened or be broken into better clauses. Perhaps the following way? "Prior analysis of centriole elimination in C. elegans oogenesis uncovered that this process takes place during diplotene..."

The text will be modified accordingly.

What are the orange arrowheads in the figure panels? They are not stated explicitly in the figure legends. My prediction was that they point to regions where centrioles are in another plane (though the overview is depicted from a different slice in the stack). Is this right? Either way, it will be useful to over-guide the reader on these orange arrowheads.

The meaning of the orange arrowheads is explained in lines 520-521.

If I am not wrong, the data/graph in Figures S2G and 2E are essentially the same (i.e., the data are duplicated). I couldn't find any statement in the figure legends indicating this. This should be added.

Apologies about this oversight -the reviewer is correct and we will make a mention of this redundancy in the legend of Fig. S2.

Some may consider the discussion on C2CD3 a little far-fetched, as this protein localizes to the distal end of centrioles (completely unlike SAS-1). Also, unlike the C. elegans centrioles, mammal centrioles do not contain a discernible central tube, casting doubt on the possibility of speculations made in the Discussion section. I suggest to remove out this paragraph, and instead to explicitly state whether the SAS-1 dependent mechanism could be applicable to other species is unclear.

We will nuance these thoughts, further stressing their speculative nature, but intend to maintain them in some form as they provide a potential parallel that will be of interest to the human cell biology community.

Could the authors add in their Discussion section some comment/thought on what the remaining GFP::SAS-7 pool (line 300-302) might possibly be? Curiously, there doesn't seem to be any structure associated with it in their EM tomograms, so it would be helpful to guide the reader further on this interesting finding.

Although we would love to comment on this further, the remaining GFP::SAS-7 foci lack ultrastructural organization and do not exhibit recognizable electron densities. That this is the case will be stated explicitly in the revised manuscript.

Reviewer #3 (Significance (Required)):

General Assessment: This paper's strength is in its rigorous cell biology approaches to tackle a fundamental developmental biology problem. However, some of their conclusions are too firm while not being well-supported by the data, so the paper requires major revision before its publication.

Advance: Discovery of a new molecular player in the centriole elimination process in worm oocytes, which can pave the way for future discoveries of centriole elimination mechanisms in other species. It is not yet clear whether the results will be broadly applicable, as some of the findings presented are in stark contrast to previous studies published on centriole elimination processes in Drosophila oocytes (e.g., Pimenta-Marques et al. 2016 Science). However, as summarized in the above section, these conclusions require further experimental evidence/support.

Audience: Centriole elimination mechanisms are not widely studied, so I am not entirely sure whether this piece will be of immediate interest to the broad cell biology community. It will certainly be of general interest to several groups studying centriole elimination mechanisms, as well as developmental biologists trying to understand the oocyte maturation process.

My expertise: Molecular and cellular mechanisms of cytoplasmic organization in development

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

Pierron et al. uses C. elegans oocytes to tackle a fundamental, yet heavily under-studied question in developmental biology: how are centrioles are eliminated during gamete formation/maturation? The paper's main conclusion is that SAS-1 (a key protein that make up the central tube in C. elegans centrioles) plays a critical part to regulate the timing of centriole elimination. I congratulate the authors on all the experiments related to SAS-1 part of their story, as they are done meticulously and in unprecedented detail (particularly all the fascinating EM and expansion microscopy data!).

The paper also concludes that the Polo-like kinase family does not have a central role in this process, in stark contrast to a previous report demonstrating their importance for centriole elimination in Drosophila oogenesis (Pimenta-Marques et al. 2016 Science). Unfortunately, I am less convinced about this part of the paper, and half of my major comments below relate to the experiments/analyses in this regard. I was similarly not very enthusiastic about a part of story that I didn't find very relevant to the main point of the paper: half of the centrioles detach from the nucleus and translocate to plasma membrane prior to their elimination. I find the observations here quite epiphenomenal and lacking a direct/mechanistic relevance to either the PLK or SAS-1 part of the story. In my view, the authors should consider taking this part out.

Overall, the piece is well written and organized, however it suffers from several shortcomings that preclude it from publication in its current form. I list my criticisms and suggestions below.

Major comments:

1. The authors state firmly at several places in the text that PCM components do not contribute to the timing of centriole elimination (e.g., lines 420-421), particularly given their experiments with Polo kinase paralogs. In my view, the data speaks otherwise. The centriole elimination process appears strikingly premature when SPD5 (another PCM component) is overexpressed with the fluorescent transgene (Figure 1I). The opposite is also true - when another PCM component, PCMD-1, is knockdown by a temperature sensitive allele, the centriole elimination process is severely delayed (Figure 4C). Even more extremely in the epistatic Polo mutant conditions (Fig. S3B), the centrioles do not appear to be eliminated at all (though the authors prefer to interpret this result differently in line 260-263, which could be flawed per my second comment below). How do the authors explain all these intriguing results?

Also, I believe these claims (on the PCM components and their role in centriole elimination) will benefit from more nuanced statements. For instance, although Plk paralogs may not be necessary for the centriole elimination process, some other centrosome components clearly are. Paradoxically, the effects observed here (when disrupting or promoting PCM formation) has the totally opposite effects observed in Pimenta-Marques et al. 2016 Science. The 2016 piece claimed that the loss of PCM renders centrioles more vulnerable to losing their stability (which makes sense). How do the authors interpret their own results (i.e. that a disturbed PCM leads to slower centriole elimination, and vice versa)?

I invite the authors to more carefully tread these nuances throughout their manuscript, which otherwise may cast major doubt on their claims. 2. When investigating the role of Polo-like kinases, the authors assume that centriole elimination must follow (or correlate with) the dynamics of RME-2 (as a proxy for oocyte maturation). What guarantees that the centriole elimination process has to follow oocyte maturation? As far as I could tell, there is no direct evidence presented in the paper about this point. Do the authors have direct data (or reference to another work) that this trend must hold true at all times? I can readily see several places in the paper where this correlation doesn't appear to hold (e.g., in Fig. 4D the centriole elimination precedes the oocyte maturation under pcmd-1 condition).

To correctly interpret their results on the epistatic Polo mutants, the authors could examine centriole elimination timing with mutants that can pre-maturely trigger or delay oocyte maturation (and do so without affecting the centriole biology itself). <br /> 3. Lines 155-159 on the dimness of the SAS-6 signal make me worried about how successfully the transgenes were generated. Could the authors comment on, or perhaps extend in detail in the Methods section, through what assays the transgenes were validated? For example, did the authors try to rescue a SAS-6-/- with a SAS-6::GFP transgene? I would like to see further support for their validities.

If the authors can demonstrate the validity of their transgenes more reliably, could they possibly comment on the bunch of seemingly random SAS-6::GFP foci in Fig. 1G? 4. Starting from line 204, the authors use the percentage of oocytes with detached centrioles (from the nucleus) as a proxy for movement to plasma membrane. This can be very confounding in my view (due to erroneous detachments etc.). As the authors explicitly state that the detachment is a process followed by a directed movement (with a defined velocity) towards the plasma membrane, this calls for a much better measurement in general. The authors should directly measure how far the centrioles are from the closest plasma membrane region in each condition they are examining (and should do this as a function of the "time progression" in different oocytes as they get closer to fertilization).<br /> 5. Do the authors observe any propensity in sas1(t1521ts) oocytes as to where the centrioles are being degraded more prominently in the cytoplasm (i.e., when attached to the nucleus vs. when near the plasma membrane)? They could perform analyses à la their assessments in Fig. S2 and see whether they can extract some more information about this. In other words, I am wondering whether SAS-1 regulates the centriole elimination process more prominently at near the nucleus or near the plasma membrane.

I ask this because the section about "centrioles moving to plasma membrane" appears epiphenomenal and rather random (i.e., the chances of a centriole moving to plasma membrane appears 50-50 under some control conditions - see control RNAi in Fig. 2G for example). Could the authors explore their existing data more closely (like suggested above), to see whether they could find intriguing correlations that tells us a little more about whether the centriole elimination at these two places are achieved differently? Otherwise, I frankly do not think this section contributes significantly to the essence of the story.

Finally, the statements about a deterministic function for the plasma membrane re-localization should be toned down, because unlike what the authors claim in the paper (that ~80% of the centrioles move to plasma membrane), the control data (in Fig. 2B) clearly demonstrates that this number is more like ~60% (hence close to its chances being 50-50). 6. The paper carefully quantifies most of the data (for which I sincerely congratulate the authors!), however the experiments in Fig. S3 fall short of this. It would be nice if the authors could do the same here for completion.

Minor comments:

1. Sentence in lines 110-113 is too long and perturbs the flow. This should be shortened or be broken into better clauses. Perhaps the following way? "Prior analysis of centriole elimination in C. elegans oogenesis uncovered that this process takes place during diplotene..."
2. What are the orange arrowheads in the figure panels? They are not stated explicitly in the figure legends. My prediction was that they point to regions where centrioles are in another plane (though the overview is depicted from a different slice in the stack). Is this right? Either way, it will be useful to over-guide the reader on these orange arrowheads.
3. If I am not wrong, the data/graph in Figures S2G and 2E are essentially the same (i.e., the data are duplicated). I couldn't find any statement in the figure legends indicating this. This should be added.
4. Some may consider the discussion on C2CD3 a little far-fetched, as this protein localizes to the distal end of centrioles (completely unlike SAS-1). Also, unlike the C. elegans centrioles, mammal centrioles do not contain a discernible central tube, casting doubt on the possibility of speculations made in the Discussion section. I suggest to remove out this paragraph, and instead to explicitly state whether the SAS-1 dependent mechanism could be applicable to other species is unclear.
5. Could the authors add in their Discussion section some comment/thought on what the remaining GFP::SAS-7 pool (line 300-302) might possibly be? Curiously, there doesn't seem to be any structure associated with it in their EM tomograms, so it would be helpful to guide the reader further on this interesting finding.

Significance

General Assessment: This paper's strength is in its rigorous cell biology approaches to tackle a fundamental developmental biology problem. However, some of their conclusions are too firm while not being well-supported by the data, so the paper requires major revision before its publication.

Advance: Discovery of a new molecular player in the centriole elimination process in worm oocytes, which can pave the way for future discoveries of centriole elimination mechanisms in other species. It is not yet clear whether the results will be broadly applicable, as some of the findings presented are in stark contrast to previous studies published on centriole elimination processes in Drosophila oocytes (e.g., Pimenta-Marques et al. 2016 Science). However, as summarized in the above section, these conclusions require further experimental evidence/support.

Audience: Centriole elimination mechanisms are not widely studied, so I am not entirely sure whether this piece will be of immediate interest to the broad cell biology community. It will certainly be of general interest to several groups studying centriole elimination mechanisms, as well as developmental biologists trying to understand the oocyte maturation process.

My expertise: Molecular and cellular mechanisms of cytoplasmic organization in development

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

In this manuscript Pierron et al. explore the mechanisms of centriole elimination during oogenesis in C. elegans. Centriole elimination is a common feature of oogenesis in many species, but it is relatively poorly understood and understudied. Here, the authors characterise the kinetics with which several key centriole and centrosome proteins are lost during this process in living worms, and they correlate this with an EM and expansion microscopy (U-Ex-STED) analyses of fixed tissues. They conclude that centriole elimination begins with the loss of SAS-1 from the central region of the centrioles, which correlates with the widening of the structure and the loss of the centriole MTs. A remnant structure containing several core centriole proteins remains, however, and this often ultimately detaches from the nuclear envelope and moves towards the plasma membrane in a MT-motor-dependent fashion before it dissipates (although detachment from the nucleus does not seem to be required for the eventual elimination of this residual structure). Intriguingly, centriole loss in this system does not appear to require the down-regulation of PLK activity, which is in contrast to the situation in Drosophila oogenesis.

The manuscript is generally well written and the data is of a high quality and is logically and clearly presented. Although the ultimate mechanisms regulating centriole elimination remain obscure (i.e. what triggers the loss of SAS-1, and how is this regulated?), the data presented here will be of significant interest to the centriole/centrosome field and I am supportive of publication. I have a few points that the authors should consider prior to publication.

Major comment:

In the EM shown in Figure 5F the authors claim that the central tube of the centriole is disrupted, but the other elements (inner tube, MTs and paddlewheel) are not. I don't think this is as clear cut as the authors claim-at least from comparing the images of the one normal centriole (5E) and one centriole that is starting to be eliminated (5F). It seems much harder to distinguish the MTs and the inner tube in the image in 5F. Perhaps this is obvious to the authors as they have compared many more images, but I think they need to find some way of showing this more convincingly (a montage of multiple centrioles)?

This same issue is compounded in Figure 6D where, using a different technique (U-Ex-STED), the authors claim that the centriolar distribution of SAS-1 is gradually disrupted as centriole elimination proceeds. It does look like the amount of SAS-1 has decreased from early prophase to late pachytene, but the central tube it stains doesn't look particularly disrupted and, if anything, the MTs look more disrupted (and also possibly of lower intensity, perhaps explaining why the ratio of SAS-1/tubulin doesn't change very much over these stages, as shown in Figure 6G).

These points are important, as throughout the manuscript the authors assume it as a fact that SAS-1 leaves the centriole early (which is clear), and that this leads to the specific loss of the central tube (which, at least on the basis of this data, is not so clear).

Minor comments:

1. The authors state that the kinetics of GFP-SAS-7 or SAS-4 loss were not altered in pcmd-1 mutants (Figure 4A-C; Figure S3E,F). This doesn't look correct to me, as both proteins seem to stay brighter for longer in the mutant embryos (and this is quite easy to see on the quantification graph for SAS-7 in Figure 4C). It looks similar for SAS-4 from the pictures shown in Figure S3E,F, although this data is not quantified (and is there any reason why this data is not quantified?).
2. The authors state that they demonstrate that SAS-1 is the first component to leave the disassembling centrioles. I would rephrase as they can't know this for sure (i.e. there could be some untested component that leaves earlier).
3. In the latter part of the Discussion the authors state that SAS-1 is critical for centriole elimination. I would rephrase, as this seems to suggest it is required for centriole elimination, which is not the case. It might also be worth discussing that the elimination machinery clearly seems to target SAS-1 early on, but we don't yet know what this machinery is or how it is regulated.

Significance

The manuscript is generally well written and the data is of a high quality and is logically and clearly presented. Although the ultimate mechanisms regulating centriole elimination remain obscure (i.e. what triggers the loss of SAS-1, and how is this regulated?), the data presented here will be of significant interest to the centriole/centrosome field and I am supportive of publication. I have a few points that the authors should consider prior to publication.

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

Centrioles are small cylindrical structures with roles in cell division, motility, and signaling. Typically, centrioles are highly stable structures which can persist for many cell generations. However, in some cells, such as the female germ line of many species, centrioles are programmed for elimination. This process is essential for maintaining centriole number from one generation to the next in sexually reproducing organisms, yet in nearly all species the molecular mechanisms underlying how centrioles are eliminated is unknown. The current study utilizes the nematode C. elegans to explore how centriole architecture changes during the elimination program in the female germ line. Using a suite of light microscopy techniques, the authors provide a stunning visual perspective of how centrioles are disassembled during oogenesis and show that removal of the central tube component SAS-1, a key regulator of centriole stability, is an early event in elimination. I have no major objections to the work and enthusiastically endorse its publication with the following minor revisions.

Page 9 line 200: In the pcmd-1 mutant, the authors state that centriolar foci devoid of nuclei are present in rachis, but they do not mention in the text that there are also nuclei that lack centriole foci in early pachytene. This is mentioned in the figure legend, but I felt it was important enough to mention in the text.

Page 9 line 211. The authors found that in the absence of dynein heavy or light chain that centrioles remain associated with the nuclear envelope (rather than moving to the periphery). To me this was striking as dynein depletion in the embryo results in the opposite phenotype with centrioles losing attachment to the nuclear envelope and moving to the cell periphery (Gonczy et al. 1999 JCB 147:135). It might be worth pointing this out somewhere in the manuscript and speculating about the reasons for this difference.

Page 11 line 277: The authors state that elimination timing is not affected by the loss of SPD-5. This is a small but important point. It really is the absence of PCMD-1 and not SPD-5, as SPD-5 is still present in the cell. An alternative would be to say "in the absence of PCM" or "in absence of a pericentriolar accumulation of SAS-5".

Figure 4D: Why does loss of PCMD-1 result in a delay in oocyte maturation as judged by RME-2 accumulation? This is not mentioned in the paper. Is this a general response to a loss of PCM or is this specific to a loss of PCMD-1?

Figure 7 E and F. The authors measure the tubulin and SAS-4 intensity in wild-type and sas-1(t1521) embryos and conclude that microtubules and SAS-4 signals decay faster in the sas-1 mutant than in the control. To me, this is convinceingly the case with microtubules in panel E but I am not so sure this is the case with SAS-4 as shown in panel F. The differences in SAS-4 levels are much smaller between mutant and control. Could the authors provide statistical analysis to show how significant the differences are?

Page 15 line 363. I think this sentence should be reworded to: "Finally, we demonstrate that the central tube protein SAS-1 is the first of the factors analyzed here to leave centrioles..."

Significance

The work contained in this manuscript represents a fundemental step forward in understanding the process of centriole elimination. The authors have carefully described the stepwise disassembly of the centriole including changes in the architechure during oogenesis. They have identified loss of the centriole stability factor SAS-1, as an early event in the elimination program and have found that in a sas-1 mutant, the centriole disassembles prematurely. They have also shown that loss of SAS-1 is followed by expansion of the centriole and ultimately loss of structural integrity. This work should be of interest to a broad range of scientists including those interested in centrosome dynamics, germ line development, and more generally cell biologists.

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

To simplify the reading of our answers, we have numbered the questions of the reviewers. Similarly, to make the identification of the changes easier, we have written the changes in the manuscript in blue, orange and green when they were asked by reviewer 1, 2 or 3 respectively.

__Reviewer #1 __

Major comments:

1. One issue concerns the conclusion that microglial TNFa signaling shapes slow waves during NREM sleep (e.g., title; lines 148, 175-176; 180; 222-223; 288) on the basis of the data shown in Fig. 4b-d. Slow waves normally consist of two components, The amount of changes reported in the study might indeed seem small if the role of microglia was to gate slow-wave-sleep itself. In contrast, the effects we report are in the upper range of the modulations reported for NREM sleep slow-waves: depletion of microglial TNFα yielded a ~17 % decrease (relative to controls) in maximal slope (fig 4d); while sleep deprivation or physiological sleep-wake cycle have been associated with similar changes in the slope of slow-waves:

2. 6% increase after sleep deprivation and 17% increase after SOM-IN optogenetic stimulation (Funk et al. 2017, ref 17);

3. 15% above or below average in early and late sleep (Vyazovskiy et al. 2007, ref 79);
4. Daily fluctuation of the slow-wave slope is smaller that 15% (Hubbard et al. 2020, ref 50). Noteworthy, as noticed by this reviewer, slow saves consist of slow oscillations and delta waves. We also measured the slope and duration of slow oscillations and delta waves and found significant differences in both between control and micTNFα-KO mice. (see response to Reviewer 2 point 8).

The change in delta peak reported here following microglial TNFα depletion reveals a pre-eminence of slower waves (“d1” waves 0.75-1.75Hz), over the faster waves (“d2” waves 2.5-3.5Hz) the latter being the main form of oscillations potentiated by sleep need (Hubbard et al. 2020; ref 50). Similarly, enhanced SW-slope and short SW-period is associated with high sleep need (Hubbard et al. 2020; ref 50) and we found that microglial TNFα depletion reduces SW slope and enhances SW duration. Together, these results suggest a role of microglial TNFα in modulating delta oscillations in response to sleep need, which gives further support to our recent publication on the role of microglia in the expression of sleep need and sleep homeostasis (Pinto et al 2023; ref 80). Accordingly, we have now replaced in the figure 4 the panel 4b with the value of the peak of delta (whose significance is less clear) with the ratio of power in the two delta bands which has a clearer interpretation in light of the work cited above. Figure 4b and the text were changed accordingly (line 161 of the results and line 277 in the discussion). Similarly, microglia may contribute to changes in SWA related to memory consolidation following intense recruitment of cortical circuits (Huber et al. 2004, ref 83). This is now more clearly stated in the text (line 292).

We do not claim to have demonstrated that the change in slow oscillations fully explains the loss of memory consolidation, but instead we report the convergent findings that TNFa depletion in microglia produces alterations in sleep slow oscillations of the order of magnitude of sleep-need induced effects, and disrupts memory consolidation known to be sleep-dependent.

VGAT was used to identify GABAergic synapses in conjunction with GABAA receptors. Of various GABAergic interneurons, somatostatin (SOM)-containing GABAergic interneurons are known to be crucial for generating slow waves during NREM sleep through their axon terminals that target and concentrate in L1 (e.g., Funk et al., 2017, ref. 17). However, not all GABAA receptors in L1 would be associated with the inputs from SOM-containing GABA interneurons. For example, there are parvalbumin-containing GABA interneurons and their activation has been reported to DECREASE slow waves (Funk et al. 2017). This is relevant and should be discussed in relation to the results.

Answer: As pointed out by the reviewer, we do acknowledge that not all GABAergic synapses in L1 are associated to inputs of SOM+ interneurons. Other than the axonal projections of SOM+ interneurons, axons of the inhibitory neuronal types neurogliaform cells and canopy cells can be found in L1. This does not seem to be the case for PV+ interneurons whose somata is not located in L1 and lack projections to L1 (Schuman et al. 2021 - ref 19). We have now acknowledged and discussed this in the discussion:

Line 285: “In this study we show that L1 inhibitory synapses are modulated by microglia in a sleep-dependent manner. Our data favor the possibility that the microglia-targeted GABAergic synapses arise from SOM-IN. However, inhibition on L1 also originates local axonal arbors of neurogliaform and canopy cells and the identity of the microglia-regulated presynaptic terminals remains to be established.”

To follow up on the above, (1) it is unclear why NeuN was used to delineate cell bodies (Fig. 1e). In fact, SOM-containing GABA neurons (see above) have been shown to inhibit pyramidal neurons through presynaptic inhibition of excitatory inputs as well as postsynaptic inhibition of dendrites, but not cell bodies, of pyramidal neurons (see Funk et al. 2017 for references). Some discussion along this line would be useful and potentially important. (2) In addition, it would have been interesting to add an immunolabel for SOM to identify SOM-containing axon terminals associated with VGAT (Figs 1, 2), and this could be done for parvalbumin (see above) terminals as well; however, this analysis is optional and not required.

Answer:

• The rationale for analyzing GABAergic synapses in L5 was to assess whether the daily change in synaptic GABAAR observed in L1 had some degree of regional specificity or was rather a widespread event throughout the cortical layers. In this regard, NeuN was used to delineate cell bodies in order to assess changes in synaptic GABAARs at somatic synapses, known to be targeted by parvalbumin inhibitory neurons. In addition to the quantification provided in figure 1e, we also analyzed synaptic GABAARs at non-somatic VGAT clusters in L5 and observed no ZT6/ZT18 difference (data not shown).
• We do agree with this reviewer that looking selectively at GABAAR located at somatostatin pre-terminals would be extremely interesting. Accordingly, we have now tried to quantify the synaptic GABAR facing SOM positive presynaptic boutons identified in mice expressing tdTomato in SOM interneurons (figure S3). However, GABAAR immunostaining requires using an antigen retrieval method (see below) that, in our hands, bleaches tdTomato fluorescence. Our attempt to retrieve the lost SOM-tdTomato signal by IHC with RFP antibodies was unsuccessful (see below) suggesting that the structure of tdTomato is altered, thereby preventing such quantification.

---

---

Quantification of changes in synaptic GABAARs at synapses targeted by SOM interneurons (SOM-IN) is not technically possible. Top, Confocal images of adult brain cortical layer 1 show signal of GABAARγ2, gephyrin and VGAT obtained with different protocols for tissue processing (fresh-frozen, perfusion with paraformaldehyde and perfusion with paraformaldehyde followed by decloaking chamber-based heat-induced epitope retrieval). Visualization of bona-fide gephyrin+VGAT+ synapses and synaptic GABAAR clusters was only possible by epitope retrieval. Bottom, Confocal images of mice expressing tdTomato in SOM-IN (SOMCre/+R26tdTom/+) show expected signal revealing SOM-tdTomato+ presynaptic boutons after perfusion. When the protocol for epitope retrieval is used, SOM-tdTomato+ signal is lost, which could not be retrieved by IHC with 3 different antibodies (source and catalog number are indicated).

Minor comments:

It appears that n's are not consistently reported. Please check.

Answer: We have now:

-added n’s in the legend of figure 4.

-corrected a typo in the legend of figure 5 (line 664: “(f-i) n= 11” instead of “(f, g) n= 11”).

-in the legend of figure S2, we have added “g. Mean intensity…b-g, n= 53-60 FOVs from 5 mice per group”.

The Y-axis does not start from zero in some graphs. Although this might be a matter of preference, it can be misleading.

Answer: The Y-axis that did not start from zero have now been explicitly signaled (figure 4b, d; figure 5b, d, e)

In the supplementary information PDF, under Immunohistochemistry (IHC): "In direct IHC" in the first line of the paragraph should be "Indirect IHC".

Answer: this is corrected. The sentence now starts by “Cryostat sections…”. Also, “Table 2 – Antibodies and IHC method used” has been corrected accordingly.

---

---

__Reviewer #2 __

Major Comments:

1) (1) There are several instances where the authors state the experiments occurred "across the 24 h light/dark cycle" (Lines 42, 139), "during the sleep/wake cycle" (Lines 87, 242, 248), or "during sleep" (Lines 155, 220, 254). These statements are imprecise and can lead to erroneous interpretations of the data. For molecular studies, data were collected at a light period timepoint (Zeitgeber Time (ZT) 6) and a dark period timepoint (ZT18). While I appreciate the comparisons of the light and dark phases, 2 timepoints are not sufficient to claim that phenomena were tested across the light-dark cycle. More importantly, though, it is not accurate to claim outcomes from data collected during ZT6 occurred "during sleep" (or ZT18 outcomes occurred during wake). Although mice sleep more in the light period vs. the dark period, they are polyphasic sleepers and thus can be awake at ZT6 and asleep at ZT18. Therefore, statements should be edited for accuracy to instead state that phenomena were observed at ZT6/ZT18 or light/dark periods. (2) In addition, any figures (e.g., Figure S1) using x-axis labels of "W" and "S" should be relabeled as "ZT18" and "ZT6," respectively.

Answer : (1) We agree with this reviewer and we have consistently corrected these imprecise wordings.

(2) We have relabeled the figure S1 and replaced “W” and “S” by “ZT18” and “ZT6” respectively.

2) The authors claim that microglial TNFα plays a role in sleep-dependent memory consolidation (Title and Lines 20, 22, 178, 198, 224, 276, 288) based on a series of experiments using tests previously shown to have a sleep-dependent consolidation component. However, the authors did not assess sleep-dependent consolidation in the micTNFα-KO and the tCtl mice, and thus this conclusion cannot be drawn. This is because the experimental paradigms did not include sleep deprivation. Claims that outcomes are sleep-dependent need to be shown as absent/impaired after sleep deprivation especially in mutant (and control) lines that have not been previously tested in this context. As such, claims of sleep-dependent memory consolidation (including in the title) should be removed OR new experiments including sleep deprivation should be included.

Answer: Previous studies have shown that in the learning tasks that we have used, memory consolidation is lost in sleep-deprived animals (refs 53 to 55). Our current study indicates that memory consolidation is also lost in micTNFα-KO mice. Thus, we believe it would not be informative to perform sleep-deprivation in micTNFα-KO as requested by this reviewer since the outcomes (loss of memory consolidation) cannot be additive. However, we acknowledge that even though microglial TNFα depletion impacts slow waves that are known to play a causal role in consolidating the memory during sleep (refs 51 and 52), it cannot be excluded that TNFα depletion impairs memory consolidation by a non-sleep dependent pathway. Therefore, we have modified the title of the study to prevent misleading interpretations:

New title: “Microglial TNFα controls synaptic GABAARs, sleep slow waves and memory consolidation”

We have changed wording (lines 225 and 643). We have further added a cautionary note in the discussion:

Line 299 : “We now show that mice lacking microglial TNFα display impaired memory consolidation when tested in a complex rotator motor learning task or in the floor-texture recognition (FTR) task. In these two tasks, memory consolidation is sleep-dependent. This suggests a possible involvement of microglial TNFα in the sleep processes that promote memory consolidation, while leaving open the possibility of a non-sleep-dependent mechanism.”.

We are now more cautious in our conclusion and write (line 313) “This work demonstrates that microglia tune slow waves and support memory consolidation probably by acting during sleep”

3) (1) "This shows that P2RX7 and microglial TNFα drive daily fluctuations in CaMKII Thr286-phosphorylation and are required for sleep-dependent GABAAR synaptic upregulation in L1 during the light phase" (Lines 144 - 146). Similar to the above comment, it cannot be definitively concluded the P2X7R or microglial TNFα are required for sleep-dependent GABAAR synaptic upregulation because sleep deprivation studies were not conducted in the P2rx7-KO or micTNFα-KO mice. (2) Furthermore, there is no analysis (or citation) of P2rx7-KO mice sleep-wake expression nor has the micTNFα-KO sleep data been presented at this point to make any determinations on how (possibly perturbed) sleep-wake expression in these mice could affect the stated outcomes.

Answer:

(1) As discussed in our previous answer, both sleep deprivation (fig 1d) and inactivation of TNFα or P2RX7 (fig 3) completely prevent the synaptic GABAAR accumulation and CaMKII phosphorylation at ZT6. Performing sleep deprivation on micTNFα-KO and P2RX7 KO is thus not expected to exert an effect since the outcomes (loss of synaptic accumulation and phosphorylation of CaMKII at ZT6) cannot be additive. We actually conducted sleep deprivation studies in micTNFα-KO as suggested by this reviewer (see below). As expected however, sleep deprivation (SD6) has no further effect on GABAAR accumulation and CaMKII phosphorylation on micTNFα-KOs when compared to ZT6.

---

Impact of sleep deprivation on synaptic GABAAR and CaMKII phosphorylation in micTNFα-KO mouse brain. In complement to figure 3, SD prevents that increased accumulation of synaptic GABAARγ2 (left) and CaMKII phosphorylation (right) in tCTR, but has no effect in micTNFα-KO (green).

left: Mean intensity of GABAARγ2 clusters at gephyrin+VGAT+ synapses normalized to ZT18. n= 48 to 65 FOVs from 4-5 mice per group.

Right: Mean intensity of Thr286-phosphorylated CaMKII signal in L1 normalized to ZT18. n= 37 to 50 FOVs from 4-5 mice per group.

(2) Please note that analysis of sleep structure of micTNF-KO mice is shown in figure 4, which reveals “that microglial TNFα has limited effects on sleep-wake patterns as shown by the lack of major alterations in the amounts of wake, NREM and REM sleep between micTNFα-KO and tCTL mice along a light/dark cycle (fig. 4a).” (line 158). Similarly, analysis of baseline sleep-wake structure in P2RX7-KO mice revealed no abnormalities (Krueger et al 2010, ref 45). This has now been discussed in the text (lines 143).

4) There are some details regarding data analysis that are lacking:

1. How were bouts defined for each arousal state? Answer: We have now defined the bouts in the Materials and Methods (section : “Sleep recording and analysis”) : “Bouts were defined as consecutive 10-s epochs of similar vigilance state and could be as short as one epoch.”

2. (1) It seems more details are needed for EEG spectra analysis. From what values was the median derived and over what time period? How was each spectral bin normalized and over what time period? (2) What data (i.e., from what time period and duration) are shown in Figure 4b? Same question for Figure 4c-d? Were these time periods the same for controls and mutants given that NREM SWA changes across the light-dark cycle? Answer:

(1) The spectrum of bouts of 2.56s (512 points at 200Hz) was computed by FFT and the spectrum corresponds to the median of the FFT of all bouts. This information is now in the material and methods section “Spectral analysis”.

(2) The graphs reported in the text correspond to the 24h time period for both tCTL and micTNFα-KO mice. This has now been clearly indicated in figure 4 legend and in the material and methods section “Slow-wave analysis”.

1. How was NREM delta power normalized and analyzed and over what time period? Answer: Normalization corresponds to the division for each animal of the spectrum by the total power of the spectrum. Data reported correspond to the 24h time period.

5) Claims that GABAAR enrichment at synapses is sleep-dependent is based primarily on the data presented in Figure 1d reporting no increase in cortical GABAAR after sleep deprivation. A previous study (not cited) showed sleep deprivation increased GABAAR expression in CaMKIIα+ neurons in barrel cortex (Del Cid-Pellitero et al., Front Syst Neurosci, 2017). It would be helpful if the authors cited and discussed this study.

Answer: Indeed, Del Cid-Pellitero et al. show that GABAARs located around the soma of layer 5 neurons are increased upon sleep deprivation. Importantly, they used brain tissue collected after perfusion with paraformaldehyde without antigen retrieval. This protocol was shown to result in uniform surface labeling of GABAARs without staining the pool of receptors clustered at synapses (Gasser et al. 2006 Nature Protocols, PMID: 17487173). In this study we performed antigen retrieval that allows visualization of synaptic GABAARs (see figure 1). We and them are thus labelling different pools of GABAARs: synaptic vs. membrane at the soma level (comprising both synaptic and extrasynaptic). On the other hand, as we did not observe a difference in synaptic GABAARs at the soma level in L5 between ZT6 and ZT18, we did not assess sleep-dependency by performing sleep deprivation in this cortical layer. Together, we do not interpret their results as conflicting to our findings, but rather that different sleep- and wake-dependent mechanism exist to regulate the abundance of GABAARs at the subcellular level. We have now cited this work and include a brief discussion:

Line 58: “Sleep deprivation has previously been shown to increase GABAARs located around excitatory somas60. This suggests that the expression of GABAARs are differentially regulated depending on their subcellular localization”.

6) Some sentences/conclusions are overstatements:

1. "...discarding the possibility that lack of synaptic GABAARs enrichment upon PLX3397 treatment results from perturbed sleep during the light phase" (Lines 68 - 69). Only sleep time is reported to make this claim, but bout frequency, bout duration, and EEG spectra could be perturbed with this manipulation. This claim should be edited for accuracy or additional data (e.g., bout and spectral analysis) should be presented. In addition, Line 68 should be edited to state that "...microglia depletion does not alter sleep time during the light phase..." unless additional analyses are provided. Answer: We have now added the bout analysis in microglia depleted mice in extended table 2.

"TNFα, which is mostly if not exclusively produced by microglia in the brain..." (Lines 93 -94). Although microglia are a major source of TNFα, there is evidence other brain cell types also release TNFα. In addition, the citation provided does not support this exclusivity claim.

Answer: The reference 29 (Zeisel et al., reference 28 in the previous version) is a single cell transcriptomic study. The data are available online: http://mousebrain.org/adolescent/genesearch.html and show that TNFα mRNA is only detected in microglia. We are not aware of evidence showing that other brain cell types release TNFα. To our knowledge there is no brain RNA seq repository that shows TNFα expression in other brain cell-types e.g :

• https://celltypes.brain-map.org/rnaseq/mouse_ctx-hpf_smart-seq;
• http://biogps.org/#goto=genereport&id=21926
• http://www.brainrnaseq.org/
• "We thus anticipate that microglial TNFα may control REM by acting at the basal forebrain..." (Line 163). This statement is based on a cited study that reported REMS suppression (and increased NREMS time) after TNFα injection in the subarachnoid space of the basal forebrain. It is unclear to me why this statement is included when ICV and IV TNFα administration also reduce REMS (Shoham et al, Am J Physiol, 1987). Given these data and this statement is not being tested, it does not seem like it needs to be included. It should also be noted a previously reported (but not cited) global TNFα KO mouse (Szentirmai and Kapás, Brain Behav Immun, 2019) also showed increased REMS and REMS bouts, but this seemed to be a dark period phenotype (NREMS and Wake time, bout frequency, and bout duration were unaffected). This is an interesting detail to at least include in the second paragraph of the Discussion. Answer: To comply with this comment, we have now removed the sentence “We thus anticipate…” (line 163, now 160), and we have modified the second paragraph of the discussion so as to include the dark-period specificity described in Szentirmai and Kapás that we now cite.

7) It is unclear to me why the authors believe ATP in these studies has a neuronal origin (Lines 106, 132, 218) when other cell types also release ATP. Is this because of NMDA treatment? If so, NMDA receptors are also expressed on other cell types like astrocytes (Verkhratsky and Chvátal, Neurochemical Research, 2020). Answer: We do not believe and we did not write that ATP has a neuronal origin. Indeed, we wrote:

-line 104: “We next identified the signaling pathway between neuron and microglia”.

-line 130: “ATP released downstream neuronal activity activates microglial P2RX7”

-line 218: “…microglia sense neuronal activity through an ATP/P2RX7 signaling pathway”.

However, to rule out any misinterpretation, we have now added a sentence that explicitly recall the possible involvement of other cell types:

Line 132 “Noteworthy, our results do not exclude the possible involvement of other cell types acting between neurons and microglia”

8) Because the authors rationalize investigating memory consolidation based on micTNFα-KO changes in NREM SWA, I am curious if the authors considered parsing NREM SWA into slow oscillations and delta waves as Vaidyanathan et al (eLife, 2021) did. The reason for this is because slow oscillations are shown to be associated with memory consolidation, but delta waves are associated with weakening memories.

Answer: The parsing between delta waves and slow oscillations (SO) in the Vaidyanathan et al. article is based on a quantile separation of the size of the events (the top 15% of events are called SO while the rest may qualify as delta waves events; this is quite different from our definition which is based on deviations larger than 3 times the estimated standard deviation of slow fluctuations during Wake); it is worth noting that applying the definition from Vaidyanathan et al. to compare groups may introduce a bias in the interpretation if the rate of slow waves is modified between groups of animals. We have performed the parsing of slow events according to Vaidyanathan et al. and found similar changes for Slow Oscillations as using our definition (wider and “slower” SO). The same effect was observed in delta waves, suggesting that microglial TNFα affects both types of slow waves similarly. __ __

9) For the complex wheel task, micTNFα-KO mice seem to start and end with better performance compared to tCTL on S1 (although it is not clear if this difference was statistically evaluated). Would the conclusions from this experiment change if data were normalized to account for the apparent better starting performance? Answer: Despite a trend for a better performance of micTNFα-KO at S1, no significant difference in the mean performance was found between controls and micTNFα-KO mice at S1. We show below the mean performance at S1 and S2 for controls and micTNFα-KO mice. Moreover, learning within each session (both at S1 and S2) is not altered in micTNFα-KO (figure 5c) revealing that the ability to learn is not affected in microglia TNF-KO mice neither in S1 nor in S2 suggesting that memory impairment across sessions is not the result of saturation of learning capacity.

---

---

---

---

Mean performance in the two sessions (S1 and S2) of the complex wheel learning task in control and micTNFα-KO as measured by the mean time on the complex wheel (in seconds) of all trials in each session. ***p Nevertheless, we acknowledge that the apparent better starting performance could lead to misinterpretation of the results, and so as suggested by this reviewer, we normalized the data to the average of the last 3 trials in session 1 and computed using the normalized values the performance improvement (figure 5d) and consolidation (figure 5e). The same results were obtained. We thus interpret that the differences in memory consolidation between S1 and S2 (fig 5d, e) do not stem from changes in baseline performance.

---

---

---

Results on the complex wheel task following normalization to performance in S1. For each mouse, latency to fall off the complex wheel in each S1 and S2 trials was normalized to the average of the last 3 trials in S1. The normalized values were used to plot the graphs as in figure 5:

Left, latency to fall in S1 and S2;

middle, performance improvement;

right, consolidation of motor learning.

We have now modified figure 5 accordingly.

10) Many of the molecular studies emphasized a layer-specific effect in L1 vs. L5. It would be helpful if the authors could link (at least in the Discussion) this cortical-layer specificity with reported microglial TNFα effects on sleep parameters and memory consolidation.

Answer: According to this comment, we have now proposed a mechanism for the L1 vs L5 specificity in the discussion:

Line 254: “The layer 1 vs layer 5 specificity may arise from the molecular difference of GABAergic synapses across the somato-dendritic arbour as proposed. Alternatively, but not exclusively, it may result from a differential expression of TNF-R1 along the cortical layers and/or from layer-specific behavior of microglia”.

We hope that it will help understand our hypothesis of a link between microglial TNFα effect on upper layer synapses with the effect on sleep parameters and memory consolidation that are proposed from line 277 onwards.

Minor Comments:

1) For the experiments investigating TNF receptor (TNFR) involvement (fig. S5), it would have been interesting to see the response to human recombinant TNFα which interacts with TNFR1 but not TNFR2 whereas mouse recombinant TNFα interacts with both receptors (Lewis et al, PNAS, 1991).

Answer: The differential effect of human and mouse TNFα was not known by us. We do agree with this reviewer that the proposed experiment would have been interesting and would further confirm the data shown in Supp fig 5b showing an involvement of TNFR1 but not of TNFR2.

2) "Synapse plasticity in the sleeping brain likely supports crucial functions of sleep" (Line 33). I believe it is more accurate to instead state sleep supports synapse plasticity. The sentences immediately following also provide examples of sleep/wake mediating plasticity.

Answer: We have now replaced the sentence in line 29 for “Sleep drives plasticity of excitatory synapses”.

3) "Moreover, microglia depletion also abolished the reduction of synaptic AMPA receptor subunit GluA2 at ZT6 (fig. S1)..." (Lines 70 -71). The data in this figure shows an increase GluA2, but the data cited showed decreased excitatory transmission. This discrepancy is not discussed.

Answer: We agree with this reviewer that the we did not discuss the increased GluA2 at synapses upon microglial depletion. However, we believe that this aspect is beyond the scope of this study and although showing the data seemed important, we believed that discussing these data might lose the reader. We have now rephrased this sentence so as not to mislead the reader (line 68):

“Moreover, microglia depletion also reversed the reduction of synaptic AMPA receptor subunit GluA2 at ZT6”

4) If you normalize within treatment (e.g., PLX, SAP, minoc, 4-OHT, apy, PPADS, A74, PSB) rather than to controls (e.g., normalize PLX-iLTP to PLX-bsl instead of CTL-bsl) in Figures 2b-f, 3b-c, do you get the same results? Similarly, if you normalize PLX treatment to CTL ZT18 in Figure 1c-d, do you get the same general outcome?

Answer: As requested by this reviewer, we have normalized to treatment rather than to control for all the experiments shown in figure 2b-f. The graph below with this normalization shows that the same results are obtained.

Concerning figures 1c-d and 3b-c, the normalization suggested cannot be performed because CTL and PLX treated mice or mutant mice were not processed for immunohistochemistry simultaneously, and so intensity values are not comparable.

5) Is there a significance marker missing in Figure 2g for bls vs. BzATP?

Answer: The significance marker it is not missing. Even though the individual experiments performed consistently show an increase in CaMKII T286 phosphorylation with BzATP treatment, the difference does not reach statistical significance between bsl and BzATP using a nested one-way ANOVA followed by Sidak’s multiple comparison test (p=0.09). There was however a typo in the legend of figure 2g about the statistical test used, which has been corrected (lines 611).

6) It is unclear if the fig. 5a callout is the correct one at Line 143. If it is correct, then it is confusing (to me) in the way it is currently associated with the text.

Answer: at line 143 (now 142), the callout is fig S5a (supplementary figure 5a that confirms TNFα depletion) and not 5a.

7) There is a missing reference in Line 279 after "NOR."

Answer: we have now added the reference.

__Reviewer #3 __

Summary: This paper tries top address how microglia-related TNFa modulate the REM sleep and motor behavior consolidation, using layer I gabaergic synapse enhancement (possible by SOM interneurons).

The methods and results are solid/convincing and easily to be followed and they seem logically for the conclusion which they state.

However there are major issues which need to be addressed before formal submission.

Reviewer #3 (Significance (Required)):

These works have serious limitations to be addressed before submitting to formal journals.

(1) The in vivo sleep experiments with micTNFa-KO (fig. 4 and extended table S1, Fig S8 a/b) indicate that layer I GABAergic synapse potentiation modulates the start of down-state of slow-waves, which supposes to affect the NREM sleep. Controversially, NREM sleep is not affected (with SW down state duration increased in KO mice) and only REM sleep is influenced. This is opposite to the literature that GABAergic synapses in the cortex or thalamus determine the power of slow-waves and The fast transition to down state suggests that cortical neurons should be less active for REM sleep compared with NREM sleep.

Answer This reviewer pointed out that in extended table 1, micTNFα-KO mice spent more time in REM sleep over 24 hours. However, we believe that this increase is marginally significant because there was no parallel decrease in the time spent in Wake or NREM sleep. Accordingly, the mean duration of REM sleep is not affected (extended table 1) and when measuring the amount of REM sleep in 2-hour bins, figure 4a shows no difference between micTNFα-KO and tCTL mice. Additionally, figure 4 shows that there is no change in the electrophysiological features of REM sleep. On the other hand, the electrophysiological parameters of NREM sleep, such as the duration and slope of slow waves, are affected, as pointed out by this reviewer. Therefore, we do not view our findings as controversial. While we have not tested this hypothesis, we do not exclude that the lack of synaptic plasticity observed in micTNFα-KO mice is linked to the prolonged REM sleep.

(2) REM sleep can consolidate motor learning. However, the data (Fig. 5) do not consistently support this, although the authors cite the literature to support their results with complex motor learning.

Answer: In line with the comment of this reviewer, Li et al. 2017 (ref 13) study showed that REM sleep deprivation impairs the performance improvement in a motor learning tasks. However, and as noticed by this reviewer, the amount of REM sleep is increased in micTNFα-KO and we are not aware of any study that has shown the consequences of increased REM sleep on motor learning consolidation. The floor-texture recognition task that we have used to show an alteration of consolidation in micTNFα-KO was shown to be NREM sleep-dependent (Miyamoto et al 2016, ref 54).

(3) NMDA-iLTP needs to be further addressed since NMDA with CNQX in oganotypic slices supposes not be able to activate NMDARs that need co-activation of AMPARs to depolarize to remove the Mg-blockade before NMDAR activation in neurons. To further strengthen this result, ACh should be co-applied with NMDA (if not AMPA) since only REM sleep is enhanced in this study indicating that ACh should be involved to activate neurons during rem sleep, more relevant to REM sleep enhancement.

Answer: we followed the iLTP protocol in culture described by Petrini et al. (reference 26) and verified an increase in GABAAR accumulation at synapses. Furthermore, in preliminary experiments (not shown), we found that this effect depended on CaMKII, as previously reported by Chiu et al. who used NMDA only on acute slices (reference 27). These findings indicate that the application of NMDA+CNQX on organotypic slices replicates the synaptic effects observed in primary cultures (in which NMDA+CNQX is used) and acute slices (only NMDA). While investigating a novel protocol of synaptic plasticity using Ach and its possible connection to REM sleep is intriguing, we believe it would be more appropriate to explore this in a separate 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 #3

Evidence, reproducibility and clarity

Summary: This paper tries top address how microglia-related TNFa modulate the REM sleep and motor behavior consolidation, using layer I gabaergic synapse enhancement (possible by SOM interneurons).

The methods and results are solid/convincing and easily to be followed and they seem logically for the conclusion which they state.

However there are major issues which need to be addressed before formal submission.

Significance

These works have serious limitations to be addressed before submitting to formal journals.

1. The in vivo sleep experiments with micTNFa-KO (fig. 4 and extended table S1, Fig S8 a/b) indicate that layer I GABAergic synapse potentiation modulates the start of down-state of slow-waves, which supposes to affect the NREM sleep. Controversially, NREM sleep is not affected (with SW down state duration increased in KO mice) and only REM sleep is influenced. This is opposite to the literature that GABAergic synapses in the cortex or thalamus determine the power of slow-waves and The fast transition to down state suggests that cortical neurons should be less active for REM sleep compared with NREM sleep.
2. REM sleep can consolidate motor learning. However, the data (Fig. 5) do not consistently support this, although the authors cite the literature to support their results with complex motor learning.
3. NMDA-iLTP needs to be further addressed since NMDA with CNQX in oganotypic slices supposes not be able to activate NMDARs that need co-activation of AMPARs to depolarize to remove the Mg-blockade before NMDAR activation in neurons. To further strengthen this result, ACh should be co-applied with NMDA (if not AMPA) since only REM sleep is enhanced in this study indicating that ACh should be involved to activate neurons during rem sleep, more relevant to REM sleep enhancement.

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:

Pinto et al. sought to determine the role of microglia in sleep and plasticity. Using a combination of in vivo experiments in mice and organotypic slices, they report a molecular circuit whereby microglia respond to ATP via the purinergic receptor P2X7R to release tumor necrosis factor α (TNFα). TNFα then acts in its soluble form at TNF receptor 1 (TNRF1) to increase synaptic enrichment of GABAA receptors (GABAAR) in layer 1 (L1) of cortex (but not L5) during the light phase. These results suggest that microglial TNFα mediates cortical inhibitory synapses in a layer- and time-of-day-specific manner. The authors also showed selectively disrupting microglia TNFα alters slow wave activity (SWA) during non-rapid eye movement sleep (NREMS) and impairs memory consolidation.

Major Comments:

The authors performed several well-designed and controlled studies to uncover microglia regulation of GABAAR enrichment at synapses during the light period uncovering a nicely presented molecular circuit that included upstream and downstream mechanisms. They also nicely probed this circuit in vivo with a conditional, microglial-specific depletion of TNFα to determine the role of microglial TNFα in sleep and memory consolidation. Although the experiments were well done, the authors made some conclusions that cannot be determined by the experiments presented in this manuscript. Specifically, there are several claims that some of the phenomena occurred "during sleep" or are "sleep-dependent" when the experiments were not designed to test these claims. I provide more detailed comments below:

1. There are several instances where the authors state the experiments occurred "across the 24 h light/dark cycle" (Lines 42, 139), "during the sleep/wake cycle" (Lines 87, 242, 248), or "during sleep" (Lines 155, 220, 254). These statements are imprecise and can lead to erroneous interpretations of the data. For molecular studies, data were collected at a light period timepoint (Zeitgeber Time (ZT) 6) and a dark period timepoint (ZT18). While I appreciate the comparisons of the light and dark phases, 2 timepoints are not sufficient to claim that phenomena were tested across the light-dark cycle. More importantly, though, it is not accurate to claim outcomes from data collected during ZT6 occurred "during sleep" (or ZT18 outcomes occurred during wake). Although mice sleep more in the light period vs. the dark period, they are polyphasic sleepers and thus can be awake at ZT6 and asleep at ZT18. Therefore, statements should be edited for accuracy to instead state that phenomena were observed at ZT6/ZT18 or light/dark periods. In addition, any figures (e.g., Figure S1) using x-axis labels of "W" and "S" should be relabeled as "ZT18" and "ZT6," respectively.
2. The authors claim that microglial TNFα plays a role in sleep-dependent memory consolidation (Title and Lines 20, 22, 178, 198, 224, 276, 288) based on a series of experiments using tests previously shown to have a sleep-dependent consolidation component. However, the authors did not assess sleep-dependent consolidation in the micTNFα-KO and the tCtl mice, and thus this conclusion cannot be drawn. This is because the experimental paradigms did not include sleep deprivation. Claims that outcomes are sleep-dependent need to be shown as absent/impaired after sleep deprivation especially in mutant (and control) lines that have not been previously tested in this context. As such, claims of sleep-dependent memory consolidation (including in the title) should be removed OR new experiments including sleep deprivation should be included.
3. "This shows that P2RX7 and microglial TNFα drive daily fluctuations in CaMKII Thr286-phosphorylation and are required for sleep-dependent GABAAR synaptic upregulation in L1 during the light phase" (Lines 144 - 146). Similar to the above comment, it cannot be definitively concluded the P2X7R or microglial TNFα are required for sleep-dependent GABAAR synaptic upregulation because sleep deprivation studies were not conducted in the P2rx7-KO or micTNFα-KO mice. Furthermore, there is no analysis (or citation) of P2rx7-KO mice sleep-wake expression nor has the micTNFα-KO sleep data been presented at this point to make any determinations on how (possibly perturbed) sleep-wake expression in these mice could affect the stated outcomes.
4. There are some details regarding data analysis that are lacking:
   • a. How were bouts defined for each arousal state?
   • b. It seems more details are needed for EEG spectra analysis. From what values was the median derived and over what time period? How was each spectral bin normalized and over what time period? What data (i.e., from what time period and duration) are shown in Figure 4b? Same question for Figure 4c-d? Were these time periods the same for controls and mutants given that NREM SWA changes across the light-dark cycle?
   • c. How was NREM delta power normalized and analyzed and over what time period?
5. Claims that GABAAR enrichment at synapses is sleep-dependent is based primarily on the data presented in Figure 1d reporting no increase in cortical GABAAR after sleep deprivation. A previous study (not cited) showed sleep deprivation increased GABAAR expression in CaMKIIα+ neurons in barrel cortex (Del Cid-Pellitero et al., Front Syst Neurosci, 2017). It would be helpful if the authors cited and discussed this study.
6. Some sentences/conclusions are overstatements:
   • a. "...discarding the possibility that lack of synaptic GABAARs enrichment upon PLX3397 treatment results from perturbed sleep during the light phase" (Lines 68 - 69). Only sleep time is reported to make this claim, but bout frequency, bout duration, and EEG spectra could be perturbed with this manipulation. This claim should be edited for accuracy or additional data (e.g., bout and spectral analysis) should be presented. In addition, Line 68 should be edited to state that "...microglia depletion does not alter sleep time during the light phase..." unless additional analyses are provided.
   • b. "TNFα, which is mostly if not exclusively produced by microglia in the brain..." (Lines 93 -94). Although microglia are a major source of TNFα, there is evidence other brain cell types also release TNFα. In addition, the citation provided does not support this exclusivity claim.
   • c. "We thus anticipate that microglial TNFα may control REM by acting at the basal forebrain..." (Line 163). This statement is based on a cited study that reported REMS suppression (and increased NREMS time) after TNFα injection in the subarachnoid space of the basal forebrain. It is unclear to me why this statement is included when ICV and IV TNFα administration also reduce REMS (Shoham et al, Am J Physiol, 1987). Given these data and this statement is not being tested, it does not seem like it needs to be included. It should also be noted a previously reported (but not cited) global TNFα KO mouse (Szentirmai and Kapás, Brain Behav Immun, 2019) also showed increased REMS and REMS bouts, but this seemed to be a dark period phenotype (NREMS and Wake time, bout frequency, and bout duration were unaffected). This is an interesting detail to at least include in the second paragraph of the Discussion.
7. It is unclear to me why the authors believe ATP in these studies has a neuronal origin (Lines 106, 132, 218) when other cell types also release ATP. Is this because of NMDA treatment? If so, NMDA receptors are also expressed on other cell types like astrocytes (Verkhratsky and Chvátal, Neurochemical Research, 2020).
8. Because the authors rationalize investigating memory consolidation based on micTNFα-KO changes in NREM SWA, I am curious if the authors considered parsing NREM SWA into slow oscillations and delta waves as Vaidyanathan et al (eLife, 2021) did. The reason for this is because slow oscillations are shown to be associated with memory consolidation, but delta waves are associated with weakening memories.
9. For the complex wheel task, micTNFα-KO mice seem to start and end with better performance compared to tCTL on S1 (although it is not clear if this difference was statistically evaluated). Would the conclusions from this experiment change if data were normalized to account for the apparent better starting performance?
10. Many of the molecular studies emphasized a layer-specific effect in L1 vs. L5. It would be helpful if the authors could link (at least in the Discussion) this cortical-layer specificity with reported microglial TNFα effects on sleep parameters and memory consolidation.

Minor Comments:

1. For the experiments investigating TNF receptor (TNFR) involvement (fig. S5), it would have been interesting to see the response to human recombinant TNFα which interacts with TNFR1 but not TNFR2 whereas mouse recombinant TNFα interacts with both receptors (Lewis et al, PNAS, 1991).
2. "Synapse plasticity in the sleeping brain likely supports crucial functions of sleep" (Line 33). I believe it is more accurate to instead state sleep supports synapse plasticity. The sentences immediately following also provide examples of sleep/wake mediating plasticity.
3. "Moreover, microglia depletion also abolished the reduction of synaptic AMPA receptor subunit GluA2 at ZT6 (fig. S1)..." (Lines 70 -71). The data in this figure shows an increase GluA2, but the data cited showed decreased excitatory transmission. This discrepancy is not discussed.
4. If you normalize within treatment (e.g., PLX, SAP, minoc, 4-OHT, apy, PPADS, A74, PSB) rather than to controls (e.g., normalize PLX-iLTP to PLX-bsl instead of CTL-bsl) in Figures 2b-f, 3b-c, do you get the same results? Similarly, if you normalize PLX treatment to CTL ZT18 in Figure 1c-d, do you get the same general outcome?
5. Is there a significance marker missing in Figure 2g for bls vs. BzATP?
6. It is unclear if the fig. 5a callout is the correct one at Line 143. If it is correct, then it is confusing (to me) in the way it is currently associated with the text.
7. There is a missing reference in Line 279 after "NOR."

Significance

The role of non-neuronal cells in sleep and sleep-related processes like learning and memory is relatively unexplored and this is especially true of microglia. The authors present a nicely done series of rigorous experiments that reveal a microglial-centric molecular circuit of inhibitory synaptic modulation that differs between the light and dark periods and plays a role in NREM slow wave activity and memory consolidation. However, most of the claims of sleep dependency of the reported phenomena have not been directly tested in this manuscript and thus await additional experiments or re-framing of the stated conclusions. Re-framing these conclusions without claims of sleep dependency still provides very interesting and informative data about the mechanistic role of microglia in inhibitory synapse plasticity and memory consolidation that may be of interest to researchers interested in learning and memory, synaptic plasticity, and sleep.

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:

Using both in vivo and ex vivo approaches in mice, the authors showed that microglia in layer 1 (L1) of the frontal cortex modulate the level of GABAA receptors at L1 GABAergic synapses depending on the light/dark phase and, more specifically, sleep/wake state (high during sleep). This was shown to be mediated through purinergic signaling via microglial P2RX7 receptors followed by microglial release of TNFa then CaMKIIa phosphorylation in neurons. Microglia-selective TNFa-KO mice showed normal sleep/wake cycles except for an increase in REM sleep amount, but cortical EEG slow waves during NREM sleep, a measure of sleep propensity, were slower in the delta range (1-4 Hz), but without any change in delta power. These animals also showed deficits in memory consolidation in two out of three memory tasks used.

Major comments:

1. The paper is clearly written and easy to follow, with virtually no typos or editorial errors. Both the introduction and discussion are informative and well-referenced.
2. The experiments are generally carefully designed including appropriate controls and comparison groups (e.g., L1 vs. L5; GABAAR vs. AMPAR; PLX vs. saporin vs. minocycline). The results are presented appropriately and often in detail including supplementary figures and tables.
3. One issue concerns the conclusion that microglial TNFa signaling shapes slow waves during NREM sleep (e.g., title; lines 148, 175-176; 180; 222-223; 288) on the basis of the data shown in Fig. 4b-d. Slow waves normally consist of two components, < 1Hz (slow oscillations) and 1-4 Hz (delta waves), and Fig. 4b shows a modest slowing in delta range (from ~2.2 Hz to ~1.7 Hz, from reading the graph for means). Importantly, there was no change in the spectral density in delta range. In the opinion of this reviewer, this is a modest effect and its significance and impact remain to be investigated. Arguably, the effects on memory consolidation could have been a result of microglial TNFa gene deletion elsewhere in the brain of these KO mice. Modification of the claim on slow waves should be considered.
4. VGAT was used to identify GABAergic synapses in conjunction with GABAA receptors. Of various GABAergic interneurons, somatostatin (SOM)-containing GABAergic interneurons are known to be crucial for generating slow waves during NREM sleep through their axon terminals that target and concentrate in L1 (e.g., Funk et al., 2017, ref. 17). However, not all GABAA receptors in L1 would be associated with the inputs from SOM-containing GABA interneurons. For example, there are parvalbumin-containing GABA interneurons and their activation has been reported to DECREASE slow waves (Funk et al. 2017). This is relevant and should be discussed in relation to the results.
5. To follow up on the above, it is unclear why NeuN was used to delineate cell bodies (Fig. 1e). In fact, SOM-containing GABA neurons (see above) have been shown to inhibit pyramidal neurons through presynaptic inhibition of excitatory inputs as well as postsynaptic inhibition of dendrites, but not cell bodies, of pyramidal neurons (see Funk et al. 2017 for references). Some discussion along this line would be useful and potentially important. In addition, it would have been interesting to add an immunolabel for SOM to identify SOM-containing axon terminals associated with VGAT (Figs 1, 2), and this could be done for parvalbumin (see above) terminals as well; however, this analysis is optional and not required.

Minor comments:

1. It appears that n's are not consistently reported. Please check.
2. The Y-axis does not start from zero in some graphs. Although this might be a matter of preference, it can be misleading.
3. In the supplementary information PDF, under Immunohistochemistry (IHC): "In direct IHC" in the first line of the paragraph should be "Indirect IHC".

Significance

There is increasing evidence for and interest in the role of microglia in modulating synapses and neuronal circuits underlying various behaviors including sleep. However, the role of cortical microglia in sleep or NREM slow waves (a measure of sleep propensity or sleep need) is unclear. GABAergic synapses in cortical L1 are known to play an important role in NREM slow waves. Thus, the evidence that microglial P2X7-TNFa signaling enriches synaptic GABAA receptors selectively in L1 (not in L5) and during sleep is important and novel.

On the other hand, in this reviewer's opinion, the effect of microglia-selective TNFa gene deletion on slow waves during NREM sleep seems modest (minor slowing but no change in power) and it is also unclear whether the effects on memory consolidation in two out of three tasks were due to the observed change in slow waves, or other alterations that are likely in these KO mice.

The reviewer's expertise: sleep neurobiology. Familiar with microglia and much of the techniques, but not ex vivo techniques.

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

Further work required on divergence to address the observed level of gene flow between wild and crop populations and the lack of admixed/hybrid genomes.

• Authors should plot Fst and Dxy along the main scaffolds. It would permit to see whether there is huge peaks of divergence and differentiation between the two populations.

We will measure Dxy across the genome and plot against Fst in the larger contigs. Check for similarities and differences in these two measures and their relationship with effector locations. Relate results to effectors as well as broader demography as highlighted by the network.

Virulence can be mediated at the expression level by effector silencing and the reviewer suggests looking for premature stop codons, deletion of an exons, or mutations in the promoter region.

We will run SNPeff to annotate and classify the severity of variants. We will relate that to potential roles in silencing in effectors relative to non-effector genes.

Simulations would provide stronger support for conclusions.

Absolutely. Simulations are important to validate our conclusions and improve our hypotheses as to the levels of selection, rates of gene flow and/or boom and bust. Our simulation would look at:

• The strength of balancing selection required to preserve diversity in pathogen virulence genes.
• The strength and direction of selection required to partition that diversity to produce increased Fst we observe at virulence genes.
• The impact of differential rates of recombination in the wild and crop populations
• The impact of clonality in the crop population at multiple levels:
• On divergence between populations and levels of inbreeding (Fis)
• On boom-and-bust dynamics and the frequency of successive invasions

To parameterise this model, we would need to estimate the rate of recombination using linkage decay across a near/pseudo chromosomal assembly. Our assembly isn’t contiguous enough to estimate recombination rates for our populations. Given the parameter space we must investigate, combined with unknowns, we feel that the investment required to design and test such a model is significant, including requiring a new (long read, phased) genome assembly. This is our aim but without that data now, we feel that a simulation would not be strong enough to get through the rigor of peer review. We are happy to add a discussion of the importance these next steps to validate our conclusions, in the Discussion section.

Figure 2 network suggests strong divergence between populations, and this needs further exploration because divergence and the locus level is very low.

• The reviewer requests a PCA
• (Reviewer #2 missed the Machine Learning in the supplementary which also speaks to this work).

Reorder Figure 2 to reduce confusion. Reduce the content in Figure 2 to include only the map, the network, and the admixture plot. Add a new Figure 3 which would include a larger PCA and the supplementary data which uses Machine Learning to attempt to partition individuals into clusters/populations.

Authors should also look for how many effector genes are non-expressed in cultivated population face to wild population.

The reviewer’s suggestion of analysis of premature stop codons etc will be done using SNPeff.

Run Selscan (ZA Szpiech and RD Hernandez (2014) or similar to look at indicators of selection.

It is feasible to run selection scan software, although this would be heavily caveated because these methods often do not account for clonal expansion in a single population.

Address Minor Comments

Reviewer #2

• Effector candidates were not evaluated/characterized in any form.
• Authors should compare pathogen features to other related species and try to contrast what stands out, especially the effectors' diversity

The reviewer referred to a statement in which we suggest that it is difficult to functionally annotate effectors (“According to the authors, it is difficult to functionally annotate these genes in general”). This statement was not intended to suggest that we did not annotate them, only that, because effectors are quickly evolving, fewer of them tend to receive an annotation, as compared to non-effector genes. In fact, we use shared annotations to refer to the presence of shared effector annotations in other rust species.

Details of cross species functional annotation were included in Supporting Information 01. They included annotation using AHRD, UniProt (Swss-Prot and TrEMBL) blast and InterProScan. AHRD uses a database of unbiased ground truth set of high-quality protein annotations with minimal redundancy to assign GO annotations. UniProt is the world’s leading high-quality protein sequence and functional information. It contains more than 190 million sequences with which to assign functional annotations to proteins. InterProScan was used to assign proteins into families as well as predict domains. All these methods utilise cross species information to assign gene/domain function and ontology (GO), the output from these is included for each gene along with that gene’s population genetic signature (Supporting Information 08).

In the results section we did highlight effector functional annotations and conservation among the Pucciniales (e.g. Rust Transferred Protein, Alpha-amylase, CSEP-06 and PriA, among others). We will clarify that statement to reflect our efforts in that area and throughout.

Where exactly the machine learning was used?

It’s in the supplement. Accounting for this comment as well as Reviewer 1 & 2 about the complexity of Figure 2 we plan to bring those results into the main document. This would allow us to unpack genetic diversity and differentiation as a separate figure from the map and network.

Address Minor Comments

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: In the present study, using beet (Beta vulgaris) - rust (Uromyces beticola) as a model system, the authors set out to assess how crop pathogens evolve to evade resistance using wild reservoirs. They tested whether 1) the genes necessary to success in wild and crop environments are more genetically differentiated between pathogen populations and 2) the rate of clonal to sexual reproduction is higher in crop pathogen populations. Using freshly obtained 42 pathogen isolates from both crop and wild beets across the east of England, the authors assessed the genetic variation among virulent genes(effectors) between wild and crop pathogens. They found evidence for higher signals of diversity and differentiation in effectors and significant differences in reproductive rates between the wild and crop pathogen populations. They highlight that these findings can be used to identify candidate genes for pathogen survival in crops and develop methods to circumvent crop pathogen resistance. Additionally, they developed a new DNA peel extraction protocol for pathogens and produced a new annotation of Uromyces beticola genome annotation.

Major comments:

• The study design and the methodologies are appropriately explained and the statistical analyses are strong enough to draw the conclusions presented in the manuscript. The results are adequately explained and the inferences drawn from them are satisfactory.
• The putative effectors (virulent genes) were identified based on the assumption that gene products secreted outside of the fungal cell and into the host are host interaction genes, potentially facilitating infection. However, these candidates were not evaluated/characterized in any form. According to the authors, it is difficult to functionally annotate these genes in general. However, I believe at least the predicted functionality can be checked with published adequately annotated genomes of related species. This comparison is lacking in the analysis. Not having confirmed the functionality of at least some of the effectors undermines the finding that the study reflects the actual genetic differentiation in infectious genes.
• Similarly, comparisons of current findings to a related species/system are missing. Authors should compare pathogen features to other related species and try to contrast what stands out, especially the effectors' diversity.
• Although the authors claim that machine learning was utilized in the manuscript, where exactly the machine learning was used is not clear. The models used in the analyses are already implemented in the software packages and methods described in the manuscript. I did not see any machine learning method being applied to improve the analysis either. If it is actually used, it would be beneficial to highlight for what and where it was used and how it improved specific analyses.

Minor comments:

• Lines 184 - 186: Can the lack of admixture and gene low among these wild isolates also explain this observation? what about the levels of FIS in these isolates? Clonality in these populations may have a significant impact on the genetic diversity in these populations.
• lines 216-220: Is this also reflected in the excess of heterozygosity non-effectors in these crop populations? The mutations should equally accumulate in both gene categories.
• lines 219-220: it is not clear which CDS are being referred to here; Are you talking about the correlation between the CDSs of wild and crops or effectors and non-effectors?
• Figure 1: I suggest separating F & G from the rest
• Figure 3: D. Unless this is a noe to one window comparison of pi, this plot does not necessarily show a correlation. Please explain how the windows were treated in this comparison.
• Figure 4: A. I would expect a relatively high correlation between the FST and pi in effectors. Does this include both wild and crop effectors?
• I spotted a number of typos throughout the manuscript. So I suggest the authors pay attention to punctuation and typos.

Significance

This study presents a critical comparative analysis of crop pathogens in their wild populations. It highlights the significance of assessing the crop pathogen genetic diversity against their wild background/relatives to identify how crop pathogens evolve to evade crop resistance. And in turn, it will help us to improve our crop varieties to be better resistant to pathogens in this era of ever-increasing demand for crop production.

Further, the present study also provides a new methodology with an annotated genome of beet pathogen Uromyces beticola to identify candidate crop resistance genes in other related pathogens. The scientific community will also benefit from the protocol they developed to extract pathogens from host peels.

Therefore, I believe this work will reach a wide audience in genetics, genomics, agriculture, crop development, and landscape genomics.

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 study presents a pathosystem beet/Uromyces beticola in order to understand how reservoirs can play a role in ermergence of new virulences. To do this, the authors sample cultivated beets and wild beets in England and resequence 42 genomes of U beticola found on these two beet species (24 from sugar beet hosts and 18 from coastal places on wild beet hosts). The authors use population genomics tools to explore population structure and compare diversities of the two populations found. Indeed they found a population of U beticola exclusively living on wild beets, and a population infecting both sugar and wild beets. They compare genes encoding for effectors (important in interactions with hosts) with genes encoding for other proteins. They found that genes encoding for effectors are more diverse in wild compartiment than genes encoding for other types of proteins. In general, the wild compartiment is more diverse than the cultivated one. The authors draw conclusions about the role of reservoir of wild population for emergence of new virulence in the cultivated population. At last, as the authors found excess of heterozygosity in the cultivated population, they conclude about clonal reproduction in this population.

Major comments:

• Are the claims and the conclusions supported by the data or do they require additional experiments or analyses to support them?

Although the paper is well written and the population genomics studies were well done, the analyses are still preliminary and hence conclusions are not accurate. Indeed, in Fig2 authors show a tree representing the clear divergence between strains found in sugar beets and the ones found in wild beets (excepted for five isolates found in wild beets but belonging to the cultivated clade). Despite the fact that this apparent divergence is supported by other analyses, the authors do not conclude about the lack of gene flow between theses two populations. Indeed, the gene flow occurred there would have been admixed genomes and no clearcut delineations between the two populations. In other word, they authors have not found hybrids in their sampling. In general, divergence is not really studied in this paper. The comparison between genes encoding for effectors and the ones encoding for other genes is very interesting. However, the authors just forget that sometimes virulence is acquired by effector silencing. Indeed an effector that is no more expressed can not be recognised by host, and then resistance can be overcome. The authors should look for effectors that are no more expressed (with stop codon for example, deletion of an exon, or mutated in the promoter region) in crop population. They could find other good candidates for adaptation. In general, conclusions are badly supported. The authors should use simulations for their model validation. This study strongly deserves it. I will detail this in the following. - Please request additional experiments only if they are essential for the conclusions. Alternatively, ask the authors to qualify their claims as preliminary or speculative, or to remove them altogether. - First validate your assumations with models simulated. If the authors assume that population infecting cultivated beets come from population infecting wild beets, they should be able to confirm this hypothesis by simulations. For instance, authors could use ABC method in order to check the posterior probability of such a model. - Authors should use divergence statistics in order to check whether there is divergence or not on their data. For example, use Dxy in order to check the degree of divergence between wild and cultivated population. As for evidence in Figure 2, there is a strong divergence between the two populations. It could be interesting to check whether there is gene flow or not between these two populations. - Authors should also look for how many effector genes are non expressed in cultivated population face to wild population. - Authors should plot Fst and Dxy along the main scaffolds. It would permit to see whether there is huge peaks of divergence and differentiation between the two populations. - Are the data and the methods presented in such a way that they can be reproduced?

Yes, the data can be reproduced as well as the analyses. - Are the experiments adequately replicated and statistical analysis adequate?

Both experiments and statistics are adequate.

Minor comments:

• Specific experimental issues that are easily addressable.

Yes they are - Are prior studies referenced appropriately?

Yes they are - Are the text and figures clear and accurate?

Figure 2 is somewhat hard to understand. There is too much data here. - Do you have suggestions that would help the authors improve the presentation of their data and conclusions?

I would prefer a PCA plot, just showing strains plotted along the first axis and showing that there is no hybrid.

Referee Cross-commenting

Hello everyone I just start the comment session by providing a few points that I think to be important to be treated in a future version of the paper. 1- Characterize the relationships between wild and agricultural populations. As shown on figure 2, the tree presented clearly indicates divergence between wild and agricultural strains. A PCA would be interesting to be plotted, as it may indicate that there is no hybrid between populations. In a general manner, statistics like Dxy or Da as well as Fst should be plotted along the genome. 2- The scenario should be validated using simulations and tested against a null hypothesis. 3- Virulence can be acquired through effector losing function. Thus, variations like occurrence of codon stop, delection in ORF, or mutations altering the promoter region should be checked.

Significance

Provide contextual information to readers (editors and researchers) about the novelty of the study, its value for the field and the communities that might be interested.

The following aspects are important:

• General assessment: provide a summary of the strengths and limitations of the study. What are the strongest and most important aspects? What aspects of the study should be improved or could be developed? This study is interesting in showing the crucial role of wild compartiment as a reservoir of virulence. However, as the data are well produced, their analysis suffer from several flaws. As divergence is not analysed, and only differentiation is shown. In addition, the model proposed is not validated by simulation, not even tested by ABC. In order to be more conclusive, selection should be tested. Fst variance is not a good predictor of selection. I would recommend to use Selscan (ZA Szpiech and RD Hernandez (2014) ) in order to test for selection on genomic data. It would give real clues for selection acting on cultivated population.
• Audience: describe the type of audience ("specialized", "broad", "basic research", "translational/clinical", etc...) that will be interested or influenced by this research; how will this research be used by others; will it be of interest beyond the specific field? This paper is of a broad audience as it treats of large problematic of evolution of plant pathogen.
• Please 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.

I am a population genomicist working on evolution of pathogens.

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

We would like to thank reviewers for their insightful comments.

Overall, there were two major concerns/suggestions:

• Applicability to humans of the increase of BTC in non-alcoholic steatohepatitis (NASH) and mechanisms of downregulation of BTC by omega-3. We now analyzed __3 __additional human gene expression datasets and show that BTC not only is increased in human NASH (as we have already shown for liver cancer meta-analysis), but is also decreased in livers of patients who received omega-3.

• One of the reviewers suggested investigating a potential mechanism of how BTC is regulated by omega3 fatty acids. Although a complete answer to this question would require entirely new studies to be done, we still performed additional investigation that was possible within a reasonable timeframe. We found that transcription factor FOXO3 (well-known inhibitor of carcinogenesis) is a highly probable mediator of the DHA inhibitory effect on BTC.

See all details of items 1 and 2 as well as answers to other (less critical concerns) below after each specific question.

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

This work by Padiadpu and colleagues investigate the mechanism by which pufa of the n-3 series (mostly DHA) may influence NAFLD progression using systems biology analysis and multiple omics analysis. The work is interesting and may provide a novel view of the topic. However, there are a number of issues the authors may wish to consider in order to improve their manuscript.

Major issues: Clarity: Since the authors refer to previously published experiments, they must refer to this work in the figure legends and improve the clarity of such legends. Here are a list of issues that must be fixed:

Fig.1: First panel is not clear. What does the table tell the reader? What are the effects of the different diets on NAFLD?

All the transcriptomic data are newly generated from the samples of previously published studies. The table shows the number of features changed by DHA and/or EPA in each of the -omics and phenotypic data used in the analysis.

I understand that the results are published elsewhere, but the authors must provide information regarding the NAFLD/ NASH scores.

We now added a supplementary table 1a showing the scores.

Fig.4: Why is there sometimes a DHA diet, sometimes DHA and EPA. Legend is not clear. What does WD + Mean? I guess it is olive oil... But the legend must be improved.

We added details in the legend for more clarity. Specifically, WD+O means WD + olive oil added as a control for WD+DHA, WD+EPA. As described in the 2nd paragraph of results, when both EPA and DHA had a similar and significant effects in reversing WD effect, it was defined as “EPA&DHA category” of parameters. When only WD+DHA or WD+EPA were significantly changed vs WD+O, those were assigned as “DHA category” or “EPA category”, respectively.

One issue the authors may consider trying to fix is the specificity of the effect of DHA on BTC.

Is it really specific? It seems to me that EPA has more or less the same effect. If the effect is DHA-specific, than make this clearer through the text.

Although BTC expression was reduced by both DHA and EPA comparing to WD, DHA had a statistically significant stronger effect than EPA (Fig. 3D).

Another issue the authors may wish to investigate is the relationship between W3 consumption and BTC expression in studies performed by other labs (if available on Gene expression omnibus?).

Thanks for the suggestion. We used publicly available data of human and mouse studies that showed significant increase in liver BTC gene expression in NASH in multiple datasets while a human trial with Omega 3 treatment for one year showed its significant reduction (Figures 3F - human data, S3G-mouse data).

Finally, a key issue would be to identify the mechanism by which DHA inhibits BTC expression? How does this happen? could such inhibition be induced by other fatty acids of the W3 series? I understand that this is not easy to address but it would significantly strengthen the manuscript.

Thanks to your question we investigated and found at least one of potential mechanisms contributing to how “DHA inhibits BTC expression”. See details in the answer to next question. As for “other fatty acids” while we agree this is important question, it is outside of the scope of the current study but will be investigated in future studies.

Moreover, it might be possible to identify the set of genes highly co-regulated with BTC expression and to investigate the possible transcription factors at play in the control of such gene set.

We really appreciate this question as our efforts in this direction provided one potential mechanism. A direct screen of transcription factor (TF) motifs in genes co-regulated with BTC did not provide any clear results. Therefore, we implemented a combination of network analysis and screen for motifs in BTC gene with the in vivo and in vitro treatment results and found FOXO3 as a candidate TF regulated by DHA upstream of BTC.

See details of the analysis and results in a new Supplementary Figure S6 and corresponding text located at the end of the results.

Minor: the authors use the term "beneficial" transcriptome alterations by DHA.

I do not think it is correct to use "beneficial".

We agree and removed the word "beneficial”.

Reviewer #1 (Significance (Required)):

Strength: This paper uses new approaches to investigate the relationship between W3 consumption and liver gene expression and its relevance to chronic metabolic liver diseases.

The experiments and data set used to perform systems biology are from an excellent lab (the authors lab) who has published a lot of important and reproducible discoveries in the field of regulation of gene expression by dietary fatty acids.

The work has high translational relevance in medicine / hepatology / metabolism.

I am not a qualified reviewer to assess the systems biology that has been done.

Limitation: The mechanistic link between DHA consumption and BTC expression is not very clear. The specificity of this effect could also be tested (DHA vs other W3 and/or W6).

Although BTC expression was reduced by both DHA and EPA comparing to WD, DHA had a significantly stronger effect than EPA (Fig. 3D). Other omega fatty acids were not tested but it can be done in future studies.

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

The authors files a manuscript describing the impact of the suppression of betacellulin as a key mechanism to counteract fibrosis and inflammation in NASH by modulating fatty acids in WD-fed mice.

Major Comments: (i) No histological analysis was presented and indeed this is of clinical relevance for NASH since diagnosis is still based on biopsy.

While histological evaluation was presented in the originally published papers (PMID: 28422962, 23303872), it is now provided in Supplementary Table S1a.

(ii) Human comparative analysis: is done with HCC not with NASH patients.

This cancer-related dataset is most likely obtained from different etiologies.

I would suggest comparing these mouse datasets with GSE48452 (human NAFLD-NASH spectra).

Thanks for this important question. We now analyzed available human data of NASH and show significant increase of BTC expression in two datasets while a human trial with omega-3 treatment for one year showed its significant reduction of BTC expression (Figure 3F) resembling our observations in mice.

(iii) to compare the inflammation and fibrosis (also lipid metabolism), one can compare these mouse datasets with GSE222576 and cite this preprint (https://doi.org/10.21203/rs.3.rs-2009380/v1)

Using the suggested dataset (of a chemically induced liver fibrosis), we first observed that Btc gene expression was significantly increased over 10 weeks of the model and now included this result in Fig. S3G.

We also queried the 66 genes from the network modules described by the authors to check their changes in our NASH model. We observed that 28 genes were differentially expressed in NASH with 14 of them belonging to the module that authors named as “Pathways in Cancer”. Other genes were from the lipid metabolism (4 genes), immunity (2) and inflammation (2 genes). In addition, we observed that several genes we found regulated by omega-3 and changed in this fibrosis model contained other inflammatory genes such as classical macrophage genes (Mmp12, Lgals3, Cd68, Trem2), fibrosis (Col4a1, Col27a1, Itga2b, Itga8) and lipid metabolism (Scd2, Lpl, Soat1). Of note, the preprint has been published and we now cite the corresponding article.

Minor comments:

(i) The heatmap in Figure 1B and another heatmap should show all mice not the average to see the variability

The supplementary figure with all the individual mouse data as another heatmap is added to show the variability and similarity (Figure S1D).

Reviewer #2 (Significance (Required)): The authors files a manuscript describing the impact of the suppression of betacellulin as a key mechanism to counteract fibrosis and inflammation in NASH by modulating fatty acids.

This is well designed experiment, and the results are of interest to hepatologists and should be indeed published after consideration of the following points

Strength is multiOMICs approach.

Weakness is human applicability.

We improved human applicability by investigating 3 additional human datasets of NASH (Fig. 3F) and finding consistent changes in BTC expression closely resembling our observations in mouse NASH model, including one trial with omega-3 treatment of patients for one year showing significant reduction in BTC gene expression.

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 authors files a manuscript describing the impact of the suppression of betacellulin as a key mechanism to counteract fibrosis and inflammation in NASH by modulating fatty acids in WD-fed mice.

Major Comments:

• (i) No histological analysis was presented and indeed this is of clinical relevance for NASH since diagnosis is still based on biopsy.
• (ii) Human comparative analysis: is done with HCC not with NASH patients. This cancer-related dataset is most likely obtained from different etiologies. I would suggest comparing these mouse datasets with GSE48452 (human NAFLD-NASH spectra).
• (iii) to compare the inflammation and fibrosis (also lipid metabolism), one can compare these mouse datasets with GSE222576 and cite this preprint (https://doi.org/10.21203/rs.3.rs-2009380/v1)

Minor comments:

• (i) The heatmap in Figure 1B and another heatmap should show all mice not the average to see the variability

Significance

The authors files a manuscript describing the impact of the suppression of betacellulin as a key mechanism to counteract fibrosis and inflammation in NASH by modulating fatty acids. This is well designed experiment and the results are of interest to hepatologists and should be indeed published after consideration of the following points

Strength is multiOMICs approach

Weakness is human applicability

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 work by Padiadpu and colleagues investigate the mechanism by which pufa of the n-3 series (mostly DHA) may influence NAFLD progression using systems biology analysis and multiple omics analysis.

The work is interesting and may provide a novel view of the topic.

However, there are a number of issues the authors may wish to consider in order to improve their manuscript.

Major issues:

Clarity:

Since the authors refer to previously published experiments they must refer to this work in the figure legends and improve the clarity of such legends. Here are a list of issues that must be fixed: Fig.1 : Firts panel is not clear. What does the table tell the reader? What are the effects of the different diets on NAFLD? I understand that the results are published elsewhere, but the authors must provide information regarding the NAFLD/ NASH scores. Fig.4: Why is there sometimes a DHA diet, sometimes DHA and EPA. Legend is not clear. What does WD + Mean? I guess it is olive oil... But the legend must be improved.

One issue the authors may consider trying to fix is the specificity of the effect of DHA on BTC. Is it really specific? It seems to me that EPA has more or less the same effect. If the effect is DHA-specific, than make this clearer through the text. In this current version of the manusript, the authors alternatively use the term DHA or W3. Related to this issue, it would be nice to know what the composition of the WD is? More specifically, it would be important to know whether it might be W3 deficient.

Another issue the authors may wish to investigate is the relationship between W3 consumption and BTC expression in studies performed by other labs (if available on Gene expression omnibus?).

Finally, a key issue would be to identify the mechanism by which DHA inhibits BTC expression? How does this happen? could such inhibition be induced by other fatty acids of the W3 series? I understand that this is not easy to address but it would significantly strengthen the manuscript. Moreover, it might be possible to identify the set of genes highly co-regulated with BTC expression and to investigate the possible transcription factors at play in the control of such gene set.

Minor: the authors use the term "beneficial" transcriptome alterations by DHA. I do not think it is correct to use "beneficial".

Significance

Strenght:

This paper uses new approaches to investigate the relationship between W3 consumption and liver gene expression and its relevance to chronic metabolic liver diseases. The experiments and data set used to perform systems biology are from and excellent lab (the authors lab) who has published a lot of important and reproducible discoveries in the field of regulation of gene expression by dietary fatty acids.

Limitation:

The mechanistic link between DHA consumption and BTC expression is not very clear. The specificity of this effect could also be tested (DHA vs other W3 and/or W6).

The work has high translational relevance in medicine / hepatology / metabolism.

I am not a qualified reviewer to assess the systems biology that has been done.

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

Manuscript number: RC-2023-01919

Corresponding author(s): Fumio, Matsuzaki and Quan, Wu.

[The “revision plan” should delineate the revisions that authors intend to carry out in response to the points raised by the referees. It also provides the authors with the opportunity to explain their view of the paper and of the referee reports.

The document is important for the editors of affiliate journals when they make a first decision on the transferred manuscript. It will also be useful to readers of the reprint and help them to obtain a balanced view of the paper.

If you wish to submit a full revision, please use our "Full Revision" template. It is important to use the appropriate template to clearly inform the editors of your intentions.]

General Statements [optional]

This section is optional. Insert here any general statements you wish to make about the goal of the study or about the reviews.

We thank two reviewers very much for their comments. Their comments greatly contribute to our revision plan. Reviewer 1 fairly evaluated our data and provided us constructive and supportive comments. We incorporated responses to Reviewer 1’s comments to our revision plan, in which we made some novel analyses and discussions according to Reviewer1’s comments. Reviewer 2 also provided us very helpful comments, which are based on his/her careful reading of our manuscript, especially from the viewpoints of a ferret specialist. These comments help us to improve our manuscript very much, whereas some of the reviewer 2’s requests appear beyond the scope of our paper and against the policy of Review Comments; the standard policy of the review for Review Commons is “do not add new pipeline of experiments” such as adding additional replicates for scRNAseq. We have made revision plans (section 2) according to the order of comments given by reviewer 1 and then next by reviewer 2, considering all the statements of the two reviewers on balance; there are 6 comments from reviewer 1, and 25 comments from reviewer 2. In the section 2, we selected revision plans that we have reflected to the preliminary revision of our manuscript.

Finally, we would like to note our fundamental interest; we are studying the cortical development of ferrets as a model of brain development to understand what mechanisms are conserved or species-specific during brain size expansion in the mammalian evolution, which, of course, includes humans. It would be great if the ferret model can be a tool used to study tRG cell biology, contributing to understanding the human cortical development.

For this purpose, it’s been critical to create series of single cell transcriptomes along cortical development. A comparison between humans and ferrets, focused in this paper, is the first attempt, because human data of single cell transcriptomes have been extraordinary enriched. These attempts of comparisons between ferrets and humans will provide valuable information about which mechanism is shared and which is not shared for the cortical development in the gyrencephalic mammals. To represent the usefulness of our approach, we chose finding of tRG in ferrets as a symbolic example, and analyzed its origin and fates.

Description of the planned revisions

Insert here a point-by-point reply that explains what revisions, additional experimentations and analyses are planned to address the points raised by the referees.

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

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.

- 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.

1-1. 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.

We thank the reviewer for the suggestion to revise the definition of tRG parent cells in lines 194-204. This issue is also pointed out by the reviewer 2.

Revision plan:

1. We revise the term “IPC” as “mitotic sibling of vRG” and stated that these cells might be tRG (CRYAB+) or non-tRG (CRYAB-) intermediate progenitors. By the term of “intermediate progenitors”, we did not intend to refer to TBR2+ neurogenic IPCs, but rather to an intermediate state of progenitors, in a general sense, with a similar morphology as tRG. To avoid any confusions on this terminology, we revised our manuscript by replacing “IPC” with “a sibling of vRG”.
2. We delete all statements relevant to Tsunekawa et al. data from the manuscript. We regret that we are not able to include Tsunekawa et al. data because we are planning to submit this data as a separate manuscript, which describes that in ferrets, vRG frequently (30% of apical division) generate non-Tbr2-positive mitotic sibling cells bearing a short basal process during the entire neurogenesis. This study includes a large volume of data with human ones and largely concerns stages that are earlier than that of tRG formation. It is, therefore, not appropriate to combine these data with those described in this manuscript.
3. As also pointed by reviewer 2, we cannot exclude the possibility that the mitotic sibling cells of vRG with a short basal process (IPC in the previous version of the manuscript) are also CRYAB positive tRG. To clarify the identity and variety of vRG sibling cells at tRG-generating stage, we are examining the sibling pairs of vRG by immunostaining for a mitotic marker Ki67 and CRYAB during P0-P5 after incorporating EGFP by electroporation to label vRG lineages. We will increase the sample size for a quantification and statistical analyses of this newly provided data to incorporate in our fully revised manuscript.

1-2. 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.

We agree with the Reviewer 1 that identifying the presence of astroglial and ependymal tRGs in datasets spanning later developmental stages would provide convincing evidence for the potential of human tRG.

Revision plan:

1. We compared our ferret dataset to the human postnatal dataset recommended by the Reviewer 1 (Herring et al., 2022). As a conclusion of our analyses shown below, we found that Herring et al., (2022) dataset was not favorable for a comparative analysis with our ferret dataset regarding the fates of human tRG, because Herring’s human dataset was derived from the prefrontal cortex; This human dataset does not include neither tRG cell population nor ependymal clusters. We have also elaborated our discussion after analyzing Herring et al. dataset in the discussion.
2. We, therefore, pare down our claim in lines 365-366, by removing “(and presumably human)” to state that “Our pseudotime trajectory analyses and immunohistochemistry analyses strongly suggested that…”.
3. We also tone down the statements as for the discussion of the relationship between human and ferrets regarding the tRG progeny fates (originally lines from 372 to the end) also elaborated our discussion after analyzing Herring et al. dataset in the same paragraph.

We will describe the details of our analysis of Herring et al. (2022) below.

https://www.dropbox.com/scl/fi/a0m72orxfsub66dh3hdbg/reviewer1_2ABC.pdf?rlkey=uzrd8ngclp87p5c8v24mqd1j7&dl=0

As mentioned above, Herring’s human dataset was derived from the prefrontal cortex, and that it did not include a specific subtype defined as tRG nor other HES1-expressing progenitor clusters such as RG in the original cluster annotation. We, therefore, re-clustered the raw dataset from GW22 (the earliest stage available) up to 10-months after birth by using Seurat pipeline with default parameters (B), and found a CRYAB-expressing population in the original “Astrocyte_GFAP” subtype among astrocyte clusters (A), which distribute in the most of collected stages, from late development through the adulthood. We then examined this dataset to find out whether tRG or its progenies are present.

After reclustering, CRYAB-expressing cells (with more than 1 raw count) represented 0.15% of the dataset and were grouped as a part of cluster 44, which was mostly derived from postnatal stages (among which 4-months was the most enriched one; C). Several astrocyte markers, such as SPARCL1, HOPX, CLU, and GJA1, as well as CRYAB, were enriched in the cluster 44 as revealed by FindMarkers (Methods). FOXJ1 expression was nearly absent overall in this dataset, indicating the absence of the ependymal cell population, a tRG-descendant cell types in ferrets (C).

To evaluate the similarity between cluster 44 and tRG or astroglial tRG, we next integrated Herring dataset with our ferret subset (about 15,000 cells) and the human GW25 subset from Bhaduri et al. (2021) of approx. 3,000 cells, both of which contained only progenitor cells. As we have done in Figure 7 of our original manuscript; we have removed cells other than progenitors, astrocytes and oligodendrocytes, such as neurons, microglia, endothelial cells. This resulted in about 20,000 cells in Herring dataset.

https://www.dropbox.com/scl/fi/nz3iulya5199i95ecr1un/reviewer1_2D.pdf?rlkey=kp7lwxtkn562un1uf9l1axn2p&dl=0

This integration (D) reveals that Herring’s cluster 44 is closely located to Bhaduri’s human and our ferret tRG clusters on UMAP, but does not overlap with these tRG clusters. This result further suggested that tRG population might be lacking or very rare in this neuron- and glia-dominated dataset, which might be due to the sampling method that targeted the enrichment of neuronal layers (Herring et al., 2022). It is also possible that this fragmented information on astrocyte and ependymal lineages could be due to the regional and/or temporal difference of samples between two human datasets.

1-3. 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.

We absolutely agree that human organoids are good models to study human brain development.

Revision plan:

According to the suggestion of reviewer 1, we analyzed two cortical organoid datasets (Bhaduri et al., 2020; Herring et al., 2022) to examine whether different tRG populations are present in organoids. Our analyses led us to conclude that tRG-like populations seem to be lacking in available organoid datasets; organoids can have CRYAB-expressing astrocyte-like cells in single-cell transcriptome datasets, but the presence of tRG-like cells seem to be unstable and dependent of lines and protocols how organoids are generated. A further assessment on tRGs’ cellular features is required on organoids by immunostaining experiments. In the light of this analysis, we elaborated our discussion by describing observations shown below. Below is our analysis of organoid data.

Bhaduri dataset contained organoids generated from 4 different lines, which showed a variability in terms of cell distribution on UMAP while overall temporal and differentiation axes were recapitulated (A). While keeping the original cluster annotations except for YH10 line, we performed reclustering. CRYAB was expressed in clusters 26 and 30 enriched in YH10 line, and cluster 29 enriched in 13234 line (B).

https://www.dropbox.com/scl/fi/8mj6u94t3hkzw6q61o7od/reviewer1_3AB.pdf?rlkey=10xiks25nzn9r90guw9l0onqh&dl=0

To confirm the identity of these clusters, we integrated organoid dataset with the dataset of primary tissues from the same paper (Bhaduri et al., 2020; C).

https://www.dropbox.com/scl/fi/qnqv2e87t74uom2pg836d/reviewer1_3CD.pdf?rlkey=mv370b3dlogwvgh6ig8bdathp&dl=0

As a result of the integration, tRG cells from the primary tissue were not overlapped with organoid-derived CRYAB-expressing cells, although a part of CRYAB-expressing organoid cells were localized in the integrated cluster 16 where primary tRG resided (D). Other cell types that were included in the integrated cluster 16 were “lateRG”, “vRG”, “oRG” from primary tissue dataset, and “glycolyticRG” from organoid dataset. We found that CRYAB-expressing organoid clusters 26 and 30 overlapped with “oRG/astrocyte” clusters of primary tissues.

Furthermore, we have analyzed another organoid dataset in stages including 5-months, 9-months and 12-months (Herring et al., 2022; E), but found no clusters that specifically expressed CRYAB (F).

https://www.dropbox.com/scl/fi/b4kiqoqyhhzk4vm5hi1bb/reviewer1_3EF.pdf?rlkey=dd00hju5n4b90wpz2zexi9gxa&dl=0

1-4. 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.

We thank the Reviewer for appreciating the difficulties associated with isolating and sequencing rare cell types. We were able to identify a total of 409 tRG cells (in tRG-like cluster) after merging all timepoints of sequencing, (Figure 1C, S3C) as stated in line 162 of the original manuscript. However, to perform pseudotime analyses, we subset our dataset using 6,000 cells in total (excluding neuron and non-progenitor clusters; Methods), which included 162 tRG cells. Pseudotime analysis transcriptomically distinguished tRG into 3 subgroups (Figure 4E). Remaining 247 tRG cells also appear to distribute similarly into these subgroups rather than forming a distinct subregion within tRG cluster (right panel in figure below). Furthermore, we conducted extensive immunohistochemical analyses of tRG-like cells, and we found that both the morphology and gene/protein expression were consistent with the notion that “tRG-like” cluster in our ferret dataset represents tRG defined in humans (Nowakowski et al., 2016).

Revision plan:

As for human dataset, we agree that the population of committed tRG was minor. Thus, we pared down our statements regarding the fates of tRG as mentioned in other comments, both in the Results and Discussion.

https://www.dropbox.com/scl/fi/aqsg5xlbxyoybzwq0xezp/reviewer1_4.pdf?rlkey=oxhmtko08nhvzkmsqxcjf9qua&dl=0

- Are prior studies referenced appropriately?

1-5. 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.

We thank the Reviewer for bringing up this issue. This is an important issue because we wanted in this study to use the ferret as a good model for the complex brain development in gyrencephalic animals, in general, to know what characteristics are shared or not, across gyrencephalic species (such as the presence of the OSVZ vs. the temporal scale).

Revision plan:

Our study demonstrated the presence of tRG cells up to P10 by immunohistochemistry and scRNA-seq. P5~P10 is the stage where neurogenesis became dominated by gliogenesis in the dorsal cortex in ferrets, although its timing is delayed in the visual cortex. On the other hand, Nowakowski et al. (2016) originally identified and defined CRYAB-expressing tRG, based on morphology and gene expression on human primary tissues during mid-neurogenic stages, while cortical neurogenesis is mostly declined in human postnatal stages. We have failed to find literatures or textbooks describing the presence of CRYAB-expressing tRG, while an ependymal layer was detected in the postnatal human cortices (Honig et al., 1996; preprint Nascimento et al., 2022). At the moment, the lack of information thus makes it difficult to compare the relationship of birth timing with the period of tRG persistence between ferrets and humans. In the revised manuscript, the “Discussion” will include this argument as well as the following difference between humans and ferrets in the RG scaffold.

Besides birth timing, Nowakowski et al. also reported that radial glia scaffold spanning from the VZ to the pial surface undergoes a transformation during neurogenic stages; tRG becomes the major RG population in the VZ, disconnecting VZ and OSVZ. In contrast, we did not find a discontinuous scaffold stage over the course of ferret neurogenesis. Instead, we still detected CRYAB-negative vRG with an apical attachment and a basal process extending beyond the OSVZ during stages where the peak of tRG expansion is achieved (such as P5 in Figure 2A, S3A). This appears to be a prominent difference between human and ferret corticogenesis.

- 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?

1-6. 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.

Revision plan:

We prepared the images for GFP+/CRYAB- vRG cells in an adjacent panel in Figure 2A as recommended by the reviewer (below). To better distinguish the morphology of an isolated vRG cell from other labelled cells, we sparsely labeled RG cells with EGFP at P3 by electroporation (Methods), and fixed the samples two days later (right panel). We highlighted the morphology (cell body and basal fiber) of a CRYAB- GFP+ vRG and that of a neighboring CRYAB+ GFP- tRG on the same panel to clarify that vRG did not express CRYAB.

https://www.dropbox.com/scl/fi/3wrmqdswt69t8pkdy30h7/reviewer1_6.pdf?rlkey=90ixbadan3mxx10m85jnpwphn&dl=0

Reviewer #1 (Significance (Required)):

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.

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

2-1. 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.

We would like to thank the reviewer for carefully reading our manuscript and providing us with valuable feedback. However, we would like to clarify that there might have been a misunderstanding regarding our conclusion about the identification of oRG-like cells in ferrets.

Our study did not conclude that we have identified oRG cells in ferrets with “a quite different transcriptomic profile than in human”. Instead, our findings indicate that unlike oRG cells in human, ferret oRG-like cells did not exhibit specificity for human oRG markers (such as HOPX and CLU) that would enable us to distinguish them from other late RG cells in ferrets. Despite this, oRG score derived from human oRG marker expression showed higher values in predicted ferret oRG-like cells compared to other ferret RG cells, reflecting a similarity of the transcriptome profile between human oRG and ferret oRG-like cells (Figure 7H-I). We will carefully describe our methodology to reach this conclusion in response to reviewer 2’s comment regarding how we determined ferret oRG in a later comment.

Major issues:

2-2. 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.

The reviewer 2 seems to misunderstand that we took cortical strips shown in Figure S1A as samples for scRNA seq. If our description in the main text is confusing, that would be our fault.

In Figure S1A of the original manuscript, we showed the cropped images of the medial part to emphasize the distinguishment of different germinal layers (VZ/iSVZ/oSVZ) and their temporal changes in ferret cortices.

Revision Plan. To avoid such a misleading, we inserted the dotty lines in the revised Figure S1A to demarcate the tissue parts for scRNAseq, which correspond to almost all lateral cortices, mainly including the somatosensory area 1 and 2 with surrounding areas. We accordingly added the following sentence in the legend, “The approximate boundaries of dorsal cortex area used for scRNA sequencing are highlighted with dotty line segments in the dorsal cortex hemisphere above each strip.”.

We also show actual sampling for single-cell transcriptomics below. As our sampling was not restricted to the somatosensory cortex, we have revised “somatosensory cortex” as “dorsal cortex” in Lines 131 and 1191 of our manuscript.

https://www.dropbox.com/scl/fi/9gg508iood73zl02836g6/reviewer2_2.pdf?rlkey=lufevala88ihvc1p6mts463as&dl=0

2-3. 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.

We disagree with the reviewer 2’s comment. We would like to clarify that we collected brain tissues in two different ways for the same set of developmental stages; one brain tissue by removing cortical plate (T); another independent brain tissue at the same developmental stage by sorting GFP-labelled lineage from neural progenitors that were electroporated at embryonic stages (AG, Methods). Both manipulations of samples aimed to increase progenitor cell populations in scRNAseq. Therefore, we have two sets of samples of the same temporal series, each prepared in a totally different way. All cell types were present in both methods of collection shown as Supplementary Figure 2E’ (below left) that separates samples by different preparations at each stage (by modifying Supplementary Figure 2E; below right). We believe that the biological replica (n=2) in this manuscript would be sufficiently reliable, judged by its reproducibility.

https://www.dropbox.com/scl/fi/levyqy9ngvpyio1yl9oif/reviewer2_3.pdf?rlkey=r4aw0hu9cdn68f1pvhp734vxx&dl=0

Here, we also cite several examples of papers important in the field of single-cell or bulk transcriptomics of brain tissue, where only a single replicate or pair (replica) was taken for experiments on mice, humans and ferrets:

mice: Ogrodnik et al., 2021 PMID: 33470505, Hochgerner et al., 2018 PMID: 29335606, Joglekar et al., 2021 PMID: 33469025;

human: Herring et al., 2022 PMID: 36318921, Polioudakis et al., 2019 PMID: 31303374, Mayer et al., 2019 PMID: 30770253, Fietz et al., 2012 PMID: 22753484;

macaque: Schmitz et al., 2022 PMID: 35322231;

ferret: Johnson et al., 2018 PMID: 29643508.

2-4. 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.

The reviewer pointed out an important criterion, the abundance of mitochondria.

Revision Plan:

We have now added the mitochondrial QC metrics in the new Figure S2A, and revised the legends as follows: “Violin plots showing the number of genes, mRNAs and the percentage of mitochondrial genes per cell in each sample and time point”. We have computed the percentage of mitochondrial genes for each cell type and found that the majority of cells in each cell type had a value less than 5% while the content value in some cells distributed along the range between 0% and 10%, up to a maximum of 28% (Figure S2A). Despite this, we have decided to include all cells that had less than 30% of mitochondrial genes in our analysis based on the percentage of reads mapped on mitochondrial genome for the following reasons:

1. The percentage of mitochondrial indicates respiratory activity, rather than apoptosis and the percentage of mitochondrial quite depends on the tissue type and species. For example, in human case, such percentage range from 5%~30% (Mercer et al., 2011 Cell; The human mitochondrial transcriptome).
2. Unfortunately, unlike human and mouse brains, there is no reference to show the percentage of mitochondrial in ferret brains. Therefore, the suitable way is to keep all of these cells.
3. These cells showing high percentage of mitochondrial genes are not clustered as an apoptosis cluster in UMAP, instead, these cells are observed in most of clusters (below). Therefore, we believed that these cells are not apoptotic cells and include these cells in further analysis.

https://www.dropbox.com/scl/fi/4kp3fczxzo6x4fx8hqt8m/reviewer2_4_1.pdf?rlkey=ypojzbuwgelt51qlf56g883s9&dl=0 4. After all, we have obtained similar clustering overall after filtering cells with a higher percentage for mitochondrial genes; we set the threshold to 10%. This filtering resulted in 28,686 cells in our dataset. We then performed our workflow from the normalization step with the same settings that we applied to our original ferret dataset (Methods). Below, we show the results comparing newly generated clusters in this filtered subset on UMAP (left), and the original clusters shown in Figure 1B (right). 26 clusters were obtained in both conditions, and both major cell types and subtypes were conserved after filtering.

https://www.dropbox.com/scl/fi/0mlk69z7hckpiw03ivfjb/reviewer2_4_2.pdf?rlkey=hfvjrifrytmnywc4vchjvf0ms&dl=0

Clustering resolution: Our choice of the resolution was based on avoiding over- or under-clustering of ferret cells. After trying several resolution values, including 0.6, 0.8, 1.0 and 1.2, we have decided to use the resolution of 0.8 as the separation of cell types was the most reasonable among other resolutions that we have tried, in a similar way to actual known cell types. For example, the resolution of 0.6 did not distinguish “tRG” cells from “late_RG1” cells, as well as “early_RG” subtypes which were distinctly enriched with different cell cycle markers (Figure S2D). On the other hand, the resolution of 1.2 resulted in an over-clustering of IPC, OPC, DL neurons and microglia.

2-5. 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.

Revision plan:

We focused on Figure 2A and S3A (2D is a histogram) to show the full morphology of CRYAB+ tRG, because Figure 2A is the initial presentation of tRG in this paper, and Fig. 2A and Fig.S3A images are taken on a 200-micrometer thick section, originally aiming to indicate that CRYAB-positive fiber is short, spanning nearly along the VZ and the SVZ. We made 3D-reconstructions of those images, which are rather better than orthogonal projections, in order to show that CRYAB+ fibers are shorter than those of vRG (terminating at positions around the upper boundary of the SVZ) and that the short basal processes are not due to the cut of long radial fibers during tissue sectioning.

We will show these 3D-reconstruction below. Please download movie files from the following URLs to look at them clearly.

Figure 2A

https://www.dropbox.com/s/qocve596c5xhtlc/%E2%98%85fig2A-Ver02.mp4?dl=0

Figure S3A

https://www.dropbox.com/s/v8gqwfi1r8ff5n5/%E2%98%85figS3A-P0%20movie-ver2.mp4?dl=0

2-6. 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.

We thank the reviewer for appreciating our time-lapse imaging data as very nice and a technical challenge. The number of time-lapse imaging that could follow the cell fates was from “4” samples instead of 3. It is indeed very infrequent and difficult to obtain a complete set of consecutive divisions from vRG, followed by histochemical examinations (fixation, cryo-sectioning and immunostaining of slices). This is because some of EGFP-labeled cells are frequently indistinguishable from each other by overlapping within a clone or with cells in other clones. Therefore, we decided to take a different way to clarify the pathways from vRG and its variety to generate tRG at the tRG-generating stage.

Revision plan:

Increasing the number of time-lapse image series will be extremely inefficient because of the reasons described above, perhaps taking a long time such as 3-5 months according to our breeding schedule of ferrets. Therefore, we take an alternative way to clarify the division patterns from vRG to generate tRG, especially focusing on the identity and variety of vRG sibling cells at the tRG-generating stage; we are examining the sibling pair of vRG and/or precursor of tRG to see what kind of cell the vRGs actually generate at their mitosis. For this purpose, we electroporate ferret cortices with the EGFP-expressing plasmid approximately one cell cycle prior to fixation (E38 or P0). We then stain ferret cortices for a mitotic marker Ki67 and tRG marker CRYAB and other markers during the tRG-generating state (P0-P5), assuming the cell cycle length of vRG and IPC as approximately 33h~45h based on our own consecutive EdU labeling experiments and time-lapse imaging.

2-7. 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.

We appreciate the reviewer 2’s question about “why is this identified as IPC, when it looks exactly like vRG, NOT as an IPC?”

Revision plan.

1. We are confident that this blue-labeled cells in Figure 3A and D are not vRG but mitotic sibling cell (of vRG) with a short basal fiber (that we named IPC in the initial manuscript). We now made the morphological features of these cells clearly visible by constructing 3D-views of the images with different snapshot images (we show below and in the preliminary revision as a supplementary movie). In addition, it divides once as time-lapse imaging revealed, hence this cell is still mitotic, instead of a postmitotic cell. Therefore, we used the term that is generally used for this type of cells, namely, intermediate progenitor cells (IPC), by which we did not intend to refer to TBR2+ neurogenic IPC. We plan to include these revised images into our fully revised manuscript.
2. We agree the reviewer 2 on the point that this blue-labeled cell may express CRYAB (the next comment of reviewer 2 essentially claims the same point), as we also wrote this possibility in line 204-207 of the original manuscript. It could not be technically possible at the moment to examine CRYAB expression in a cell emerging only in the course of time-lapse imaging. If we could label vRG with a transgenic or knock-in fluorescence marker, which mimics CRYAB gene expression, we could have figured out whether blue cells (the mitotic vRG sibling cells) express the CRYAB gene. Indeed, we tried to knock the EGFP gene in the CRYAB gene many times over a year, but have so far failed. Given that tRG is defined as the cell type expressing CRYAB with a short basal fiber at late-neurogenic stage, irrespective of its mitotic activity, this blue labeled vRG sibling cell in Fig. 3A (and/or Fig. 3D) might express CRYAB, hence can be a “mitotic tRG” (although its possibility seems to be low as shown in Fig. 2E). To avoid any possible misleading, we have changed the term of these cells to a “mitotic vRG sibling cell (or mitotic tRG parental cell) with a short basal process”, and add a comment that “this cell might be mitotic tRG with CRYAB expression”.
3. As for the TBR2 expression, we do not know these cells that appeared in the course of time-lapse imaging express TBR2 or not. As shown in Fig. 2F, 10% (P10) to 30 % (P5) of CRYAB+ cells express TBR2. On the other hand, “intermediate progenitors” do not necessarily express TBR2 in general. Therefore, we disagree on the reviewer 2’s comment “their analyses in Fig 4A contradict their interpretation on tRG’s parent cells”, but “our analyses in Fig 4A is compatible with our interpretation on tRG’s parent cells in time-lapse imaging”, and that is “a mitotic vRG sibling (or mitotic tRG parental cell) with a short basal fiber divides to produce CRYAB+ tRG at the end of timelapse imaging”. However, to avoid any overstatements or misunderstanding on this issue, we have revised related text as described above.
4. We are not able to include the data taken by Tsunekawa et al.. This is because we are going to submit a separate paper, which includes a large volume of data with human ones in collaboration with another group and largely concerns stages that are earlier than that of tRG formation. It is, therefore, not practical to combine these data with those described in this manuscript. Therefore, we remove all descriptions related with Tsunekawa et al.

Below we show snapshot images and 3D-reconstructions for Figure 3A and 3D. Please download movie files from the following URLs to view at them at the highest resolution.

@Figure 3A:

1)A time lapse movie (20 min interval) showing images around time 40:00 at which vRG underwent the second division.

https://www.dropbox.com/s/znx3bboxefhj0jt/%E2%98%85Fig_3A%20movies%20around%2040%20h.mp4?dl=0

2)Snapshot images for time 40:00

https://www.dropbox.com/s/6y25mk4jhwqy6v7/%E2%98%85E38-fig3A-sRG-2.png?dl=0

3) 3D-reconstruction images at the same time point (40:00)

https://www.dropbox.com/s/so8hesjzy63yxmb/%E2%98%853D-reconstruction%20%2840.00%202nd%20div%29.mp4?dl=0

4) The entire time-lapse movies of time 0:00-84:00; The mitotic sibling cell of the vRG is indicated by a white arrow.

https://www.dropbox.com/s/ywua95f8fmohsmc/%E2%98%85Fig3A-arrow-time.mp4?dl=0

@Figure 3D:

A revised time-lapse snapshots of Figure 3D.

https://www.dropbox.com/s/xyet4virt3j9u3t/%E2%98%8520211220%EF%BC%8DP0%EF%BC%8Dtimelaps-xt04corrected.psd?dl=0

The assignment of the cell has corrected to the right one for the same mitotic cell because cell body position at the first two time points were misassigned in the original manuscript (at the following time points, there is no change).

Snapshot image at time point of 06:20; https://www.dropbox.com/s/hn3v6ao1qkhnfjh/%E2%98%85Fig3D%20sRG%20at%200620.png?dl=0

Rotating movie of 3D-reconstruction at time point of 06:40:

https://www.dropbox.com/s/6taqjr0u21x5tn0/%E2%98%853Drotated%20movie%20of%20time%20point%2006.40.mp4?dl=0

2-8. Alternative interpretation of time-lapse images (lines 196-197): maybe a tRG can generate one tRG CRYAB+, and one IPC CRYAB-.

We agree with reviewer 2 that there is an alternative interpretation of cell identity appearing in time-lapse imaging of Fig. 3. In line 196-197, we wrote that “These mother IPC underwent an asymmetric division to generate a non-CRYAB expressing cell and a CRYAB__+ tRG”. As pointed by reviewer 2 here and in the previous comment, we cannot exclude the possibility that this vRG sibling cell may be a mitotic tRG (see our response to the previous reviewer 2 comment). If so, what we observed in Fig. 3A and D could be interpreted as a mitotic tRG, and generate one CRYAB+ tRG and one CRYAB- climbing cell. However, as we haven’t confirmed or stated whether this parent cell was a mitotic tRG, we also did not examine the identity of this sister cell of CRYAB+ tRG. It can be an IPC or nascent neuron or even an astroglial progenitor cell. From our data, we cannot say anything about the identity of the CRYAB-negative sister cell other than that this cell is CRYAB-negative, migrating upward. That is why we did not mention about the identity of this CRYAB-negative sister cell of tRG other than that the sister cell of tRG is CRYAB-negative.

Revision plan. We changed the term of IPC to “a mitotic vRG sibling cell” and describe the possibility that “This mitotic vRG sibling cell (or mitotic tRG parental cell) can be a mitotic tRG if this cell express CRYAB, and its apical division generates one tRG and one CRYAB-negative climbing cell with an unknown identity, replacing the description of line 196-197.

2-9. 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.

We thank reviewer 2 for pointing our careless mistake.

Revision plan. We have corrected the shifted position of arrows in Figure 5E. We have removed “mitosis” in the title of Figure 5E since the initial manuscript did not include descriptions on mitosis in the text.

2-10. Line 277: “Transcriptomic trajectories were homologous across the two species”. What does this refer to? What are these trajectories? Pseudotime? Is this statistically tested?

The meaning of the term “Transcriptomic trajectories” was not clear.

Revision plan. We revised our description in this part as “Temporal patterns and variety of neural progenitors during the cortical development were similar to each other between humans and ferrets at the single cell transcriptome level”.

2-11. 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 lines 291-292, we mention that “tRG cells also showed a remarkable similarity between the two species (Fig. 6C, 6D), as represented by CRYAB, EGR1, and CYR61 expression (Fig. 6E)”. Here, what we wanted to claim is that the same combination of gene expression (CRYAB, EGR1, and CYR61) is characteristically at relatively high levels in both ferrets and human tRG. As the reviewer 2 claimed, CRYAB and CYR61 genes are highly selective for ferret tRG among mid-late RG types, while the expression of EGR1 and CYR1 are just relatively enriched in tRG than in other cell types in human RG (except for highly selective CRYAB). Irrespective of the difference in their relative enrichment in tRG between humans and ferrets, one can still state that the combination of these marker expression at higher levels is shared in these two species”. We were not able to find which part in the manuscript was the reviewer referring to for the claimed argument (“EGR1 and CYR61 are expressed selectively in human tRG”).

Revision plan. To clarify our statement, we changed this sentence into “tRG cells also showed a remarkable similarity between the two species (Fig. 6C, 6D), as represented by a high level of expression for the combination of CRYAB, EGR1, and CYR61 (Fig. 6E)”

2-12. 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.

We apologize for a confusion caused by insufficient explanations for our methodology. We want to clarify that we did not find " ferret oRG-like cells as those near human oRGs in the integrated UMAP." Rather, we try to identify oRG-like cells in ferrets based on the hypothesis that, when comparing ferret and human datasets, oRG-like cells in ferrets would exhibit a higher degree of similarity to human oRG cells than to other cell types. This hypothesis was supported by our observations of other clusters such as tRG, later RG, and IPC (Figure 6 C and D).

To identify oRG-like cells in ferrets, we utilized the mutual nearest neighbor (MNN) method to determine the similarity between cells from different species (Stuart et al., 2019 PMID: 31178118). For example, when attempting to identify the human cell that was most similar to a given ferret cell (F), we calculated the distance between cell F and all the cells in the human dataset in the high dimensional expression space. This allowed us to identify a human cell (H) that exhibited the smallest distance to cell F. Subsequently, we computed the distance between cell H and all the cells in the ferret dataset. If cell F had the smallest distance to cell H in the human dataset, we considered cells H and F as a pair of mutual nearest neighbors.

Using this method, we can find all pair of mutual nearest neighbors in two datasets. We then find these pairs that one is human oRG and define the other is oRG-like in ferret. However, upon further investigation of the characteristics of these cells, we would not find any specific markers (such as HOPX and CLU in human oRG) that would enable us to distinguish them from other later RG cells in ferrets.

Accordingly, only when our strategy to find mutual nearest neighbors is suitable, the selected cells can get higher oRG score, otherwise, the selected set of ferret cells will not show a high oRG score. Therefore, we disagree with the notion that “The fact that the selected cells have higher oRG scores is expected and obvious”.

We hope this explanation provides a clearer understanding of our methodology and the rationale behind our approach to identifying potential oRG cells in ferrets.

2-13. 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.

Thank reviewer 2 for his/her precious advice.

Revision plan.

We added several issues discussed in the responses to the reviewers to Discussion. Please look at our responses to comment 2-14 and 2-15 as well as the preliminary manuscript.

2-14. 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.

We appreciate Reviewer’s comment regarding the difference between transcriptomic trajectories and cell lineage. We agree that transcriptomic trajectories do not necessarily reflect cell lineage. However, relationships along transcriptomic trajectories provides useful information about the differentiation potential of cells. Furthermore, in this study, we examined the temporal and spatial relationships between CRYAB+ tRG and FoxJ1+ ependymal cells that were predicted as tRG descendant cells by transcriptomic trajectories. We could confirm an increasing overlap of FoxJ1+cells with tRG cells along the course of post-natal development in Figure 5. We thus accessed the relationship of the two cell types by not only in silico but also in vivo analyses.

Revision plan. We disagree with the reviewer 2 as for ferrets, because we accessed the relationship of tRG and their progeny cells by not only in silico but also in vivo analyses.

On the other hand, as for progenies of human tRG, they were predicted certainly depending on the molecular relationship by comparison with ferrets without histochemical evidence, as pointed by reviewer 2, and the populations of these committed tRG are small. Therefore, we removed “(and presumably also human)” and we tone down about the progeny relationship of tRG as a prediction. We also acknowledge that further studies are needed to confirm the lineage relationships among cell types, as we discussed in the Discussion part.

2-15. 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.

The statement “the notable role of tRG as progenitors contributing to the formation of the ependyma and white matter” is certainly a speculation based on our results, but not experimentally indicated yet by such as gene knockout, as the reviewer pointed out. Although we repeatedly tried to knock out the CRYAB gene in ferrets for a year, we have so far failed.

Revision plan. Taking the comments from reviewer 1 and 2 into account, we largely revised “Discussion” with a more moderate expression, by incorporating comparative analyses with other human datasets, and we also emphasize the importance of in vivo studies as the next step. We just paste the last paragraph of the preliminary revised Discussion. Please see the “Discussion” in the preliminary revision of our manuscript.

“In ferrets, genetic manipulations can be achieved through in utero or postnatal electroporation, as well as via virus-mediated transfer of DNA (Borrell, 2010; Kawasaki et al, 2012; Matsui et al, 2013; Tsunekawa et al, 2016). Thus, it is theoretically possible to disrupt the CRYAB gene in vivo in ferrets to investigate its role in tRG and their progeny, including ependymal cells, and to track the tRG lineage. If the CRYAB gene is essential to form ependymal layers, we will be able to explore how the ventricle contributes to cortical folding and expansion. Despite extensive efforts over a year, we have thus far been unsuccessful in knocking in and/or knocking out the CRYAB gene. Nevertheless, we anticipate that technical advances will surpass our expectations, both in ferret and human organoids. Taken together, these functional studies in ferrets as well as in human organoids hold promising insights into the understanding of the tRG lineage and its contribution to cortical development in the near future”.

Minor issues:

2-16. 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.

Revision plan. We removed the mentioned statement from our manuscript and revised lines 58-59 as follows: “In many mammalian phylogenic states, cerebral cortex evolved to gain an additional germinal layer (Smart et al. 2002; Zecevic et al. 2005; Kriegstein et al. 2006; Reillo et al. 2011)”.

2-17. 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.

We appreciate Reviewer 2’s remark.

Revision plan. We now added these citations in lines 60-61 and in the Reference list as Reillo I, De Juan Romero C, García-Cabezas MÁ & Borrell V (2011). A role for intermediate radial glia in the tangential expansion of the mammalian cerebral cortex. Cereb Cortex 21: 1674–1694.

2-18. When introducing ferret as an interesting or important animal model, suitable original studies should be cited.

Revision plan:

For ferrets, there is a long history as experimental animals for electrophysiology similarly with cats and monkeys, but this is not a review of ferret biology. We thus added 6 additional references regarding ferret brain morphology and development listed below.

Jackson, C.A., J.D. Peduzzi, and T.L. Hickey (1989) Visual cortex development in the ferret. I. Genesis and migration of visual cortical neurons. J. Neurosci.9:1242–1253. PMID: 2703875.

Chapman B & Stryker MP (1992) Origin of orientation tuning in the visual cortex. Curr Opin Neurobiol 2: 498–501.

Chenn A., and McConnell S.K. (1995) Cleavage orientation and the asymmetric inheritance of Notch1 immunoreactivity in mammalian neurogenesis. Chenn A, et al. Cell PMID: 7664342.

Noctor SC, Scholnicoff NJ, and Juliano SL. (1997) Histogenesis of ferret somatosensory cortex. J Comp Neurol. 387(2):179-93.PMID: 9336222.

Reid CB, Tavazoie SF, Walsh CA. (1997) Clonal dispersion and evidence for asymmetric cell division in ferret cortex. Development. 1997 124(12):2441-2450. doi: 10.1242/dev.124.12.2441.PMID: 9199370

2-19. 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.

Revision plan.

Layer borders: We now labeled the approximate position of the boundary of the VZ in Figure 2E-G. We have revised the legends as follows; “The border of the VZ is shown with a white line”. For counting, we have determined borders by the distribution of DAPI, and radial glia-specific markers in our hands and determined the approximative distance of the VZ border from the ventricular surface in the antero-posterior axis where we performed the imaging in Figure 2E-G. The distance was approximately determined as 80 µm at P5 and 40 µm at P10.

Discrepancy in the intensity of CRYAB: We apologize for the unclear statement on how the images were acquired in the legends of Figure 2E-G. We now revised as follows; “Representative images taken with a 100X-objective lens are shown with MAX projection.”. In Figure 2E-G, images were taken as optical sections of 1.5 µm interval for 12 µm-thick sections. Those images were processed as MAX-projection onto the Z plane. On the other hand, In Figure 2B, we have used 20X-objective lens, instead of 100X-objective lens and did not perform any image projection procedure such as a MAX-projection and only 1 z-plane is shown. Therefore, the visual difference in the CRYAB intensity between Figure 2B and Figure 2E-G derives from whether max projection of several consecutive images was done.

2-20. Co-expression of CRYAB and FOXJ1. In Fig 5B this must be demonstrated with merged channels.

Revision plan.

We added the images with merged channels as requested and revised corresponding legends as follows: “Images with merged channels in A are shown with the same color codes, antibodies and scale bars as A.”.

2-21. 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.

Revision plan.

We will revise the cited sentence and will change the referred figure as follows: “These cells often aligned on a line parallel to the ventricular surface (Fig. 5A)”. We show these nuclear rows by arrows.

2-22. 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.

Revision plan.

We thank the Reviewer for their remarks on typos. We corrected the typos indicated by Reviewer 2. We agree with the Reviewer and also modified the title of Figure 5B as suggested by the Reviewer.

Reviewer #2 (Significance (Required)):

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.

We disagree the reviewer’s view that this study is clearly of interest to a very limited audience. This study first enabled a precise comparative analysis in which we could compare rich human single cell transcriptomes and the ferret dataset of single cell transcriptomes, which were based on greatly improved genomic information (especially, gene models). This study is also first to show global temporal patterns of cortical progenitors of a carnivore species, a famous gyrencephalic mammalian model, and have been shown to be similar to a primate species at the single cell transcriptomic level. Indeed, upon uploading this manuscript in BioRxiv, many non-ferret specialists as well as specialists have inquired datasets and requested some collaborations with us. So we believe that this paper has already attract a general interest of brain scientists.

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.

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

Please insert a point-by-point reply describing the revisions that were already carried out and included in the transferred manuscript. If no revisions have been carried out yet, please leave this section empty.

1-1. 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.

1. We revise the term “IPC” as “mitotic sibling of vRG” and stated that these cells might be tRG (CRYAB+) or non-tRG (CRYAB-) intermediate progenitors. By the term of “intermediate progenitors”, we did not intend to refer to TBR2+ neurogenic IPCs, but rather to an intermediate state of progenitors, in a general sense, with a similar morphology as tRG. To avoid any confusions on this terminology, we revised our manuscript by replacing “IPC” with “a sibling of vRG”.
2. We delete all statements relevant to Tsunekawa et al. data from the manuscript. We regret that we are not able to include Tsunekawa et al. data because we are planning to submit this data as a separate manuscript, which describes that in ferrets, vRG frequently (30% of apical division) generate non-Tbr2-positive mitotic sibling cells bearing a short basal process during the entire neurogenesis. This study includes a large volume of data with human ones and largely concerns stages that are earlier than that of tRG formation. It is, therefore, not appropriate to combine these data with those described in this manuscript.

1-2. 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.

1. We compared our ferret dataset to the human postnatal dataset recommended by the Reviewer 1 (Herring et al., 2022). As a conclusion of our analyses shown below, we found that Herring et al., (2022) dataset was not favorable for a comparative analysis with our ferret dataset regarding the fates of human tRG, because Herring’s human dataset was derived from the prefrontal cortex; This human dataset does not include neither tRG cell population nor ependymal clusters. We have also elaborated our discussion after analyzing Herring et al. dataset in the discussion.
2. We, therefore, pare down our claim in lines 365-366, by removing “(and presumably human)” to state that “Our pseudotime trajectory analyses and immunohistochemistry analyses strongly suggested that…”.
3. We also tone down the statements as for the discussion of the relationship between human and ferrets regarding the tRG progeny fates (originally lines from 372 to the end) and also elaborated our discussion after analyzing Herring et al. dataset in the same paragraph.

We will describe the details of our analysis of Herring et al. (2022) below.

https://www.dropbox.com/scl/fi/a0m72orxfsub66dh3hdbg/reviewer1_2ABC.pdf?rlkey=uzrd8ngclp87p5c8v24mqd1j7&dl=0

As mentioned above, Herring’s human dataset was derived from the prefrontal cortex, and that it did not include a specific subtype defined as tRG nor other HES1-expressing progenitor clusters such as RG in the original cluster annotation. We, therefore, re-clustered the raw dataset from GW22 (the earliest stage available) up to 10-months after birth by using Seurat pipeline with default parameters (B), and found a CRYAB-expressing population in the original “Astrocyte_GFAP” subtype among astrocyte clusters (A), which distribute in the most of collected stages, from late development through the adulthood. We then examined this dataset to find out whether tRG or its progenies are present.

After reclustering, CRYAB-expressing cells (with more than 1 raw count) represented 0.15% of the dataset and were grouped as a part of cluster 44, which was mostly derived from postnatal stages (among which 4-months was the most enriched one; C). Several astrocyte markers, such as SPARCL1, HOPX, CLU, and GJA1, as well as CRYAB, were enriched in the cluster 44 as revealed by FindMarkers (Methods). FOXJ1 expression was nearly absent overall in this dataset, indicating the absence of the ependymal cell population, a tRG-descendant cell types in ferrets (C).

To evaluate the similarity between cluster 44 and tRG or astroglial tRG, we next integrated Herring dataset with our ferret subset (about 15,000 cells) and the human GW25 subset from Bhaduri et al. (2021) of approx. 3,000 cells, both of which contained only progenitor cells. As we have done in Figure 7 of our original manuscript; we have removed cells other than progenitors, astrocytes and oligodendrocytes, such as neurons, microglia, endothelial cells. This resulted in about 20,000 cells in Herring dataset.

https://www.dropbox.com/scl/fi/nz3iulya5199i95ecr1un/reviewer1_2D.pdf?rlkey=kp7lwxtkn562un1uf9l1axn2p&dl=0

This integration (D) reveals that Herring’s cluster 44 is closely located to Bhaduri’s human and our ferret tRG clusters on UMAP, but does not overlap with these tRG clusters. This result further suggested that tRG population might be lacking or very rare in this neuron- and glia-dominated dataset, which might be due to the sampling method that targeted the enrichment of neuronal layers (Herring et al., 2022). It is also possible that this fragmented information on astrocyte and ependymal lineages could be due to the regional and/or temporal difference of samples between two human datasets.

1-3. 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.

According to the suggestion of reviewer 1, we analyzed two cortical organoid datasets (Bhaduri et al., 2020; Herring et al., 2022) to examine whether different tRG populations are present in organoids. Our analyses led us to conclude that tRG-like populations seem to be lacking in available organoid datasets; organoids can have CRYAB-expressing astrocyte-like cells in single-cell transcriptome datasets, but the presence of tRG-like cells seem to be unstable and dependent of lines and protocols how organoids are generated. A further assessment on tRGs’ cellular features is required on organoids by immunostaining experiments. In the light of this analysis, we elaborated our discussion by describing observations shown below. Below is our analysis of organoid data.

https://www.dropbox.com/scl/fi/8mj6u94t3hkzw6q61o7od/reviewer1_3AB.pdf?rlkey=10xiks25nzn9r90guw9l0onqh&dl=0

Bhaduri dataset contained organoids generated from 4 different lines, which showed a variability in terms of cell distribution on UMAP while overall temporal and differentiation axes were recapitulated (A). While keeping the original cluster annotations except for YH10 line, we performed reclustering. CRYAB was expressed in clusters 26 and 30 enriched in YH10 line, and cluster 29 enriched in 13234 line (B).

To confirm the identity of these clusters, we integrated organoid dataset with the dataset of primary tissues from the same paper (Bhaduri et al., 2020; C). https://www.dropbox.com/scl/fi/qnqv2e87t74uom2pg836d/reviewer1_3CD.pdf?rlkey=mv370b3dlogwvgh6ig8bdathpdl=0

As a result of the integration, tRG cells from the primary tissue were not overlapped with organoid-derived CRYAB-expressing cells, although a part of CRYAB-expressing organoid cells were localized in the integrated cluster 16 where primary tRG resided (D). Other cell types that were included in the integrated cluster 16 were “lateRG”, “vRG”, “oRG” from primary tissue dataset, and “glycolyticRG” from organoid dataset. We found that CRYAB-expressing organoid clusters 26 and 30 overlapped with “oRG/astrocyte” clusters of primary tissues.

1-4. 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.

As for human dataset, we agree that committed tRG was minor. Thus, we pared down our statements regarding the fates of tRG as mentioned in other comments, both in the Results and Discussion.

https://www.dropbox.com/scl/fi/aqsg5xlbxyoybzwq0xezp/reviewer1_4.pdf?rlkey=oxhmtko08nhvzkmsqxcjf9qua&dl=0

1-5. 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.

Our study demonstrated the presence of tRG cells up to P10 by immunohistochemistry and scRNA-seq. P5~P10 is the stage where neurogenesis became dominated by gliogenesis in the dorsal cortex in ferrets, although its timing is delayed in the visual cortex. On the other hand, Nowakowski et al. (2016) originally identified and defined CRYAB-expressing tRG, based on morphology and gene expression on human primary tissues during mid-neurogenic stages, while cortical neurogenesis is mostly declined in human postnatal stages. We have failed to find literatures or textbooks describing the presence of CRYAB-expressing tRG, while an ependymal layer was detected in the postnatal human cortices (Honig et al., 1996; preprint Nascimento et al., 2022). At the moment, the lack of information thus makes it difficult to compare the relationship of birth timing with the period of tRG persistence between ferrets and humans. In the revised manuscript, the “Discussion” will include this argument as well as the following difference between humans and ferrets in the RG scaffold.

Besides birth timing, Nowakowski et al. also reported that radial glia scaffold spanning from the VZ to the pial surface undergoes a transformation during neurogenic stages; tRG becomes the major RG population in the VZ, disconnecting VZ and OSVZ. In contrast, we did not find a discontinuous scaffold stage over the course of ferret neurogenesis. Instead, we still detected CRYAB-negative vRG with an apical attachment and a basal process extending beyond the OSVZ during stages where the peak of tRG expansion is achieved (such as P5 in Figure 2A, S3A). This appears to be a prominent difference between human and ferret corticogenesis.

1-6. 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.

We prepared the images for GFP+/CRYAB- vRG cells in an adjacent panel in Figure 2A as recommended by the reviewer (below). To better distinguish the morphology of an isolated vRG cell from other labelled cells, we sparsely labeled RG cells with EGFP at P3 by electroporation (Methods), and fixed the samples two days later (right panel). We highlighted the morphology (cell body and basal fiber) of a CRYAB- GFP+ vRG and that of a neighboring CRYAB+ GFP- tRG on the same panel to clarify that vRG did not express CRYAB.

https://www.dropbox.com/scl/fi/3wrmqdswt69t8pkdy30h7/reviewer1_6.pdf?rlkey=90ixbadan3mxx10m85jnpwphn&dl=0

2-2. 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.

To avoid such a misleading, we inserted the dotty lines in the revised Figure S1A to demarcate the tissue parts for scRNAseq, which correspond to almost all lateral cortices, mainly including the somatosensory area 1 and 2 with surrounding areas. We accordingly added the following sentence in the legend, “The approximate boundaries of dorsal cortex area used for scRNA sequencing are highlighted with dotty line segments in the dorsal cortex hemisphere above each strip.”.

We also show actual sampling for single-cell transcriptomics below. As our sampling was not restricted to the somatosensory cortex, we have revised “somatosensory cortex” as “dorsal cortex” in Lines 131 and 1191 of our manuscript.

https://www.dropbox.com/scl/fi/9gg508iood73zl02836g6/reviewer2_2.pdf?rlkey=lufevala88ihvc1p6mts463as&dl=0

2-4. 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.

We have now added the mitochondrial QC metrics in the new Figure S2A, and revised the legends as follows: “Violin plots showing the number of genes, mRNAs and the percentage of mitochondrial genes per cell in each sample and time point”. We have computed the percentage of mitochondrial genes for each cell type and found that the majority of cells in each cell type had a value less than 5% while the content value in some cells distributed along the range between 0% and 10%, up to a maximum of 28% (Figure S2A). Despite this, we have decided to include all cells that had less than 30% of mitochondrial genes in our analysis based on the percentage of reads mapped on mitochondrial genome for the following reasons:

1. The percentage of mitochondrial indicates respiratory activity, rather than apoptosis and the percentage of mitochondrial quite depends on the tissue type and species. For example, in human case, such percentage range from 5%~30% (Mercer et al., 2011 Cell; The human mitochondrial transcriptome).
2. Unfortunately, unlike human and mouse brains, there is no reference to show the percentage of mitochondrial in ferret brains. Therefore, the suitable way is to keep all of these cells.
3. These cells showing high percentage of mitochondrial genes are not clustered as an apoptosis cluster in UMAP, instead, these cells are observed in most of clusters (below). Therefore, we believed that these cells are not apoptotic cells and include these cells in further analysis. https://www.dropbox.com/scl/fi/4kp3fczxzo6x4fx8hqt8m/reviewer2_4_1.pdf?rlkey=ypojzbuwgelt51qlf56g883s9&dl=0
4. After all, we have obtained similar clustering overall after filtering cells with a higher percentage for mitochondrial genes; we set the threshold to 10%. This filtering resulted in 28,686 cells in our dataset. We then performed our workflow from the normalization step with the same settings that we applied to our original ferret dataset (Methods). Below, we show the results comparing newly generated clusters in this filtered subset on UMAP (left), and the original clusters shown in Figure 1B (right). 26 clusters were obtained in both conditions, and both major cell types and subtypes were conserved after filtering.

https://www.dropbox.com/scl/fi/0mlk69z7hckpiw03ivfjb/reviewer2_4_2.pdf?rlkey=hfvjrifrytmnywc4vchjvf0ms&dl=0

Clustering resolution: Our choice of the resolution was based on avoiding over- or under-clustering of ferret cells. After trying several resolution values, including 0.6, 0.8, 1.0 and 1.2, we have decided to use the resolution of 0.8 as the separation of cell types was the most reasonable among other resolutions that we have tried, in a similar way to actual known cell types. For example, the resolution of 0.6 did not distinguish “tRG” cells from “late_RG1” cells, as well as “early_RG” subtypes which were distinctly enriched with different cell cycle markers (Figure S2D). On the other hand, the resolution of 1.2 resulted in an over-clustering of IPC, OPC, DL neurons and microglia.

2-5. 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.

We focused on Figure 2A and S3A (2D is a histogram) to show the full morphology of CRYAB+ tRG, because Figure 2A is the initial presentation of tRG in this paper, and Fig. 2A and Fig.S3A images are taken on a 200-micrometer thick section, originally aiming to indicate that CRYAB-positive fiber is short, spanning nearly along the VZ and the SVZ. We made 3D-reconstructions of those images, which are rather better than orthogonal projections, in order to show that CRYAB+ fibers are shorter than those of vRG (terminating at positions around the upper boundary of the SVZ) and that the short basal processes are not due to the cut of long radial fibers during tissue sectioning (we show in below and in the final version as a supplementary figure and movies).

We show these 3D-reconstruction in below. Please download movie files from the following URLs to look at them clearly.

Figure 2A

https://www.dropbox.com/s/qocve596c5xhtlc/%E2%98%85fig2A-Ver02.mp4?dl=0

Figure S3A

https://www.dropbox.com/s/v8gqwfi1r8ff5n5/%E2%98%85figS3A-P0%20movie-ver2.mp4?dl=0

2-7. 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.

1. We are confident that this blue-labeled cells in Figure 3A and D are not vRG but mitotic sibling cell (of vRG) with a short basal fiber (that we named IPC in the initial manuscript). We now made the morphological features of these cells clearly visible by constructing 3D-views of the images with different snapshot images (we show below and in the final revision as a supplementary figure and movies). In addition, it divides once as time-lapse imaging revealed, hence this cell is still mitotic, instead of a postmitotic cell. Therefore, we used the term that is generally used for this type of cells, namely, intermediate progenitor cells (IPC), by which we did not intend to refer to TBR2+ neurogenic IPC. We plan to include these revised images into our fully revised manuscript.
2. We agree the reviewer 2 on the point that this blue-labeled cell may express CRYAB (the next comment of reviewer 2 essentially claim the same point), as we also wrote this possibility in line 204-207 of the original manuscript. It could not be technically possible at the moment to examine CRYAB expression in a cell emerging only in the course of time-lapse imaging. If we could label vRG with a transgenic or knock-in fluorescence marker, which mimics CRYAB gene expression, we could have figured out whether blue cells the mitotic vRG sibling cells (or mitotic tRG parental cell) express the CRYAB gene. Indeed, we tried to knock the EGFP gene in the CRYAB gene many times over a year, but have so far failed. Given that tRG is defined as the cell type expressing CRYAB with a short basal fiber at late-neurogenic stage, irrespective of its mitotic activity, this blue labeled vRG sibling cell in Fig. 3A (and/or Fig. 3D) might express CRYAB, hence can be a “mitotic tRG” (although its possibility seems to be low as shown in Fig. 2E). To avoid any possible misleading, we have changed the term of these cells to a “mitotic vRG sibling cell (or mitotic tRG parental cell) with a short basal process”, and add a comment that “this cell might be mitotic tRG with CRYAB expression”.
3. As for the TBR2 expression, we do not know these cells that appeared in the course of time-lapse imaging express TBR2 or not. As shown in Fig. 2F, 10% (P10) to 30 % (P5) of CRYAB+ cells express TBR2. On the other hand, “intermediate progenitors” do not necessarily express TBR2 in general. Therefore, we disagree on the reviewer 2’s comment “their analyses in Fig 4A contradict their interpretation on tRG’s parent cells”, but “our analyses in Fig 4A is compatible with our interpretation on tRG’s parent cells in time-lapse imaging”, and that is “a mitotic vRG sibling (or mitotic tRG parental cell) with a short basal fiber divides to produce CRYAB+ tRG at the end of timelapse imaging”. However, to avoid any overstatements or misunderstanding on this issue, we have revised related text as described above.
4. We are not able to include the data taken by Tsunekawa et al.. This is because we are going to submit a separate paper, which includes a large volume of data with human ones in collaboration with another group and largely concerns stages that are earlier than that of tRG formation. It is, therefore, not practical to combine these data with those described in this manuscript. Therefore, we remove all descriptions related with Tsunekawa et al.

Below we show snapshot images and 3D-reconstructions for Figure 3A and 3D. Please download movie files from the following URLs to look at them clearly.

@Figure 3A:

1)A time lapse movie (20 min interval) showing images around time 40:00 at which vRG underwent the second division. https://www.dropbox.com/s/znx3bboxefhj0jt/%E2%98%85Fig_3A%20movies%20around%2040%20h.mp4?dl=0

2)Snapshot images for time 40:00

https://www.dropbox.com/s/6y25mk4jhwqy6v7/%E2%98%85E38-fig3A-sRG-2.png?dl=0

3) 3D-reconstruction images at the same time point (40:00)

https://www.dropbox.com/s/so8hesjzy63yxmb/%E2%98%853D-reconstruction%20%2840.00%202nd%20div%29.mp4?dl=0

4) The entire time-lapse movies of time 0:00-84:00; The mitotic sibling cell of the vRG is indicated by a white arrow.

https://www.dropbox.com/s/ywua95f8fmohsmc/%E2%98%85Fig3A-arrow-time.mp4?dl=0

@Figure 3D:

A revised time-lapse snapshots of Figure 3D.

https://www.dropbox.com/s/xyet4virt3j9u3t/%E2%98%8520211220%EF%BC%8DP0%EF%BC%8Dtimelaps-xt04corrected.psd?dl=0

The assignment of the cell has corrected to the right one for the same mitotic cell because cell body position at the first two time points were misassigned in the original manuscript (at the following time points, there is no change).

Snapshot image at time point of 06:20; https://www.dropbox.com/s/hn3v6ao1qkhnfjh/%E2%98%85Fig3D%20sRG%20at%200620.png?dl=0

Rotating movie of 3D-reconstruction at time point of 06:40:

https://www.dropbox.com/s/6taqjr0u21x5tn0/%E2%98%853Drotated%20movie%20of%20time%20point%2006.40.mp4?dl=0

2-8. Alternative interpretation of time-lapse images (lines 196-197): maybe a tRG can generate one tRG CRYAB+, and one IPC CRYAB-.

We changed the term of IPC to “a mitotic vRG (or mitotic tRG parental cell) sibling cell” and describe the possibility that “This mitotic vRG sibling cell (or mitotic tRG parental cell) can be a mitotic tRG if this cell express CRYAB, and its apical division generates one tRG and one CRYAB-negative climbing cell with an unknown identity, replacing the description of line 196-197.

2-9. 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.

We have corrected the shifted position of arrows in Figure 5E. We have removed “mitosis” in the title of Figure 5E since the initial manuscript did not include descriptions on mitosis in the text.

2-10. Line 277: “Transcriptomic trajectories were homologous across the two species”. What does this refer to? What are these trajectories? Pseudotime? Is this statistically tested?

We revised our description in this part as “Temporal patterns and variety of neural progenitors during the cortical development were similar to each other between humans and ferrets at the single cell transcriptome level”.

2-11. 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.

To clarify our statement, we changed this sentence into “tRG cells also showed a remarkable similarity between the two species (Fig. 6C, 6D), as represented by a high level of expression for the combination of CRYAB, EGR1, and CYR61 (Fig. 6E)”

2-13. 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.

We added several issues discussed in the responses to the reviewers to Discussion. Please look at our responses to comment 2-14 and 2-15 as well as the preliminary manuscript.

2-15. 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.

Taking the comments from reviewer 1 and 2 into account, we largely revised “Discussion” with a more moderate expression, by incorporating comparative analyses with other human datasets, and we also emphasize the importance of in vivo studies as the next step. We just paste the last paragraph of the preliminary revised Discussion. Please see the “Discussion” in the preliminary revision of our manuscript.

“In ferrets, genetic manipulations can be achieved through in utero or postnatal electroporation, as well as via virus-mediated transfer of DNA (Borrell, 2010; Kawasaki et al, 2012; Matsui et al, 2013; Tsunekawa et al, 2016). Thus, it is theoretically possible to disrupt the CRYAB gene in vivo in ferrets to investigate its role in tRG and their progeny, including ependymal cells, and to track the tRG lineage. If the CRYAB gene is essential to form ependymal layers, we will be able to explore how the ventricle contributes to cortical folding and expansion. Despite extensive efforts over a year, we have thus far been unsuccessful in knocking in and/or knocking out the CRYAB gene. Nevertheless, we anticipate that technical advances will surpass our expectations, both in ferret and human organoids. Taken together, these functional studies in ferrets as well as in human organoids hold promising insights into the understanding of the tRG lineage and its contribution to cortical development in the near future”.

2-16. 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.

We removed the mentioned statement from our manuscript and revised lines 58-59 as follows: “In many mammalian phylogenic states, cerebral cortex evolved to gain an additional germinal layer (Smartet al. 2002; Zecevic et al. 2005; Kriegstein et al. 2006; Reillo et al. 2011)”.

2-17. 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.

We now added these citations in lines 60-61 and in the Reference list as Reillo I, De Juan Romero C, García-Cabezas MÁ & Borrell V (2011). A role for intermediate radial glia in the tangential expansion of the mammalian cerebral cortex. Cereb Cortex 21: 1674–1694.

2-18. When introducing ferret as an interesting or important animal model, suitable original studies should be cited.

For ferrets, there is a long history as experimental animals for electrophysiology similarly with cats and monkeys, but this is not a review of ferret biology. We thus added 6 additional references regarding ferret brain morphology and development listed below.

Jackson, C.A., J.D. Peduzzi, and T.L. Hickey (1989) Visual cortex development in the ferret. I. Genesis and migration of visual cortical neurons. J. Neurosci.9:1242–1253. PMID: 2703875.

Chapman B & Stryker MP (1992) Origin of orientation tuning in the visual cortex. Curr Opin Neurobiol 2: 498–501.

Chenn A., and McConnell S.K. (1995) Cleavage orientation and the asymmetric inheritance of Notch1 immunoreactivity in mammalian neurogenesis. Chenn A, et al. Cell PMID: 7664342.

Noctor SC, Scholnicoff NJ, and Juliano SL. (1997) Histogenesis of ferret somatosensory cortex. J Comp Neurol. 387(2):179-93.PMID: 9336222.

Reid CB, Tavazoie SF, Walsh CA. (1997) Clonal dispersion and evidence for asymmetric cell division in ferret cortex. Development. 1997 124(12):2441-2450. doi: 10.1242/dev.124.12.2441.PMID: 9199370

2-19. 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.

Layer borders: We now labeled the approximate position of the boundary of the VZ in Figure 2E-G. We have revised the legends as follows; “The border of the VZ is shown with a white line”. For counting, we have determined borders by the distribution of DAPI, and radial glia-specific markers in our hands and determined the approximative distance of the VZ border from the ventricular surface in the antero-posterior axis where we performed the imaging in Figure 2E-G. The distance was approximately determined as 80 µm at P5 and 40 µm at P10.

2-20. Co-expression of CRYAB and FOXJ1. In Fig 5B this must be demonstrated with merged channels.

We added the images with merged channels as requested and revised corresponding legends as follows: “Images with merged channels in A are shown with the same color codes, antibodies and scale bars as A.”.

2-21. 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.

We revise the cited sentence and will change the referred figure as follows: “These cells often aligned on a line parallel to the ventricular surface (Fig. 5A)”. We show these nuclear rows by arrows.

2-22. 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.

We thank the Reviewer for their remarks on typos. We corrected the typos indicated by Reviewer 2. We agree with the Reviewer and also modified the title of Figure 5B as suggested by the Reviewer.

Description of analyses that authors prefer not to carry out

Please include a point-by-point response explaining why some of the requested data or additional analyses might not be necessary or cannot be provided within the scope of a revision. This can be due to time or resource limitations or in case of disagreement about the necessity of such additional data given the scope of the study. Please leave empty if not applicable.

2-1. 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.

We would like to thank the reviewer for carefully reading our manuscript and providing us with valuable feedback. However, we would like to clarify that there might have been a misunderstanding regarding our conclusion about the identification of oRG-like cells in ferrets.

Our study did not conclude that we have identified oRG cells in ferrets with “a quite different transcriptomic profile than in human”. Instead, our findings indicate that unlike oRG cells in human, ferret oRG-like cells did not exhibit specificity for human oRG markers (such as HOPX and CLU) that would enable us to distinguish them from other late RG cells in ferrets. Despite this, oRG score derived from human oRG marker expression showed higher values in predicted ferret oRG-like cells compared to other ferret RG cells, reflecting a similarity of the transcriptome profile between human oRG and ferret oRG-like cells (Figure 7H-I). We will carefully describe our methodology to reach this conclusion in response to reviewer 2’s comment regarding how we determined ferret oRG in a later comment.

2-3. 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.

We disagree with the reviewer 2’s comment. We would like to clarify that we collected brain tissues in two different ways for the same set of developmental stages; one brain tissue by removing cortical plate (T); another independent brain tissue at the same developmental stage by sorting GFP-labelled lineage from neural progenitors that were electroporated at embryonic stages (AG, Methods). Both manipulations of samples aimed to increase progenitor cell populations in scRNAseq. Therefore, we have two sets of samples of the same temporal series, each prepared in a totally different way. All cell types were present in both methods of collection shown as Supplementary Figure 2E’ (section 2) that separates samples by different preparations at each stage (by modifying Supplementary Figure 2E; section 2). We believe that the biological replica (n=2) in this manuscript would be sufficiently reliable, judged by its reproducibility.

https://www.dropbox.com/scl/fi/levyqy9ngvpyio1yl9oif/reviewer2_3.pdf?rlkey=r4aw0hu9cdn68f1pvhp734vxx&dl=0

Here, we also cite several examples of papers important in the field of single-cell or bulk transcriptomics of brain tissue, where only a single replicate or pair (replica) was taken for experiments on mice, humans and ferrets:

mice: Ogrodnik et al., 2021 PMID: 33470505, Hochgerner et al., 2018 PMID: 29335606, Joglekar et al., 2021 PMID: 33469025;

human: Herring et al., 2022 PMID: 36318921, Polioudakis et al., 2019 PMID: 31303374, Mayer et al., 2019 PMID: 30770253, Fietz et al., 2012 PMID: 22753484;

macaque: Schmitz et al., 2022 PMID: 35322231;

ferret: Johnson et al., 2018 PMID: 29643508.

2-14. 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.

We disagree with the reviewer 2 as for ferrets, because we accessed the relationship of tRG and their progeny cells by not only in silico but also in vivo analyses.

On the other hand, as for progenies of human tRG, they were predicted certainly depending on the molecular relationship by comparison with ferrets without histochemical evidence, as pointed by reviewer 2, and the populations of these committed tRG are small. Therefore, we removed “(and presumably also human)” and we tone down about the progeny relationship of tRG as a prediction. We also acknowledge that further studies are needed to confirm the lineage relationships among cell types, as we discussed in the Discussion part.

Reviewer #2 (Significance (Required)):

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.

We disagree the reviewer’s view that this study is clearly of interest to a very limited audience. This study first enabled a precise comparative analysis in which we could compare rich human single cell transcriptomes and the ferret dataset of single cell transcriptomes, which were based on greatly improved genomic information (especially, gene models). This study is also first to show global temporal patterns of cortical progenitors of a carnivore species, a famous gyrencephalic mammalian model, and have been shown to be similar to a primate species at the single cell transcriptomic level. Indeed, upon uploading this manuscript in BioRxiv, many non-ferret specialists as well as specialists have inquired datasets and requested some collaborations with us. So we believe that this paper has already attract a general interest of brain scientists.

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

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: This study by Magalhaes et al sheds light on the molecular underpinnings of the relative resistance of children to severe COVID-19. The authors found that priming of epithelial cells by resident immune cells to express tonic levels PRR receptors MDA-5 and RIG-I predisposes the epithelial cells for a faster and more robust onset of IFN-beta production upon SARS-CoV-2 infection. The study uses a combination of in vitro and ex vivo models, as well as mining of scRNA-Seq datasets from clinical specimens.

Major comments: The claims and conclusions are supported by the data and therefore no new experiments are needed.

Optional

1. The use of primary cells (i.e. human airway epithelial cultures cross talking to immune cells) would make this study more compelling, although I assume that the major findings would be recapitulated in such models.
2. It is not clear how the use of Yersinia enterocolitica to trigger activation of PBMC is relevant to this story. Using different (commensal) pathogens to achieve PBMC activation may yield different and more physiologically relevant results.
3. The manuscripts would greatly benefit from improved structure and focus, particularly in the Abstract, Introduction and Results sections. The text is very dense, and makes it difficult for the reader to follow the flow and to distinguish important from less important information. Particularly, the introduction starts very broadly introducing COVID-19, which I think we are by now all familiar with. Directly starting with the burning question why kids get less sick with SARS-CoV-2 would capture the readers' attention better. Figure 1 a is beautiful for a review but much too dense to help the reader as a graphical abstract. In the results section, for each experiment, leading with clearly stating the rationale of the specific question, the gap in knowledge and why the gap is there, then followed by the results, then summarizing the impact of said results, would make this a much more enjoyable read and help the reader evaluate the novelty and impact better, particularly for Figures 1, 2, and 3 (but also all others). The interaction wheel graphs (Figure 4. are amazing, but are not properly explained in the text (do I read this right that in adults, all the crosstalk is basically performed by proliferating T-cells?). In all, these scientific writing issues sell an otherwise beautiful story short.

Referees cross-commenting

I agree with reviewers 1 and 2 that the use of primary cells would significantly elevate the story. However, I think this should be "optional", as I do not think it would change the findings.

Significance

General assessment:

The main strength of the study are its topic and clearly relevant question: why do kids rarely get severe COVID-19? The main novelty is the answer to this question, that immune cell-epithelial crosstalk in children elevates the tonic expression of MDA5 and RIG-I via the IRF1 axis, leading to faster onset of IFN production and signaling upon SARS-CoV-2 challenge, which ultimately mounts an antiviral response detrimental to robust SARS-CoV-2 replication. The study uses an innovative combination of in vivo and ex vivo experiments and analysis of clinical specimens.

The significant advance of this study to the field is clear to this reviewer, although it could be much better stated in the manuscript, as described at length above. The study is of great interest to the field of immunology and virology, and also has clinical and translational impact with respect to risk assessment for severe COVID-19 per age group, as well as epidemiological considerations for infection control.

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 the manuscript entitled "Enhanced airway epithelial response to SARS-CoV-2 infection in children is critically tuned by the crosstalk between immune and epithelial cells" Gonçalves Magalhães et al address the molecular interactions that result in greater antiviral responses to coronavirus infections in children versus adults. The authors have re-analyzed previously reported scRNA data from the nasal passages of healthy donors and found heightened pro-inflammatory responses in the nasal passages of children relative to the signatures in adults. Thus, the authors posit that this could result in priming of epithelial cells subsequently boosting antiviral responses to SARS-CoV-2 infection in a RIG-I and MDA5 dependent manner. Beyond the expected antiviral protection conferred by type I IFNs, the authors demonstrate type II IFN and TNF can also promote antiviral priming. Although epithelial and immune cell crosstalk has been previously demonstrated, this study proposes an age-dependent differences in the magnitude of this crosstalk. Indeed, bacterial stimulation of PBMCs derived from children resulted in greater cytokine [production and A549 priming. The article was well written, the data presented is compelling, and appropriately controlled. Importantly, the authors appropriately acknowledge the study limitations. The mechanisms that govern the increased inflammatory responsiveness of tissues derived from children remain to be addressed.

Major comments:

• Although the authors acknowledge this limitation, do primary epithelial cells respond to priming? Is there an age difference in viral detection and/or the response to priming?
• Optional: Does the deletion of IRF3 phenocopy MAVS deficiency in the context of type I IFN priming and blockade of IFN replication (Fig 2D)? Does priming induced increases in IRF1 and IRF7 and is this sufficient to overcome the loss of IRF3 (PMID: 25520509)?

Minor comments:

• Are the differences observed in MDA5 and RIG-I expression after PBMC stimulation across A549 IFNR KO cells significant?
• Figure 2C could be strengthened by the addition of total IRF3 and STAT2 immunoblot.
• Figure 3C, include tSTAT2 control.
• There is one typo on line 610. Change OSA2 to OAS2.
• Please expand the discussion to include relevant work on IFN and TNF-mediated antiviral priming (PMID: 16537619) and epithelial-immune cell crosstalk (PMID: 36563691).

Significance

Advance: This study expands upon previous work by the authors that described that children have higher expression of RLRs and in epithelial and immune cells relative to adults. The current work, provides an incremental but important advance to the previous study by demonstrating that PBMC-mediated priming of epithelial cells (A549). However, it remains to be addressed whether epithelial cells from children have increased capacity to detect SARS-CoV-2 infection or respond to priming.

Audience: This work is of interest to pediatric clinicians, virologists, immunologists, cell biologists, bioinformaticians.

Reviewer expertise: viral innate immunity; IFN regulation

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 the manuscript entitled "Enhanced Airway Epithelial Response to SARS-CoV-2 Infection in Children is Critically Tuned by the Cross-Talk Between Immune and Epithelial Cells", the authors report a mechanism how airway epithelium of children is in a primed state to more efficiently defense SARS-CoV-2 infection. They interestingly found that the expression level of MDA5 and RIG-I are essential for this primed stage. In details, they discovered that enhanced immune-epithelial cell interactions through cytokines play a major role in the upregulation of these two RNA sensors, where IRF1 may be the most important regulator. They fully used their single cell sequencing data and established a relative convincing in vitro model to validate their hypothesis. This story is quite attractive to me since they explained the behind mechanisms for a clinical phenomenon. However, there are still some issues need to be addressed.

Major comments:

1. A549 is a cancer cell line and cannot support SARS-CoV-2 replication. The authors reply on it too much. Some important experiments need to be performed in primary airway epithelial cells.
2. MDA5 and RIG-I protein levels need to be detected in some important experiments such as Figure 4D, Figure 5A, Figure 5B, Figure 6B.

Minor comments:

1. In figure 1B, the authors claimed that "As expected, pre-treatment alone did not trigger notable IFN-β transcription". But in my mind, pretreatment elevated IFN- β mRNA at least 5 folds compared to mock group (from 10-2.5 to 10-1.8). The authors need to make the statement more accurate.
2. In figure 2A, the authors used IFNARKO cells. This should be mentioned in the manuscript.
3. The statement of Figure 2C should be more accurate, "This was also confirmed at the downstream level of STAT2 activation, at which phosphorylation was still observed upon infection of RIG-I or MDA5 single KOs with SARS-CoV- 2, but was fully diminished only in the double KO cells (Fig 2C)." IRF3 phosphorylation is not downstream of STAT2. More downstream event can be shown as STAT1/STAT2/IRF9 nuclear translocation or ISGs transcription. Also, total STAT2 levels should be shown here.
4. It is quite surprising that none of IRNAR, RIG-I, MDA5 affected SARS-CoV-2 replication in untreated cells.
5. If we compare the mRNA induction of MDA5 and RIG-I between Figure 4D and Figure 6B, they are not in the same amplitude. There are 10 folds induction in Figure 4D while 100-300 folds in Figure 6B.

Significance

This is an interesting study with clear data. They fully used their single cell sequencing data and established a relative convincing in vitro model to validate their hypothesis. They uncovered an important mechanism why airway epithelium of children is in a primed state and more resistant to SARS-CoV-2 infection.

The limitation here is that they do not have in vivo model to support their conclusions.

Nevertheless, this manuscript is still able to answer a clinical question. Why children are much more resistant to SARS-CoV-2 infections compared to adults? It has good clinical significance. I believe that a broad audience will be interested in this story.

I study virus and cancer induced metabolic reprogramming, virus and host interaction and innate immune regulation, and mass spectrometry (metabolomics and proteomics).

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-2023-01901

Corresponding author(s): Gavin, Sherlock

We thank the reviewers for their comments and their generally positive reviews – the reviews were constructive, and we have revised the manuscript to deal with all the requested changes and suggestions. We believe the manuscript is improved as a result, and hope that the reviewers agree that it is now suitable for publication. Below with provide a point-by-point reply that explains what revisions we have made. Reviewers’ comments are italicized, while our responses are highlighted.

• *

Reviewer #1:

• *

*It would be interesting have an idea of the global mutation rates and spectra in the diploid and haploid lineages across the conditions as well. The S. cerevisiae mutational spectrum has been shown to be dependent on the environment and genetic background to an extent but not ploidy. Ploidies differ in terms of just the frequency. How similar/ dissimilar are the overall mutational spectra here? Were there any homozygous mutations in the diploids? *

• *

• We have plotted the mutation types in the diploid and haploid lineages across the conditions to compare the frequencies of each type of mutation between ploidies, which is now presented as Supplemental Figure 3. The mutation types between ploidies for each of the conditions look similar. Homozygous diploids are indicated in Table 2.

  *Fitness gains and losses can happen without trade-offs if neutral home mutations are non-neutral in non-home conditions. Can the authors comment on that in this context. Physico-chemically, how different are the home/ non-home environments? How do the fitness effects correlate across the environments in the absence of these adaptive mutations? It would also be useful to know the extent of fitness variance of the populations in the home and away environments, this would aid the reader better grasp the significance of fitness gains/loss. *

• We agree that trade-offs could occur as a result of mutations that are neutral in the home condition showing trade-offs in the non-home conditions. However, in the newly added Supplemental Table 1, it is clear that most lineages have several passenger mutations, yet for lineages carrying the mutations in the same candidate beneficial mutation, they have largely similar pleiotropic profiles, suggesting that the influence of neutral mutations that arise in the home environment do not play a large role in determining fitness in other environments, at least for those tested. We have not generated strains that only contain the passenger mutations – while that would empirically test the fitness effects of the passenger mutations, it would be extremely time consuming to generate such strains, and the results would be unlikely to change our claims in the paper.

  *A summary table of the pleiotropic effects would be very useful as in Bakerlee et al. 2021. *

• We have added a summary table (Supplemental Table 3) of the pleiotropic effects as suggested. Reviewer #2

*The major conclusion of the manuscript state that "mutations in the same genes tend to produce similar pleiotropic effects", suggesting that a number of times this does not occur. For instance, the authors comment on the case of PDR3, which does not always produce 'cost-free' adaptation across environments. I believe that, to strengthen and better define their conclusion, the authors should develop a quantitative analysis of the reproducibility of pleiotropic profiles (that considers how many times genes have been found mutated). The heatmaps provided are compelling, but make it hard to generalize on how often, and to what extent, the gene mutated can predict pleiotropy across various environments. *

• *

• We have calculated pairwise correlations between pleiotropic profiles for mutations that arose in the same environment either in the same gene, or in different genes, and added this Supplemental Figure 10. These data show that by and large, correlations between mutations in the same gene are higher than those for different genes.

  *In the concluding sentence of the discussion, it is unclear whether the authors are speculating about a role of the strength of selection in determining pleiotropy based on their results, or if that only represents a suggested hypothesis to test in future studies. *

• We have modified the concluding sentence to clarify its meaning (it was a suggested hypothesis)

  *The method used to identify putative adaptive mutations should be described in more detail. For instance, I seem to understand that only one mutation per lineage is considered 'adaptive'. However, many lineages seem to have more than one mutation. Based on what reported in the method section, the adaptive mutations have been hand-picked based on previous knowledge of selection in the environments of choice ("the list of genes was curated based on those genes' interactions with other identified genes or pathways known to be involved in the adaptation of that specific condition from previous work"). If this assumption is correct, the criteria for such a curation should be specified in more detail. *

• We have further clarified our criteria in the text; note, there was not a requirement for there to be only a single beneficial mutation per lineage, though very few lineages had two candidate beneficial mutations.

  *The term 'Pareto front' is technical and left undefined. *

• We have clarified the meaning of Pareto front

  *The section ' adaptation can be cost-free' only refer to figure 4, (with adaptive mutant lineages from populations evolved in fluconazole), while it comments extensively on mutation isolated in clotrimazole (reported in Sup. Fig10, not mentioned in the section). *

• We thank the reviewer for noting our oversight – we have also now referenced the supplementary figure too (now Supplementary Figure 11). Reviewer #3

*It would be helpful if the authors could clearly provide information on the zygosity of the evolved mutations, as the presence of mutations in homozygous or heterozygous states can impact the results of the study. *

• We have added zygosity information to the genotypes in the text and in Table 2, Summary of Adaptive Mutations

  *Do any of the evolved lineages have multiple adaptive mutations or other potentially adaptive mutations? If so, it would be great if the authors could provide a table listing these lineages and mutations. *

• We have added Supplemental File 1, which enumerates the adaptive and passenger mutations found in each lineage. Candidate adaptive mutations are in highlighted in red. Of the ~200 adaptive lineages, 4 have two candidate adaptive mutations, while the rest have only one.

  *In the Pooling of the Isolated Clones section of the Methods, the ancestor and subject pools were mixed in different ratios for different types of pools. While not strictly necessary, it would be helpful to provide a brief explanation for this. *

• We have added a brief explanation

  *The conditions listed in Table 1 and Supplemental Figure 2 do not seem to match perfectly. *

• We have corrected Supplemental Figure 2 such that it matches Table 1

• *

Supplementary Figure 6 demonstrates reproducible fitness estimates across lineages with the same mutations but distinct barcodes, supporting the authors' inference of adaptive mutations. However, it also appears to show no evidence of interactions among these mutations. Can the authors clarify if this is due to the absence of lineages with multiple mutations or if no observable interactions were found?

• *

• See response above – there are very few lineages with more than one candidate beneficial mutation. The remaining passenger mutations are thus likely neutral.

• *

*In the Pleiotropy is common, strong and variable section of the results, all three conditions were noted to have their evolved lineages tested in other conditions and presented in Supplementary Figure 5. However, due to the rapid dominance of lineages evolved in clotrimazole, there is no comparison data for them in Supplementary Figure 5. *

• Unfortunately, we were not able to generate robust fitness remeasurements in the clotrimazole condition, due to the rapid takeover by lineages that were evolved in that condition

  *In the Results section on cost-free adaptation, it would be beneficial to include any compositional differences, such as pH, between the two drugs used that could have contributed to the fitness effects of the evolved lineages in pH 7.3. *

• We are not aware of any such differences – we did not pH any of the media other than the media with a specific pH.

  *Results - Adaptation can be cost-free: While the authors did state "at least across the conditions in which we remeasured fitness" at the end of the paragraph, it may be prudent to exercise caution when stating "cost-free adaptation" as only a few conditions were tested. For instance, an all-beneficial or all-deleterious result can sometimes be obtained solely based on the chosen conditions. *

• *

• We have added additional caution in the text based on the reviewer’s suggestion.

• *

*Colormaps in Figure 4, Supplemental Figure 6, 10, and 11: The colors for values below -0.2 are uniform, whereas the heatmaps exhibit darker blues. *

• *

• We have edited the color scales on Figure 4 and Supplemental Figures 6, 10, and 11 (now Supplemental Figures 6, 11, and 12) such that the scales are uniform.

• *

*Results - Pleiotropy varies according to the mutated gene: "For example, haploid lineages adapted in glycerol/ethanol with mutations in IRA1 show the same pattern of fitness effects across conditions (Supplemental Figure 6)." I believe the authors are referring to Supplemental Figure 11. *

• The reviewer is correct – we have fixed this reference to what is now Supplemental Figure 12

  *On the topic of IRA1, IRA2, and GPB2 in the section "Pleiotropy varies according to mutated gene" in the Results: Although IRA1 mutants exhibit highly similar patterns, it is challenging to ascertain which of the two genes, GPB2 or IRA2, has a more similar pattern. *

• *

• We have create a new supplemental figure showing the correlation between mutations in the same gene and mutations in different genes for lineages evolved in the same condition – see response to Reviewer #1 above.

• *

Results - Pleiotropy varies according to mutated gene: From "If lineages isolated from the same home environment have similar pleiotropic profiles..." to the end of that paragraph. While it is true that "pleiotropy varies according to target genes and not environment alone," it may be premature to suggest that the environment is the "main driving force" of pleiotropy without some form of statistical analysis.

• We did not intend to suggest that environment is the main driving force - that section was somewhat poorly worded. We have modified the wording to make that clearer.

  *Discussion - line 5, paragraph 2: "For example, in glycerol/ethanol, the haploid adapted lineages have a trade off at 37{degree sign}C but the diploid adapted lineages do not (Supplemental Figure 11)." I believe the authors are referring to Supplemental Figure 6. *

• We thank the reviewer for spotting this and have fixed the figure reference.

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 an intriguing study on the pleiotropic effects of adaptive mutations in yeast populations evolving in different environments. The study used haploid and diploid barcoded budding yeast populations to understand the pleiotropic effects of adaptive mutations in "non-home" environments where they were not selected. The findings indicate that pleiotropy is common, and most adaptive evolved lineages show fitness effects in non-home environments, which can be beneficial or deleterious. The results also highlight how ploidy influences the observed adaptive mutational spectra in different conditions. The methodology involved whole-genome sequencing and pooled fitness remeasurement assays in 12 environments with various perturbations. The study concludes that pleiotropic effects are unpredictable, but lineages with adaptive mutations in the same genes tend to show similar effects. Overall, the study provides insights into the dynamics of adaptation and the impact of pleiotropy in different environments.

Major comments

1. It would be helpful if the authors could clearly provide information on the zygosity of the evolved mutations, as the presence of mutations in homozygous or heterozygous states can impact the results of the study.
2. Does any of the evolved lineages have multiple adaptive mutations or other potentially adaptive mutations? If so, it would be great if the authors could provide a table listing these lineages and mutations.

Minor comments

1. In the Pooling of the Isolated Clones section of the Methods, the ancestor and subject pools were mixed in different ratios for different types of pools. While not strictly necessary, it would be helpful to provide a brief explanation for this.
2. The conditions listed in Table 1 and Supplemental Figure 2 do not seem to match perfectly.
3. Supplementary Figure 6 demonstrates reproducible fitness estimates across lineages with the same mutations but distinct barcodes, supporting the authors' inference of adaptive mutations. However, it also appears to show no evidence of interactions among these mutations. Can the authors clarify if this is due to the absence of lineages with multiple mutations or if no observable interactions were found?
4. In the Pleiotropy is common, strong and variable section of the results, all three conditions were noted to have their evolved lineages tested in other conditions and presented in Supplementary Figure 5. However, due to the rapid dominance of lineages evolved in clotrimazole, there is no comparison data for them in Supplementary Figure 5.
5. In the Results section on cost-free adaptation, it would be beneficial to include any compositional differences, such as pH, between the two drugs used that could have contributed to the fitness effects of the evolved lineages in pH 7.3.
6. Results - Adaptation can be cost-free: While the authors did state "at least across the conditions in which we remeasured fitness" at the end of the paragraph, it may be prudent to exercise caution when stating "cost-free adaptation" as only a few conditions were tested. For instance, an all-beneficial or all-deleterious result can sometimes be obtained solely based on the chosen conditions.
7. Colormaps in Figure 4, Supplemental Figure 6, 10, and 11: The colors for values below -0.2 are uniform, whereas the heatmaps exhibit darker blues.
8. Results - Pleiotropy varies according to the mutated gene: "For example, haploid lineages adapted in glycerol/ethanol with mutations in IRA1 show the same pattern of fitness effects across conditions (Supplemental Figure 6)." I believe the authors are referring to Supplemental Figure 11.
9. On the topic of IRA1, IRA2, and GPB2 in the section "Pleiotropy varies according to mutated gene" in the Results: Although IRA1 mutants exhibit highly similar patterns, it is challenging to ascertain which of the two genes, GPB2 or IRA2, has a more similar pattern.
10. Results - Pleiotropy varies according to mutated gene: From "If lineages isolated from the same home environment have similar pleiotropic profiles..." to the end of that paragraph. While it is true that "pleiotropy varies according to target genes and not environment alone," it may be premature to suggest that the environment is the "main driving force" of pleiotropy without some form of statistical analysis.
11. Discussion - line 5, paragraph 2: "For example, in glycerol/ethanol, the haploid adapted lineages have a trade off at 37{degree sign}C but the diploid adapted lineages do not (Supplemental Figure 11)." I believe the authors are referring to Supplemental Figure 6.

Significance

General assessment:

The research is well-conducted, utilizing both haploid and diploid barcoded yeast populations, and isolating adaptive clones to determine fitness effects in non-home environments. The double-barcoding system allowed the authors to perform pooled fitness measurements of a large number of lineages coming from different home-environments in a plethora of conditions accurately and efficiently. The inclusion of a multiple evolution conditions followed by fitness measurements in a broader range of conditions allowed the authors to study the effect of environment to pleiotropy.

The low number of generations used in this study, however, could hamper the discovery of more adaptive mutations, particularly those with smaller effects, and also make the study underpowered for studying epistasis among the evolved mutations. Furthermore, while the definition of pleiotropy in this study is reasonable and practical, it also makes most, if not all, generalist mutations pleiotropic and hence it's not surprising to see pleiotropy to be so common in this study.

Advance:

This study is an extension of a few recent publications. Jerison et al. (2020) evolved 20 haploid founder replicates in 11 environments for about 700 generations and measured the fitness of evolved clones - one clone from each replicate - across these conditions. This study provides in-depth analyses of how environments affect pleiotropy and a certain level of analyses for the underlying mutations. Bakerlee et al. (2021) evolved several hundred barcoded haploid and diploid populations in a few environments for 1000 generations and traced not only the fitness changes but also the dynamics of pleiotropy longitudinally. The environments used in this study are similar to each other, and so were the results, with the exception of environments with high (37˚C) and low (21˚C) temperatures. The current study utilized their double-barcoding system to allow for testing both haploids and diploids in a broader range of conditions. Although only three of the starting environments were chosen for further analyses, these environments are more dissimilar, and the putative underlying adaptive mutations in the evolved clones were identified and more thoroughly analyzed.

The study's findings offer valuable insights into the intricate relationship between adaptation, pleiotropy, and environmental dynamics. While the complexity of pleiotropy and the multitude of factors that influence it make it challenging to comprehensively address all aspects in a single study, the results presented here contribute significantly to our understanding of this phenomenon. Nevertheless, further research of this nature is crucial to deepen our knowledge of the underlying mechanisms and to identify overarching patterns that can be applied across diverse systems. Overall, this study represents a promising step towards advancing our understanding of pleiotropy and its role in adaptive evolution.

Audience:

While the current study focuses on yeast as a model organism for evolutionary experiments, the implications of pleiotropy extend far beyond basic research. An understanding of the pleiotropic effects of mutations is crucial for comprehending the mechanisms of evolution and developing effective clinical interventions. As pleiotropy can affect disease outcomes and drug responses, the insights gained from this study can have far-reaching implications in the fields of biology and medicine. Thus, this study contributes not only to our understanding of yeast genetics but also to broader areas of research and application.

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

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

In this manuscript, the authors address the question of pleiotropy (the multiple effects of a single mutation on different traits) of adaptive mutations occurred during evolutionary processes. To do this, they evolve populations of S.cerevisiae strains in 12 environments, they identify the major adaptive mutation occurring in a subset of them, and they use a barcoding system to address their effect on fitness in environments where they were not evolved. The study confirm a number of conclusions from previous studies, such as the frequency of positive and negative pleiotropy of evolved lines when tested in other environments. The major novelty of this work is represented by the focus on single adaptive alleles and the conclusion that mutations in the same genes tend to produce similar pleiotropic effects.

Major comments:

1. The major conclusion of the manuscript state that "mutations in the same genes tend to produce similar pleiotropic effects", suggesting that a number of times this does not occur. For instance, the authors comment on the case of PDR3, which does not always produce 'cost-free' adaptation across environments. I believe that, to strengthen and better define their conclusion, the authors should develop a quantitative analysis of the reproducibility of pleiotropic profiles (that considers how many times genes have been found mutated). The heatmaps provided are compelling, but make it hard to generalize on how often, and to what extent, the gene mutated can predict pleiotropy across various environments.
2. In the concluding sentence of the discussion, it is unclear whether the authors are speculating about a role of the strength of selection in determining pleiotropy based on their results, or if that only represents a suggested hypothesis to test in future studies.
3. The method used to identify putative adaptive mutations should be described in more detail. For instance, I seem to understand that only one mutation per lineage is considered 'adaptive'. However, many lineages seem to have more than one mutation. Based on what reported in the method section, the adaptive mutations have been hand-picked based on previous knowledge of selection in the environments of choice ("the list of genes was curated based on those genes' interactions with other identified genes or pathways known to be involved in the adaptation of that specific condition from previous work"). If this assumption is correct, the criteria for such a curation should be specified in more detail.

Minor comments:

1. The term 'Pareto front' is technical and left undefined.
2. The section ' adaptation can be cost-free' only refer to figure 4, (with adaptive mutant lineages from populations evolved in fluconazole), while it comments extensively on mutation isolated in clotrimazonle (reported in Sup. Fig10, not mentioned in the section).

Referees cross-commenting

I tend to agree with reviewers 1 and 3 that, given the focus on individual mutations of this manuscript, more information about their nature is important. On top of the zygosity, I would be curious to know whether mutations predicted to inactivate the gene (frameshifts, stop codon), have different pleiotropic profiles than AA substitutions.

To answer the reviewer's 2 second major point, my understanding is that most of the lines have accumulated other mutations (marked with a '+' sign in Fig4 and FigS6,S10,S11). I suspect, however, that none of these mutations have been considered adaptive given the criteria described in the 'identifying adaptive mutations' session (e.g. mutations in coding regions appearing in more than one clone in a given condition, and with median fitness in the original home greater than 0).

Significance

The manuscript address a question (the emerge of pleiotropy during evolutionary adaptation) which has been extensively studied. However, it does it in a more comprehensive way that previously achieved, including many environments, both haploid and diploid organisms, and by focusing on single adaptive mutations. Most of the conclusions match the ones of previous studies. Perhaps the only exception is represented by the conclusion that individual genes, more than home environments, are proposed to dictate the pleiotropy profiles. However, the fact that mutations affecting the same genes often produce similar pleiotropy profiles is not necessarily unexpected. Overall, the paper is clearly written and can represent a valuable resource for a rather specialized community interested in the origin of pleiotropy during evolutionary adaptation.

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

The objective of the study was to investigate the impact of specific adaptive mutations on trade-offs and pleiotropic effects in haploid and diploid Saccharomyces cerevisiae populations. The authors identified adaptive mutations from strains evolved in three specific home conditions and then conducted fitness assays in up to 12 environments (home/non-home). They highlight that ploidy level plays a condition-specific role in shaping the adaptive mutation spectra. Adaptive mutations showed fitness effects in both home and non-home environments, which were beneficial in some cases and detrimental in others.

Major comments

It would be interesting have an idea of the global mutation rates and spectra in the diploid and haploid lineages across the conditions as well. The S. cerevisiae mutational spectra has been shown to be dependent on the environment and genetic background to an extent but not ploidy. Ploidies differ in terms of just the frequency. How similar/ dissimilar are the overall mutational spectra here? Were there any homozygous mutations in the diploids?

Fitness gains and losses can happen without trade-offs if neutral home mutations are non-neutral in non-home conditions. Can the authors comment on that in this context. Physico-chemically, how different are the home/ non-home environments? How do the fitness effects correlate across the environments in the absence of these adaptive mutations? It would also be useful to know the extent of fitness variance of the populations in the home and away environments, this would aid the reader better grasp the significance of fitness gains/loss.

Minor comments

A summary table of the pleiotropic effects would be very useful as in Bakerlee et al. 2021. Citation errors e.g. Consistent with prior work (Jerison, et al., 2021)... is either incorrect or not referenced.

Significance

The paper is well written highlighting an interesting question, the take home message is incremental at best in the context of the overall literature. In general, my suggestion would be to tone down the conclusions a bit, as the evidence isn't very clear cut .

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

Point-by-Point Response (author’s replies in plain text)

---

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

Summary: Silao et al make the intriguing observation that yeasts that are generally considered less pathogenic are unable to catabolize proline than Candida albicans. They then, in Candida albicans, construct mutants defective for the two key enzymes (Put1, Put2) required to convert proline to glutamate, which they show to be essential for proline utilization as an energy (carbon) and nitrogen source. The authors proceed to untangle the regulatory aspects of proline degradation, including the respective cellular localization of its key enzymes. They then make the important discovery that strains lacking either Put1 or Put2 suffer from a proline-dependent growth defect, which they attribute to resulting defects in mitochondrial metabolism.

The manuscript then goes on to analyze a broad range of infection models including: reconstituted human epithelial skin model, Drosophila, mouse systemic infections, organ colonization in these mice (kidney, spleen, brain, liver and histochemistry of the kidneys) as well as survival when incubated with cultured human neutrophils. Finally, they use yeast cells constitutively expressing yEmRFP (so that yeasts can be distinguished from other host cells) and coated with FITC before incubation with the host cells (which coats the wall of the original cells, but does not spread to progeny) and they go on to perform an impressive set of analyses of C. albicans growth within mouse kidneys both in vivo and ex vivo, exploiting an implanted window together with intravital imaging with a two photon microscope at different time points. The system is impressive and visualizes tissue invasion by hyphal cells beautifully. Finally, they compare the intra vital images from WT and put2-/- cells and show that, as in vitro, put2-/- cells do not form filaments and do not show extensive invasion of the kidney tissue. While the in vivo aspect of the study includes many different models, it finds defects in virulence for different subsets of put mutants and the relative importance of filamentation vs proline utilization for virulence is not conclusively resolved.

Overall, this is an important and timely manuscript, which significantly contributes to the understanding of how proline metabolism intersects with yeast fitness in the context of infections. However, there are several major concerns regarding some of the conclusions drawn from the study. In addition, some general recommendations that would improve the manuscript are provided.

Specifically, the manuscript provides a very detailed description of experiments and observations. However, in several parts it is difficult to follow and the reader needs more guidance about the logic involved in reaching conclusion. Specifically, several aspects of the paper are written for experts in Candida (yeast) metabolism. Here, explaining the rationale for some of the experiments, and providing more background information that is not obvious to a non-expert, is required.

In particular, writing a clear and measured summary sentence at the end of each paragraph and a conclusion paragraph that summarizes key findings in simple terms would help make the manuscript more digestible for readers.

In addition, the impressive microscopy and broad range of in vivo experiments is comprehensive but only adds incremental information relevant to proline metabolism-that filamentous growth in vivo and virulence is reduced in cells carrying some mutations in one or more put genes. However, this broad sweep of model systems and the development of the in vivo imagining system might have more impact in a separate paper focused on the real-time in vivo visualization of kidney invasion.

We thank Reviewer 1 for the extensive list of comments and have endeavored to adjust the manuscript to address all of the major and minor concerns. It is evident that Reviewer 1 clearly understood the significance of the work and we appreciate that the comments are presented in a positive manner intended to improve our manuscript.

Major comments:

1. The main finding that impressed this reviewer is that "removing the ability to catabolize proline, in an organism that evolved to catabolize it, leads to (growth) defects". This point could be better highlighted throughout the manuscript.

Thanks for the comment. We will adjust the text to reflect this suggestion.

1. The authors show that deletion strains for proline metabolism have defects that are important for in vivo pathogenicity. This is an important finding. However, as the manuscript reads now, it suggests that the main findings are that the ability to use proline in the respective host niche is key. Mechanistically, the manuscript revolves primarily around defects that arise when deleting PUT1 and/or PUT2 (i.e., an "unknown" toxicity of proline in the case of put1-/- (or put1-/- put2-/-) and the additional P5C-dependent toxicity for put2-/- mutants; see below).

Yes, the reviewer is correct in that we believe that proline catabolism is necessary to initiate and power hyphal growth, which is coupled to virulence. We have previously shown that upon phagocytosis by macrophages, the expression of Put1, Put2 and even Gdh2 are induced in phagocytized C. albicans cells, which is consistent with the analysis shown in Fig. 2D and Fig. S2B. Consequently, proline, or an amino acid that is metabolized via the proline catabolic pathway, must be present in the phagosomal compartment. However, as we now report, proline inhibits growth of cells lacking the capacity to catabolize it. Although we cannot differentiate the cause of reduced virulence in put mutants, i.e., the lack of energy due to the inability to catabolize proline vs proline toxicity, proline catabolism is clearly important and a robust indicator of virulence. As point 1, we have adjusted the text to make this clearer.

1. In order to claim that catabolizing prolines promotes pathogenicity (as opposed to the alternative hypothesis that the inability to catabolize proline leads to the observed defects), additional experiments would be required. For example, the put mutants would need to be compared with mutants that significantly reduce/impair proline uptake, such as the referenced gnp2 mutant (Garbe et al 2022). While the finding that less pathogenic yeast species are unable to catabolize proline is both intriguing and important, it also remains as is presented as a loose, non-quantitative correlation that only tangentially address the question of whether "proline catabolism is key for pathogenicity".

We have in fact already shown that proline uptake is required to induce filamentation (Martínez and Ljungdahl 2003, Fig. 6). The main point of our current work, which we believe is important and of general interest, is that C. albicans is adapted to use proline as sole energy source, which reflects the environment (humans) in which it evolved. See the response to point 2. Interestingly, the differences in the expression levels of Put1 (off in the absence of proline, induced robustly by proline) and Put2 (low level of constitutive expression, induced robustly by proline) suggest that cells are primed to decrease the likelihood of becoming inhibited by P5C, i.e., the constitutive expression of Put2 is able to ameliorate the potential toxicity of P5C. Regardless, the finding that put1 and put2 mutants exhibit significantly reduced virulence in two host models provides clear support for proline catabolism being key for C. albicans pathogenicity.

1. 238 onwards: The conclusion that "the primary growth inhibitory effect of proline is linked to catabolic intermediates formed by Put1 and that are metabolized further by Put2"does not appear to be fully supported by the evidence. Addition of proline to put1 mutants already reduced OD600 by ~50% (Figure 2); and is further reduced to ~10% when put2 is deleted. This implies that there are two inhibitory effects of proline, not one primary one. At the least, this option should be discussed, including why deletion of PUT1 leads to proline toxicity. The latter is not clear-is it that too much proline accumulates in the cell and this accumulation is toxic? If this is the case, the effect would be expected to be proline concentration dependent. Performing a relatively simple experiment as performed for the put2 mutant (Fig. 3 / S3F) may clarify this issue. Particularly, if the experiment would be coupled with intracellular quantification of proline.

Precisely! Proline toxicity is evident even in put1 mutants, clearly suggesting that proline, without being further catabolized, exerts a growth inhibitory effect (Fig. 3A). We traced this inhibitory effect to decreased mitochondrial respiration (Fig. 3E). There are two parameters to consider regarding the inhibitory effects of proline in put2 mutants. First, the presence of proline induces the expression of Put1 independent of Put2 (Fig. S2C), consequently, the levels of the toxic intermediate P5C increases (Fig. 3B). P5C has previously been postulated to inhibit mitochondrial respiration, which is well-aligned with our analysis (Fig. 3E; see response Point 5). We initially tested whether a proline-P5C cycle, suggested by work in mammalian cells, would play a role in proline-mediated toxicity; however, increasing cytoplasmic pools of proline by supplying high levels of glutamate (which according to work in mammalian cells should efficiently convert to cytoplasmic proline) did not occur; we did not see glutamate-enhanced Put1 expression (Fig. 2D, S2A, S2B). We agree with the reviewer with respect to the suggested experiment, and have monitored growth of put1 in media with different proline concentrations. The results are incorporated in the revised Fig. 3.

1. The caption "P5C mediates a respiratory block" is misleading, as the evidence is not that compelling: Although P5C increases in put2, but not in put1 mutants, and given that both single mutants experience a proline-dependent respiratory defect (Fig. 3E), the results suggest a more complex relationship.

Previous work using pure P5C (Ref. 36; Nishimura et al) showed that it targets respiration, hence the caption “respiratory block” in the header. In mammals, PRODH (Put1) physically interacts with mitochondrial respiratory complex II in the inner mitochondrial membrane (line 89-90), while P5CDH (Put2) is in the matrix. The put1 mutation might affect basal activity of the respiratory chain resulting in lowered respiration, which may compound when proline accumulates in the mitochondria. The inhibitory mechanism remains unknown, and in going forward we have begun characterizing various GFP-tagged respiratory complex components in put1 mutants and in strains co-expressing Put1-RFP (for interaction studies). The results are out of the scope of this current work.

1. The virulence assays and in vivo experiments do not present a unifying view: in Drosophila put2∆∆ is less virulent than put1∆∆, which appears similar to put3∆∆. Given that put2 mutants grow slowly, likely because of P5C inhibition, this seems logical. However, in mice, put3∆∆ remains highly virulent while put1∆∆ and put2∆∆ results for survival are mixed. Furthermore, in 4 mouse organs, put1∆∆ and put2∆∆ are not significantly different from one another but are different from wt, while put3∆∆ has no significant reduction in CFU. Kidney histology shows very little invasion by put1 and put2 and more by put3, but visually put3 appears to invade much less than the WT, and the human neutrophil experiment shows effects of put2 or put3 but not put1. This leaves the reader rather confused. It may be worth discussing the reasons for different results in different models. Is the availability of proline in each of the organisms and organs similar?

We thank the reviewer for these thoughtful observations, however, we note that all of the diverse assay systems employed provide a clear and consistent indication that the inability to completely catabolize proline significantly reduces virulence. This is well-aligned with our previous data regarding the need for proline catabolism to escape macrophages (Silao et al, 2019). The requirement for Put3 may not be very strict since the Put enzymes are still expressed in the absence of Put3 (Fig. 2D/S2A/S2B), indicating the activity of additional regulatory factors; hence, this may explain why the put3 strain behaves like wildtype in the murine model (Fig. 5B). The dispensability of Put3 in the murine model could be due to a lower neutrophil count and that murine neutrophils exhibit a lower affinity for fungal cells as compared to human blood (Machata et al., 2020, Front Immunol). The more pronounced requirement of Put3 to survive in whole human blood and when co-cultured with human neutrophils could indeed be linked to the need to rapidly derepress PUT1/PUT2 (and even other target genes) as suggested by the global RNASeq analysis that shows that proline catabolism is a core response of C. albicans during neutrophil interaction (Niemiec MJ et al., 2017, BMC Genomics). In Drosophila, a well-established model to study innate immunity, the presence of hemocytes that fulfill the equivalent functions of neutrophils and macrophages could explain the increased requirement for Put3. In summary, although it is impossible to know the precise mechanistic basis underlying the observed differences, we believe it unreasonable to expect that all mutations behave identically in each virulence model. In fact, differences considered trivial such as the use of mouse background can have profound effects on virulence. Presumably the differences we report are due to the specific nutrient composition (proline and metabolites feeding into the proline catabolic network) and physical parameters intrinsic to each model. For instance, Lionakis et al. (2013) suggested that filamentation occurs faster in the kidney compared to other organs, such as the liver/spleen, indicating the presence of kidney-specific cues that drive infections of this organ.

1. The ex vivo and in vivo analysis of the dynamics of C. albicans growth in the host is visually impressive, but it distracts from the focus of the paper and the metabolic findings. Showing that put mutant cells do not form filaments in vivo (as in vitro) does not add much conceptually to the paper. Furthermore, this lovely advance in in vivo visualization is lost at the end of this paper and the authors should consider whether it might fit better in manuscript that could really highlight the in vivo visualization approach.

We appreciate this comment. Indeed, our lab is at an advanced stage of completing a manuscript focused on the use of intravital and clearing microscopy to follow the onset of an upper urinary tract infection (UTI) in a murine candidemia model. However, our ability to visualize in 3D the onset of an infection in a living host is not a trivial achievement and we were impressed that it provided a clear answer as to whether a single C. albicans cell can initiate an infection and undergo morphogenesis leading to hyphal growth. Furthermore, we tested a put2 strain, the growth of which is highly sensitive to the presence of proline, and found that it did not exhibit filamentous growth. This clearly shows that cells colonizing the kidney are exposed to an environment that requires a functional proline catabolic network to exhibit filamentous growth, a characteristic of renal infections. Our results are consistent with the kidney being a metabolic hub for arginine/proline biosynthesis, which likely increases the levels of these amino acids in this organ.

1. The discussion of cells stained with FITC and expressing yEmRFP does not clearly point out that the FITC is only an indicator for those cells that were used to innoculate the tissue and that finding cells without FITC indicates that they are mitotic progeny, indicating that they have been dividing. The authors clearly understand this, but a naive reader may miss this important point if it is not stated explicitly.

We have adjusted the text to explicitly clarify this.

Minor comments:

1. Throughout: what is the distinction between utilization of proline for C or for energy? These terms seem to be used interchangeably.

C. albicans is heterotroph that can use proline to generate biomass (gluconeogenesis, etc) and its catabolism generates sufficient amounts of ATP to power growth. Thus, when proline is used as sole carbon source, it can also serves as the sole energy source. In the text, we have tried to be consistent using “carbon source” when discussing proline as a component of growth media, and “energy source” when discussing proline catabolism.

1. Introducing the schematic in Fig. 2A at the beginning of Figure 1, would help explain proline catabolism before delving into the growth experiments that rely upon this framework. This should include an explanation, for readers less familiar with the metabolic issues, of the main limitations to catabolizing proline, and the key issues for being able to use proline for nitrogen, carbon, and energy (potentially indicated in the overview figure, e.g. pointing towards gluconeogenesis etc.).

We have considered the reviewers suggestion, however, we believe that the placement of the schematic in Fig 2 is appropriate as is, and where it will hopefully enable readers to more readily grasp the strain construction and experiments documented in Fig.2.

1. Saccharomyces can only grow on proline as a nitrogen source, but not as energy/carbon source. Could the authors briefly mention or discuss why this is the case? This is not clearly apparent after reading the manuscript and it leaves the reader confused and trying to understand if the fact that proline is required for carbon utilization is a new finding of this paper or was already known. Do the authors think this is tied to the presence of complex 1 components in C. albicans that are not found in S. cerevisiae. Is this consistent for the pathogenic, but not the non-pathogenic yeasts analyzed in figure 1?

We have adjusted the text to clarify our thoughts regarding this. Indeed, we do believe that a major reason for the ability of C. albicans to efficiently grow using proline as a sole energy source is the presence of Complex I. However, C. glabrata appears to be able to grow well using proline as sole energy source despite apparently lacking Complex I. Consequently, alternative NADH dehydrogenases exist in C. glabrata, but how this is coupled to energy metabolism will require additional work that is out of the scope of the present work.

1. 100: While Gdh2 is apparently an important enzyme for generating ammonium, why is it not necessary for macrophage escape and virulence as shown in reference 18? A recent paper from Garbe et al (ref 12) suggests that Gnp2 is the major proline permease in C. albicans and what is known, and not known, about proline uptake would be good to mention, given that PUT gene functions require that proline enters the cells.

We have recently shown that ammonia generation by Gdh2 is dispensable for macrophage escape and documented that phagosome alkalinization is not a requisite for the induction of hyphal growth (Silao et al. 2020). We have referred to the work of Garbe et al., which is consistent with our previous work (Martinéz and Ljungdahl, 2004) where we reported that proline-dependent filamentation is dependent on Csh3. Csh3 is an ER membrane-localized chaperone responsible for catalyzing the proper folding of amino acid permeases, in csh3 null mutant strains, amino acid permeases accumulate in the ER as non-functional unfolded aggregates. Consistently, we have tested and found that proline-induced Put2-GFP expression is dependent on Csh3 (unpublished), clearly establishing that the regulatory effects of proline are dependent on its uptake. We have not generated a gnp2-/- strain, but suspect that we could find growth conditions where such a mutant would be refractory to proline induction. We have adjusted the text to include this information.

1. 116: Is the "low sugar environment of the host" referring to a specific niche, such as the GI tract, or human blood? Compared to most natural environments, glucose is abundant in the host, e.g., at ~5 mM, it is the most abundant metabolite in blood, and similarly, in the GI tract, levels can go beyond 50 mM glucose (see e.g. PMIDs 34371983, 21359215). Or is this comment indicating that the in vivo sugar concentration is lower than that in common lab growth media? Please spell out the niche/concentration for clarification - and compare that to other niches that are considered "high sugar environments".

We have adjusted the text to clarify our statement. The natural environment of C. albicans is the human host. Virulent infections are not within the GI with high sugar content, but rather result when C. albicans cells successfully cross into the blood with a relatively low glucose (5 mM), which importantly is a level that does not effectively repress mitochondrial function. A major point of our recent work is that laboratory experiments with C. albicans growing on YPD or SD with 2% glucose (111 mM) examine growth of cells with repressed mitochondrial functions.

1. 123: "proline as sole energy source" - suggest "is the source of carbon, nitrogen, and energy"

The text is adjusted (see response to Minor Point 1).

1. 142: it is worth noting to readers that C. neoformans is a basidiomycete and thus VERY distant from the other yeasts studied here-it is in a different major phylum of fungi.

Again, thanks for this suggestion, the text is adjusted. We included C. neoformans since the role of proline catabolism has been characterized and linked to its pathogenicity (reviewed in Christgen and Becker, 2018, Antioxi Redox Signal, Ref. 1).

1. 143: Here it is implied that put1 and put2 mutant strains do not grow on SPD, but this is not stated explicitly.

The put1 and put2 mutants are unable to grow in/on all media containing proline as sole nitrogen source. The phenotype is very tight that we were able to exploit this as a selection phenotype for reconstitution (Fig. 1A). We have adjusted the text to make this clear.

1. 151: The abbreviation SPG is not explained in main text. This was explained in the methods (1% glycerol as primary carbon source).

As suggested, we have defined SPG in the main text.

1. Paragraph 156 onwards: this section is particularly hard to read and very dense. Also, it is difficult to understand the significance of these experiments for the overall findings of the paper. Please at least provide a small conclusion / summary at the end of the paragraph that puts the findings into perspective.

We have adjusted text to make it more accessible.

1. Figure 2 C: simplifying the scheme (e.g. lots of redundant information, P2 and Mito - just give it one name) would help. This figure may be better in the supplementary material.

The schematic of our subcellular fractionation study uses standard designations routinely used by the cell biology community. We believe that its inclusion will help readers judge the how we mapped the intracellular localization of the reporter proteins, which is essential to understand the proline catabolic network.

1. Figure 2B: It is not directly apparent from the micrographs that Put1-RFP localisation is mitochondrial. Co-localisation of the RFP with a mitochondrial dye (e.g., mitotracker) or something similar is required to validate it.

We have previously reported that Put2 is a bona fide mitochondrial protein (by confocal microscopy, subcellular fraction, and co-localization with Mitotracker (Far Red) (Silao et al., Ref 17). The fact that the Put1-RFP associated fluorescence exhibits a distinct mitochondrial signature, is spatially exclusive and exhibits no overlap with the cytosolic pattern of Gdh2-GFP, co-fractionates with Put2-HA and the mitochondrial marker Atp1, should suffice to confirm that Put1-RFP is a mitochondrial localized protein.

1. Throughout the manuscript (figure legends): Suggest using "mean" instead of "Ave."

We have adjusted the legends.

1. 175: According to the 'Yeasttract' and 'Pathoyeasttract' databases, Put1 regulates at least 36 and 22 genes, in S. cerev. and C. alb., respectively (based on DNA binding and/or regulatory changes). The only gene in common between these two lists of genes is PUT1. Thus, it is quite likely that Put3 regulates many other processes that explain its function and that its major function may not be only to regulate Put1.

We assume that the reviewer is referring to Put3 (instead of Put1). Yes, Tebung et al. (2017) suggested that Put3 also regulates other genes. However, their data show that C. albicans put3 mutant was unable to grow in medium (YCB+Pro) compared to SPD (2% glucose as carbon source) where proline is used merely as a nitrogen source (Tebung et al., Fig. 3A). Our data in Fig. 1C shows that a put3 null strain exhibits residual growth on SPD, which aligns well with the expressed levels of PUT enzymes (Fig. 2D). Our conclusion is that despite being essential for rapid proline-dependent derepression of proline catabolic genes, Put3 is not the only transcription factor operating at the promoters of the PUT genes.

1. 175: Is it clear whether the Put3-independent mechanisms are positive or negative with respect to Put1?

We have accumulated evidence that an additional transcription factor positively regulates PUT1 expression and have a manuscript in preparation to describe this factors. The manuscript will focus on the Put3-independent regulation of PUT1, PUT2, and GDH2 expression.

1. 218: Suggestion: "growth was indistinguishable".Unless growth curves or growth rates are provided and if one time-point data are the basis for this point, than "rates" is not a relevant term.

The reviewer is correct; we will adjust the text accordingly. We have performed growth assays in a multi-well microplate format (Bioscreen) and found that the growth rates are not statistically different between WT, put1, put2, and put1 put2 strains in the presence and absence of proline in SD with 2% glucose. This is consistent with glucose repression of mitochondrial function, i.e., proline toxicity depends on derepression of mitochondrial function.

1. 256 onwards: did the authors test if the ROS scavenging effectively reduced ROS? i.e. does the luminol-HRP assay yield less ROS in +proline +scavenger treatment? This is necessary to effectively conclude that the growth inhibitory effect of proline is due to blocking respiration.

Indeed, we used NAC as a control in the luminol-HRP system and we saw reduction in ROS formation. In fact, this is the underlying reason why we used high levels of NAC for growth rescue (in Fig. 3D). We include the control data as Fig S3F.

1. The Figure captions are extremely lengthy and detailed, making it cumbersome to find the relevant information. Suggest moving some of the information, such as additional experimental details, into the methods section.

We have streamlined the figure legends.

1. 277-301: Phloxine is not exclusively a live/dead cell indicator-it is an indicator of metabolic activity. In Scerev. and Calb. it also indicates slower growth, opaque growth, and it has been used as an indicator of aneuploidy in C. glabrata (https://journals.asm.org/doi/10.1128/msphere.00260-22) and of diploids vs haploids in S. pombe. The colonies illustrated aer made up of many live cells, and thus the section "Defective proline utilization is linked to cell death" needs to be presented more carefully. In addition, it appears that this section shifts from using defined medium to using rich medium and 37C instead of 30C. Why was this shift necessary?

The reviewer is correct that phloxine (PXB) has been used to identify opaque growth (EFG1-dependent). However, the fact that the accumulation of PXB in the put mutants is evident in both SC5314 and cph1 efg1 backgrounds (Fig. 3G and Fig. S4C) suggests that we are not assaying opaque switching. We mention that we have observed an increase in the number of PI+ cells in put mutants under similar conditions, but as we pointed out, we were unable to reliably quantitate this by FACS due to the clumping of put mutants. Zheng et al 2022, the paper cited by the reviewer, used PXB to assess the ploidy of C. glabrata strains, but their assay was developed using 5 μg/ml PXB, half of the concentration we used. The homogenous accumulation of PXB as the macrocolonies grow (Fig. 3G), suggests that the accumulation is not a consequence of spontaneously occurring ploidy variations. Thus, we believe that the accumulation of PXB does indeed reflect enhanced cell death. The point here is to trace the consequences of proline toxicity and to test the dependency on mitochondrial function. We used complex media, which contains multiple nitrogen sources (amino acids, peptides), to specifically highlight the contribution of proline catabolism in the fitness of C. albicans. The put1, put2 and put1 put2 mutants grow normally on YPD+PXB (30 oC) without accumulating the dye; we only observed visible PXB uptake in put2 after 2-3 days in mature macrocolonies. We attribute the gradual increase in PXB accumulation to be a consequence of glucose becoming limiting, derepressing mitochondrial functions, a requisite for proline toxicity. Consistently, the accumulation is more evident in cells grown on non-fermentable C-sources (Fig. 3G and Fig S4C).

1. 295-301: Related to the point above, these results are hard to interpret due to the switch from defined medium in all prior experiments to rich growth medium here. Also, it is not clear why a 48h old YPD culture was chosen to show that the degree of PI staining correlates with mitochondrial activity - is this due to the culture age? It would be more clear to image cells grown on glucose vs. glycerol/lactate, or under repressive / de-repressive glucose concentrations (e.g., as shown in Fig. S4C where a PI+ difference is apparent for 0.2% glucose vs. 2% glucose at 30 oC).

See response to Point 19 for our rationale to switch to rich medium. We have adjusted the text to enhance its readability. In liquid YPD, all strains grow, however, we noticed that the put mutants tend to flocculate (sign of stress in yeast) when cells enter stationary phase, giving rise to erratic OD readings, particularly evident in the put1 mutant. At 48h, the cultures become dense and cells experience glucose limitation, derepress mitochondrial functions and exhibit maximal flocculation (Fig. S4D). In put mutants, the derepression of mitochondrial function results in proline sensitivity. We tested the notion that this would also increase cell death, which it does, see Fig. S4E.

1. 313-14: The statement 'the invasion process was dependent on the ability of cells to catabolize proline' doesn't take into account that put mutant cells are defective in filamentous growth irrespective of their utilization of proline...and like the efg1 cph1 double mutant.

Proline-induced filamentous growth is dependent on the catabolism of proline, which activates Efg1 and consequently the hyphal growth program. In Fig. 4A we show that put mutants grown on Spider media, initiate filamentation (as evidence by wrinkled colonies) but do not grow invasively (no halo). In Fig. 4B we developed and used a novel invasion assay to assess growth through a collagen plug. Similar to the control cph1 efg1 mutant, the put mutants exhibit drastically reduced capacity to penetrate through the plug, and reach the D10 media in the transwell (D10 = DMEM with 10% FBS). However, it is important to note that although these results are linked to two distinct processes - the filamentation defect of cph1 efg1 is due to the inability respond to multiple filamentation cues (e.g., CO2, 10% FBS, etc.), whereas the filamentation defect of the put mutants is linked to the inability to catabolize proline and to its toxicity. Clearly, the WT strain relies on proline catabolism, coming from one or three possible sources of proline (see response to Reviewer 3): 1) DMEM/F-12 medium used in the PureCol EZ Gel; 2) diffusion of nutrients up through the collagen from the recovery medium DMEM supplemented with 10% FBS; and 3) the proteolytic breakdown of collagen. Also, in contrast to the put mutants, WT cells are refractory to inhibition by proline.

1. 316-327: The results of the experiment described can only be interpreted as an effect of proline catabolism if the three strains (efg1 cph1; put1; put2) have similar growth rates as yeast cells in vitro. Why weren't the cells competed directly (efg1 cph1 vs put cells)?

We believe that the relevant comparisons are to WT. We recovered cells from the top of the collagen (see Fig. 4B inset) to monitor their ability to survive and grow on top of the collagen. We found that the ability to catabolize proline enables WT and cph1 efg1 cells to grow equally well (recovered similar ratio as starting input). This was not the case with the put mutants, they did not grow as well and almost 100% of the cells recovered were WT.23.

Fig 6: The logical order of the experiments, and in the text, is: 1) 4 h window, 2) 26 h window and then 3) ex vivo. The cartoon in 6B should be in this order as well.

Thanks for bringing this issue up. We have adjusted the figure and text placing the schematic time-lines in proper order.

1. 337: it is not clear what the 'direct exposure...' is trying to tell us. Can this be made more explicit?

The direct exposure means that the fungal cells are in contact with the culture media at the edges/border of the 3D skin model (see schematic diagram). Hence, fungal cells are in direct contact with 10% FBS, facilitating the observed filamentous growth. The inability of the put mutants to invade the skin model should be evaluated at the center of the artificial epithelium where there is likely a local increased concentration of proline stemming from the proteolytic activities associated with fibroblasts and keratinocytes.

1. 340-346: Here proteins with high proline content were used to ask if they could be induce transcription of PUT1 or PUT2 RNA and protein. This experiment is designed only to test the role of these proteins to induce utilization of nitrogen, as glucose is included in the medium. Given that these proline-rich proteins need to be lysed by proteases before they can be imported, and since no import pathways were tested, the results appear to tell us that mucin is more readily digested to peptides that contain proline-but why that is the case is not clear and how it relates to proline utilization is also not clear.

We thank the reviewer for raising this important point. First, we monitored protein not mRNA levels. We will adjust the text to provide better context for this experiment. Briefly, these experiments were initiated as we were perplexed as to why the wildtype cells took such a long time (14 days) to fully invade the collagen matrix (Fig. 4B); we naïvely assumed that fungal cells would secrete proteases to degrade the collagen and assimilate the liberated proline. In going forward, our experimental strategy was to incubate various proteins with a dense culture of cells in HBSS medium (pH 7.4) supplemented with low glucose (3.8 mM) and lactate (0.83 mM). This condition mimics interstitial fluid, where most broad range proteolytic enzymes are inactive or at least operating suboptimal. The results were clear; with the exception of mucin, the proteins did not stimulate Put1 or Put2 expression. We conclude that host-dependent processes play an important role on the release of the amino acids/peptides from these high-proline content proteins (see line 531-553 for discussion). The capacity of mucin to efficiently induce Put1 expression is interesting since mucin is abundant in the gut where systemic infections are thought to originate. It is important to be cautious here, we used a commercial mucin preparation (Sigma, 2 batches) that may contain degradation products, e.g., proline-rich peptides, that can easily be assimilated by C. albicans. Put1 expression is an excellent readout for proline uptake since its expression responds tightly to the presence of proline derived from exogenous supply or from intracellular conversion (Fig. 2D, S2A, S2B).

1. 363-369 An alternative is that Put3 induces different proteins important for growth.

We included this possibility in the revised text.

1. 379-380-the conclusion for this paragraph is somewhat of an overstatement as there is no analysis of the degree to which proline utilization is a predictor of virulence. It simply shows that put mutants affect the ability to survive in neutrophils.

We have adjusted the text.

1. Discussion: The statement that "S. cerevisiae" evolved in high sugar environments is debatable. The natural niche could well be forest soil and tree bark, or insect/wasp guts with arguably little glucose around.

The reviewer is correct, S. cerevisiae can be isolated from diverse environments with variable sugar contents, but it is the capacity to deal with high sugar environments that makes this yeast stand out in comparison to Candida spp. The unique attribute of S. cerevisiae have been exploited and truly benefited humankind in making alcohol and bread. We have amended the text to state this more accurately.

1. 469-470-how strong is the 'correlation' between the ability to utilize proline and virulence? Given that different mutants had different effects in different models, this seems like a very loose 'correlation'; it would be good to have some quantitative measures to make this claim.

We have used directed genetic approaches to determine whether a gene/protein is essential for virulence by testing them in currently available infection models. It is important to note that all virulence assays provided a consistent and clear read-out, namely that the inability to catabolize proline significantly reduced the expression of virulence characteristics. Presumably the differences we report are due to the specific nutrient composition (proline and metabolites feeding into the proline catabolic network) and physical parameters intrinsic to each model. In fact, the expression of virulence factors (i.e., hyphal growth) can significantly differ in different organs within a same mouse model (Lionakis et al., 2013) and that virulence outcomes can change depending on mouse background. We fail to see how this can be viewed as loose. This has not been shown before. Please refer to our response to major point 6.

1. 500: Was the experiment was done in larvae, and not in adult Drosophila? Fig 5 legend says flies and shows a picture of a fly and larvae are only mentioned much later in the text.

These experiments were performed using adult flies. We now include a reference regarding the levels of arginine in hemolymph in both larvae and adult Drosophila (Priyankage et al., 2012; Anal Chem).

1. 512:Why is it presumed that proline accumulates in the mitochondria in put1 mutants? How strong is the presumption?

Despite a great deal of efforts in many labs, the mechanism of proline transport across the mitochondrial membrane is not known. What has been shown in mammalian and plant systems is that proline can readily enter and accumulate in mitochondria where it is catabolized. (https://link.springer.com/article/10.1007/s00425-005-0166-z; https://www.sciencedirect.com/science/article/pii/0003986177902089). Our presumption that proline accumulates in the mitochondria is based on our finding that proline inhibits mitochondrial respiration when Put1, catalyzing the first oxidation reaction, is absent.

1. 539: why are MMPs important for digestion of collagen? This is not clear at this point of the Discussion.

In mammalians cells, some secreted MMPs have collagenase activity (e.g., MMP-1) that degrade proteins comprising the extracellular matrix, which releases proline. We emphasize this since the 3D skin model is comprised of dermal fibroblasts and keratinocytes that are known to secrete MMPs (Ref. 69).

1. 574: Concluding sentence of this paragraph seems unsubstantiated. There are at least two defects in put2 strains-hyphal growth and growth in general, presumably because of P5C accumulation.

See response to point 21. Proline-induced filamentous growth is dependent on its catabolism, which activates Efg1 and consequently the hyphal growth program. However, there are many potential cues in hosts that could induce hyphal growth in situ. Our finding that strains unable to catabolize proline do not filament, indicates that proline is a key modulator of virulence.

1. Fewer abbreviations would make the manuscript easier for non-experts to read. For example, P5C is not defined in the abstract. Furthermore, if an abbreviation is not used more than 3 times, it is not necessary to provide it (e.g., mammalian proteins in the last paragraph).

We have adjusted the text.

typos:

1. 82: should read 'is restricted to the mitoch...'

2. 102-103: should read 'to evade macrophages'

3. Fig. S4F is mislabelled as Fig. S4G.

Thanks!

**Referees cross-commenting**

Overall, we stand by our initial assessment of the study. However, we were not aware of previous studies that investigated proline utilization in yeasts, as noted by Rev # 2 (https://onlinelibrary.wiley.com/doi/epdf/10.1002/yea.1845). The current study suggests that using proline as an energy/carbon source is more wide-spread, beyond pathogenic yeasts. Further, the C. albicans strain they used for this study (ATCC 10231) was apparently unable to grow on proline in the quoted paper. In light of this, we think the authors should reference this study, tone down the claims about the clear correlation of pathogenicity and proline utilization, and address this apparent discrepancy with the indicated Candida albicans isolate. We note that our review considered this a paper mostly of interest to specialists.

Although other non-pathogenic fungi have been shown to use proline as pointed out by Reviewer 2, this metabolic attribute has not been previously tested in members of the pathogenic Candida spp. complex. We have included the reference and included a statement that many fungi, isolated from diverse environmental niches, can use proline as a carbon source.

Reviewer #1 (Significance (Required)):

1. The advance in this paper is conceptual for the proline utilization connection to virulence in a range of species and technical for the in vivo microscopy. Limitations are that the conceptual advance is based only on qualitative work in figure 1 and that the animal studies do not provide a conceptual advance, although the technical advance of in vivo visualization of kidney tissue is impressive and (to the knowledge of this reviewer) quite new as the only prior work was in mouse ears.

In response to the reviewer’s comment regarding Fig. 1, although it is qualitative, it is very reproducible. We even tried several clinical isolates of S. cerevisiae and observed consistent behavior to the standard laboratory strains (i.e., they do not grow on SP medium where proline is used as sole carbon/nitrogen/energy source). We tried to quantify growth of all strain in liquid SP medium at 30 oC using a TECAN microplate reader, but then the results show very erratic reading among species (and replicates) as each behaves differently; C. tropicalis, C. krusei, and C. parapsilosis form pseudohyphae and clump readily, while C. albicans forms hyphae and pseudohyphae.

2.The work fits well as an extension of the body of work from the corresponding author's lab with additions from the labs with expertise in models of infection.

1. People interested in yeast metabolism and pathogenic yeast virulence will be the audience for this paper and as written it is for a specialized audience interested in pathogenic yeast metabolism and, perhaps, (although not mentioned at all in the text) for those who want to try PUT gene products as new drug targets.

This was actually mentioned in the last paragraph of the discussion (line 581-582).

1. Reveiwer expertise is in pathogenic yeast biology and yeast metabolism. Little expertise in high tech microscopy.

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

The study is part of the continuous work by the authors to dissect the mechanism of utilization of proline as a carbon source in Candida spp. In particular, this work shows that the inability to process proline leads to accumulation of the toxic intermediate P5C and subsequent inhibition of mitochondrial respiration and toxic effect on the cells. Furthermore, the study demonstrates that proline utilization is important for C. albicans kidney colonization. The experiments are meticulously designed and the study adds to the overall understanding of the metabolic utilization of proline as a carbon source and its potential relevance for infection.

I find this work interesting, but the role of Put1 and Put2 in proline utilization is not particularly novel. The novelty here is the subcellular localization of the two proteins. Also, the importance of proline utilization for infection is unclear. The host-pathogen interaction assays are ambiguous as each assay gives different result. Lastly, the authors try to generalize the importance of use of proline as a energy source by other Candida spp.. This is not very surprising, given that it has been reported previously by others (example DOI: 10.1002/yea.1845) and that many pathogenic or closely related to C. albicans species use various amino acids, not only proline, as a carbon source.

Yes, as reviewer 2, we are not surprised that many of the pathogenic members of the Candida spp. complex are able to use proline, but this needed to be checked. The fact that proline can be used as a sole carbon/nitrogen/energy source clearly set them apart from the paradigm yeast S. cerevisiae. A major question is what amino acids are important in the context of the host? To assess this, we have used mutations that specifically block proline utilization. Our past studies demonstrating that proline catabolism is rapidly activated in C. albicans cells phagocytized by macrophages indicates that proline is present in the phagosomal compartment. Furthermore, put mutations clearly affect virulence in flies and murine systems. We are at a loss to understand why the reviewer believes that our data, which consistently shows that proline catabolism is important, is ambiguous.

The expectation that all three mutant strains, i.e., put1, put2 and put3, would behave identically in the different infection models reflects an unnuanced view of how infection works. In fact, differences considered trivial such as the use of mouse background can have a profound effects on virulence. Consequently, it is striking how the diverse infections models consistently and unequivocally demonstrate that proline catabolism affects virulence. Also, it should be appreciated that we are not testing mutations affecting proteins with many overlapping functions, where it may be appropriate to challenge claims as to their direct role in virulence. Here we tested mutants that lack the enzymes that catalyze proline utilization. A more reasonable expectation is that the virulence is commensurate to the specific nutrient composition of model systems (as asked by reviewer#1), which can fluctuate among models (see our response to the major comment 6 of reviewer 1). As it is not practical to precisely test the proline levels in the models, we have worked to identify and focus on critical phenotypes that can be analyzed in vitro. Our findings provide the basis for understanding the virulence and growth properties of the mutants in the context of the complex infection models.

Moreover, the authors take C. albicans as an example to demonstrate the role of PUT in invasion and infection. Proline is known stimulus for hyphal growth in this species, but many other Candida spp., including C. auris, do not filament. So how, aside from supporting growth, proline is linked to infection in these species? I think the authors oversell the importance of proline in Candida spp. pathogenesis and should tone this part down or remove completely. A new story that validates the importance of PUT in non-albicans species can bring clarity to why and where proline is critical for survival and infection.

The fact that proline supports growth in the host environment is one of the critical aspects of our work. The lack of appreciation for this finding represents a common misconception in infection biology. It is not just the ability to gain access to a host and initiate an infection that counts, it is equally important to sustain growth and to thrive within the host. Thus, the adaptation to the host environment is critical. Here we document that proline catabolism not only initiates but sustains an infection acting as a critical carbon/energy source. The inability of the put1 and put2 mutants, which are sensitive to proline, to grow and infect multiple models clearly suggests the substantial quantity of proline is accessible. Also, we have constructed C. glabrata (Fig. S1C) and C. auris (not shown) strains that lack the ability to catabolize proline, and are currently characterizing the virulence properties of these strains. This is out of the scope of the present study.

Major comments: I am not convinced by the data that proline is important to initiate infection. Candida infections of the kidney occur only at late stages of sepsis. The authors need more compelling data to prove that proline is important for infection in the host.

Again, not sure why there is such skepticism here, regardless of whether kidney infections occur late, the fact that in contrast to WT, we do not observe put mutants filamenting, clearly suggesting that the capacity to catabolize proline plays a role in the expression of virulence characteristics of C. albicans. Based on our findings using IVM, which provides 3D information, we can at least conclude that a single isolated C. albicans cell can initiate hyphal growth, initiating a point of infection. In addition, our newly added whole human blood data suggests that proline catabolism is required for survival in the blood; human blood contains high amount of proline, arginine, and ornithine that are all catabolized via the proline catabolic network.

Minor comments: I find the manuscript difficult to read and the discussion part is overly long. Some streamlining and adding a bit more explanation for the rationale of each experiment will make the work easier to follow. Some language/style needs refining as well.

We have attempted to take this critique into account during the revision of the manuscript and have streamlined the text and added explanations regarding the rationale underlying our experimental approaches.

**Referees cross-commenting**

In this manuscripts the authors clarify the cellular compartmentalization of steps in proline catabolism. However, it is not novel that proline is a valuable carbon source. The role of proline utilization for establishing or progression of infection remains ambiguous even after the authors provide different in vivo results. The overall significance of the study is limited.

Please refer to our comments below. We do not understand that the reviewers apparently question the obvious role of proline utilization facilitating virulence.

Reviewer #2 (Significance (Required)):

The strengths of this study are in the experimental design and variety. The data is well presented and visualized. The limitations are as pointed above - I find it especially difficult to figure out where, in a real infection scenario (e.g. breach of the gut barrier and entry into the bloodstream) proline will be the primary energy source. To me the significance of this work is minor.

C. albicans is the primary human fungal pathogen placed under the “Critical Priority Group” by WHO and yet our understanding of nutrient assimilation in this fungal pathogen is only a fraction of what is known in the model yeast S. cerevisiae, which has proven not to be the best paradigm for understanding the regulatory circuits operating in human fungal pathogens. This manuscript, as well as other recent publications, have revisited and corrected earlier assumptions regarding C. albicans growth, providing novel information that reflect important regulatory differences specifically relevant to the life of C. albicans in the host. For example, had it not been for the recent findings (Ref. 10, 18, 31) that show that proline utilization in C. albicans is not subject to nitrogen catabolite repression (NCR) and that glucose represses mitochondrial function, the perception in the field would remain that C. albicans cannot utilize proline as a carbon and/or nitrogen source in the presence of a “preferred” source of nitrogen, which is applicable in the blood that contains high concentrations of possible sources of carbon and nitrogen. Furthermore, the low but constitutive expression of Put2 and the tight highly responsive Put1 expression in response to proline (Fig. 2D, S2A, S2B), suggest that C. albicans is well equipped to productively anticipate proline availability depending on the host status, entirely consistent with its “opportunistic” character. The many incorrect and previously held assumptions regarding C. albicans, uncritically propagated in several influential reviews, likely have hampered efforts to develop novel antifungal therapies. We do not understand, nor accept the view that a more precise understanding of the proline catabolism is incremental.

The type of question raised by the reviewer is exactly what we hope to achieve in the future but to get there we have to have correct assumptions in place, and this is only possible if we have a more thorough understanding of the regulatory mechanisms driving proline utilization in C. albicans. The idea that certain proteins are refractory to degradation by C. albicans suggest that other external factors are triggering the release of amino acids from these proteins. This work however, suggest that proline is likely accessible in the gut due to the presence of proline-rich proteins like mucin (Fig. S5A/B).

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

The manuscript of Silao et al. describes an in-depth investigation of the role of Put1 and Put2 enzymes in proline catabolism and virulence in Candida albicans. This is an extension of previous work in this system. The basic biochemistry and genetics are solid and support the role of these enzymes in the proposed pathway and provide evidence that the build up a toxic intermediate in the absence of Put2 is likely involved in the poor growth of the strain when proline is the only carbon source.

Note that we observe the toxic effects of proline even when it is not the sole carbon source, however, and importantly, toxicity is dependent on mitochondrial function, which is repressed by high levels of glucose. Proline toxicity is observed when glycerol/lactate are present as carbon sources in addition to proline. Under these conditions, mitochondria are not repressed and exogenous proline impairs growth, particularly evident in put2 cells that accumulate the toxic intermediate P5C.

The conclusions regarding its role in virulence are less convincing, particularly the data derived from the collagen invasion assay, the ex vivo skin model and the ex vivo/in vivo imaging. The survival and fungal burden assays support a modest role in virulence and a modest reduction in infectivity (although the presented data for survival does not have statistical significance data reported for the kaplan analysis.

See below for response regarding collagen assay. We have included the significance values derived from Kaplan analysis in the revised Fig. 5B.

The manuscript is clearly written. The methods are well described.

**Referees cross-commenting**

I remain unconvinced of the broad significance of the advances and stand by my assessment that this is for the most part a reasonable study but does not move the field forward. The novel technical aspects are either extensions of previous in vivo imaging or are not well controlled (collagen invasion assay)s.

See below for response.

Reviewer #3 (Significance (Required)):

This is a detailed study of an area that is fairly mature and thus will be of interest to those in the field but does not represent a large advance and is thus truly incremental.

See below for response.

Major limitations of the work are as follows. First, the collagen invasion assay may be flawed. The recovery media is made with DMEM which is a medium that lacks proline and is fairly stringent. Control experiments need to be done to be sure that the mutants grow in the recovery medium. Second, the data from the RHE model are hard to interpret since so few cells are present in the tissue. It is hard to see if there are few filaments of if there are just too few cells to assess in the tissue. Third, in vitro experiments assessing the filamentation of the mutants in the medium in which these assays are preformed need to be done as controls. Candida albicans filaments in many conditions such as tissue culture medium. Spider medium is a strong inducer of filamentation but is very different than in vivo/ ex vivo conditions.

Related to the collagen invasion assay, there is a misunderstanding. The reviewer appears to confuse the put mutations with proline auxotrophy. The put mutants are proline prototrophs and can synthesize proline as they possess a full repertoire of biosynthetic enzymes. In contrast, the put mutants cannot utilize proline to obtain nitrogen or energy. In fact, the presence of excess proline imposes toxicity to the put mutants. There are three possible sources of proline. 1) PureCol EZ Gel is a ready-to-use collagen solution that forms a firm gel when warmed to 37 °C. It contains purified Type I bovine collagen (5 mg/ml) dissolved in DMEM/F-12 medium, which has multiple amino acids, including a substantial amount of arginine. 2) The recovery medium DMEM supplemented with 10% FBS. The presence of FBS provides amino acids and induces filamentous growth. As the reviewer points out, C. albicans grows in this media and exhibits filamentous growth. 3) The proteolytic breakdown of collagen is expected to liberate proline. Consequently, the poor growth of the mutants clearly demonstrate the importance of proline catabolism. Also, the fact that we recovered put mutants surviving on top of the collagen (Fig. 4B, inset) suggests that they remain viable but simply are unable to efficiently invade the collagen. Consistently, microscopic inspection of the wells of the put mutants showed extremely few or even complete absence of invading cells in the recovery medium. We will adjust the text and provide a more detailed description of the experimental set-up. In summary, the main concern of the reviewer with respect to lack of proline is not relevant.

Regarding the 3D-skin model, equal numbers of fungal cells were applied on top of the RHE. To avoid overgrowth, only low numbers (100 C. albicans cells) can be applied for the WT strain, and consequently for all other strains. In contrast to WT, which clearly proliferates, the apparent low level of put1 and put2 cells at the center of the 3D skin model is the consequence of poor growth. The upper layer of the RHE consists of stratified keratinocytes. To grow, WT fungal cells obtain proline either directly from the keratinocyte, from secreted proteases that liberate proline from keratin (proline not as abundant in keratin as in collagen, the main component of the dermis), or from the medium that basolaterally feeds the RHE. At the border of the model leakage from the medium can occur. Our results, showing poor growth of the mutants in the center of the 3D-skin model, entirely consistent with the collagen plug experiments, indicates that proline catabolism plays a determinant role to enable invasive growth.

Lastly, the imaging experiments are highly problematic. First, reference must be made to previous ex vivo imaging reported by the Lionakis lab in 2013. Second, the number of cells imaged is so low that there is no power to make any conclusions. At 24 hr, the mutants may be delayed in filamentation or they may be delayed in establishing infection. There is no way to know what is causing the apparent lack of filaments. This technique as presented is not any higher resolution than traditional histology and in fact histology would provide a more convincing case for reduced filamentation.

These considerations significantly reduce the overall significance of the work.

I work on Candida albicans.

We thank the reviewer for highlighting the beautiful study by Lionakis et al which document the host response, specifically the role of macrophages in mitigating C. albicans infection of the kidney. However, the reviewer apparently failed to recognize that their method is completely differed from ours. Lionakis et al. performed ex vivo imaging of kidney slices using regular confocal imaging, and the authors express an awareness regarding the limitations of this approach. In fact, these authors even state in their discussion that intravital microscopy should be pursued in the future to further investigate Candida-macrophage interactions in the kidney. Also, they point out that kidney-specific factors seem to facilitate rapid filamentous growth of C. albicans. In our work, we have experimentally addressed both of these astute statements. To our knowledge, our work is the first report of imaging a Candida cell infecting a kidney in a living mouse, which on its own is a major development and achievement considering the complexity of the kidney microenvironment. The finding that the put2 mutant does not exhibit filamentous growth in the kidney of a living mouse (24 h) is striking and strongly suggests that a substantial quantity of proline, or amino acids (e.g., arginine) that are metabolized via the proline catabolic network, is present in the kidney. This is clear based on finding that WT C. albicans cells respond accordingly to initiate hyphal growth. Consistent to this, it is well documented that the kidney is a major metabolic hub for arginine and proline metabolism. The work by Lionakis aligns remarkably well with our previous and current work in that put mutants exhibit greatly reduced survivability in co-culture with macrophages and do not evade these primary immune cells due to their inability to induce filamentous growth within the phagosome (Silao et al., 2019). We have adjusted the text to include a discussion that places our work in the context of the Lionakis work.

We have added a Fig. 6C showing an example of the scanned area of the kidney. Further we added the following in the revised legend to indicate that large areas of kidneys were imaged in our assessment of fungal growth and filamentation:

“Sites of colonization where localized using a spiral scan in the Las-X Navigator-module in the FITC channel. The entire area of the renal surface attached to the glass imaging window was scanned; circles highlight examples of regions of interest (ROI) exhibiting stronger and deviating fluorescence from the background. Each ROI was examined in detail using FITC, yEmRFP and autofluorescence. Scale bar, 500 µm.”

CONCLUDING STATEMENT – SUMMARY RESPONSE:

Our current work is based our previous discovery that proline metabolism provides energy to induce and support filamentous growth (PLoS Genetics, 2019). This turned out to be important since we also discovered that C. albicans cells depend on mitochondrial proline metabolism to evade engulfing macrophages, implicating this process as being an important virulence determinant. Consistently, using time-lapse microscopy, we subsequently found that proline catabolic enzymes are rapidly induced in C. albicans cells upon phagocytosis by macrophages. These results demonstrated that proline is present within phagosomes. As exciting as these findings are, they focused on a single phenotype, i.e., filamentation, and were obtained using in vitro experimental approaches. These results demanded that we pursue additional avenues to further characterize and test the in vivo relevance and merely provide a solid background for the current work.

In contrast to reviewer 2 and 3, we do not believe that our finding that proline catabolism plays such a critical role in virulence as being merely “incremental”. We also could not have foreseen that the ability to use proline as an energy source is a common feature of multiple fungal pathogens capable of causing human disease. This is conceptionally very important in that human fungal pathogens, unlike the well-studied yeast Saccharomyces cerevisiae, are not readily found out in nature, and thus have evolved to use a similar spectrum of nutrients as host cells, including cancer cells. It is important for the fungal pathogen community to realize that regulatory switches operating in C. albicans are wired substantially differently to those in S. cerevisiae, and are likely optimized to reflect the actual condition in the host environment. The growing appreciation that diverse cancers are able to shift metabolism to exploit proline as an energy source is strikingly and fascinatingly similar to our findings with pathogenic fungi. This represents a conceptual advance in that it points to the wealth of proline stored within extracellular matrix proteins as providing a potential and significant source of energy for virulent fungal and cancerous growth.

Finally, we strongly believe it is improper to extrapolate virulence properties based on in vitro findings, and that it is essential to actually test host-microbial pathogen interactions using refined in vivo models. Our successful use of advanced intravital microscopy goes beyond traditional and accepted murine infection models and has provided us with a unique state-of-the-art vantage point. Our findings that a single C. albicans cell is able to initiate and establish a site of infection in a kidney within a living mouse is itself important, and coupled to the novel finding that hyphal development at sites of infection depends on the ability of the fungal cells to catabolize proline must reflect the physiological conditions in the kidney. This is not an incremental finding, and we do not understand that reviewers 2 and 3 diminish the significance of these findings. Clearly, our manuscript provides a strong foundation for more detailed and advanced studies.

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 of Silao et al. describes an in-depth investigation of the role of Put1 and Put2 enzymes in proline catabolism and virulence in Candida albicans. This is an extension of previous work in this system. The basic biochemistry and genetics are solid and support the role of these enzymes in the proposed pathway and provide evidence that the build up a toxic intermediate in the absence of Put2 is likely involved in the poor growth of the strain when proline is the only carbon source.

The conclusions regarding its role in virulence are less convincing, particularly the data derived from the collagen invasion assay, the ex vivo skin model and the ex vivo/in vivo imaging. The survival and fungal burden assays support a modest role in virulence and a modest reduction in infectivity (although the presented data for survival does not have statistical significance data reported for the kaplan analysis.

The manuscript is clearly written. The methods are well described.

Referees cross-commenting

I remain unconvinced of the broad significance of the advances and stand by my assessment that this is for the most part a reasonable study but does not move the field forward. The novel technical aspects are either extensions of previous in vivo imaging or are not well controlled (collagen invasion assay)s.

Significance

This is a detailed study of an area that is fairly mature and thus will be of interest to those in the field but does not represent a large advance and is thus truly incremental.

Major limitations of the work are as follows. First, the collagen invasion assay may be flawed. The recovery media is made with DMEM which is a medium that lacks proline and is fairly stringent. Control experiments need to be done to be sure that the mutants grow in the recovery medium. Second, the data from the RHE model are hard to interpret since so few cells are present in the tissue. It is hard to see if there are few filaments of if there are just too few cells to assess in the tissue. Third, in vitro experiments assessing the filamentation of the mutants in the medium in which these assays are preformed need to be done as controls. Candida albicans filaments in many conditions such as tissue culture medium. Spider medium is a strong inducer of filamentation but is very different than in vivo/ ex vivo conditions.

Lastly, the imaging experiments are highly problematic. First, reference must be made to previous ex vivo imaging reported by the Lionakis lab in 2013. Second, the number of cells imaged is so low that there is no power to make any conclusions. At 24 hr, the mutants may be delayed in filamentation or they may be delayed in establishing infection. There is no way to know what is causing the apparent lack of filaments. This technique as presented is not any higher resolution than traditional histology and in fact histology would provide a more convincing case for reduced filamentation.

These considerations significantly reduce the overall significance of the work.

I work on Candida albicans.

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 is part of the continuous work by the authors to dissect the mechanism of utilization of proline as a carbon source in Candida spp. In particular, this work shows that the inability to process proline leads to accumulation of the toxic intermediate P5C and subsequent inhibition of mitochondrial respiration and toxic effect on the cells. Furthermore, the study demonstrates that proline utilization is important for C. albicans kidney colonization. The experiments are meticulously designed and the study adds to the overall understanding of the metabolic utilization of proline as a carbon source and its potential relevance for infection. I find this work interesting, but the role of Put1 and Put2 in proline utilization is not particularly novel. The novelty here is the subcellular localization of the two proteins. Also, the importance of proline utilization for infection is unclear. The host-pathogen interaction assays are ambiguous as each assay gives different result. Lastly, the authors try to generalize the importance of use of proline as a energy source by other Candida spp.. This is not very surprising, given that it has been reported previously by others (example DOI: 10.1002/yea.1845) and that many pathogenic or closely related to C. albicans species use various amino acids, not only proline, as a carbon source. Moreover, the authors take C. albicans as an example to demonstrate the role of PUT in invasion and infection. Proline is known stimulus for hyphal growth in this species, but many other Candida spp., including C. auris, do not filament. So how, aside from supporting growth, proline is linked to infection in these species? I think the authors oversell the importance of proline in Candida spp. pathogenesis and should tone this part down or remove completely. A new story that validates the importance of PUT in non-albicans species can bring clarity to why and where proline is critical for survival and infection.

Major comments: I am not convinced by the data that proline is important to initiate infection. Candida infections of the kidney occur only at late stages of sepsis. The authors need more compelling data to prove that proline is important for infection in the host.

Minor comments: I find the manuscript difficult to read and the discussion part is overly long. Some streamlining and adding a bit more explanation for the rationale of each experiment will make the work easier to follow. Some language/style needs refining as well.

Referees cross-commenting

In this manuscripts the authors clarify the cellular compartmentalization of steps in proline catabolism. However, it is not novel that proline is a valuable carbon source. The role of proline utilization for establishing or progression of infection remains ambiguous even after the authors provide different in vivo results. The overall significance of the study is limited.

Significance

The strengths of this study are in the experimental design and variety. The data is well presented and visualized. The limitations are as pointed above - I find it especially difficult to figure out where, in a real infection scenario (e.g. breach of the gut barrier and entry into the bloodstream) proline will be the primary energy source.

To me the significance of this work is minor.

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:

Silao et al make the intriguing observation that yeasts that are generally considered less pathogenic are unable to catabolize proline than Candida albicans. They then, in Candida albicans, construct mutants defective for the two key enzymes (Put1, Put2) required to convert proline to glutamate, which they show to be essential for proline utilization as an energy (carbon) and nitrogen source. The authors proceed to untangle the regulatory aspects of proline degradation, including the respective cellular localization of its key enzymes. They then make the important discovery that strains lacking either Put1 or Put2 suffer from a proline-dependent growth defect, which they attribute to resulting defects in mitochondrial metabolism.

The manuscript then goes on to analyze a broad range of infection models including: reconstituted human epithelial skin model, Drosophila, mouse systemic infections, organ colonization in these mice (kidney, spleen, brain, liver and histochemistry of the kidneys) as well as survival when incubated with cultured human neutrophils. Finally, they use yeast cells constitutively expressing yEmRFP (so that yeasts can be distinguished from other host cells) and coated with FITC before incubation with the host cells (which coats the wall of the original cells, but does not spread to progeny) and they go on to perform an impressive set of analyses of C. albicans growth within mouse kidneys both in vivo and ex vivo, exploiting an implanted window together with intravital imaging with a two photon microscope at different time points. The system is impressive and visualizes tissue invasion by hyphal cells beautifully. Finally, they compare the intra vital images from WT and put2-/- cells and show that, as in vitro, put2-/- cells do not form filaments and do not show extensive invasion of the kidney tissue. While the in vivo aspect of the study includes many different models, it finds defects in virulence for different subsets of put mutants and the relative importance of filamentation vs proline utilization for virulence is not conclusively resolved.

Overall, this is an important and timely manuscript, which significantly contributes to the understanding of how proline metabolism intersects with yeast fitness in the context of infections. However, there are several major concerns regarding some of the conclusions drawn from the study. In addition, some general recommendations that would improve the manuscript are provided.

Specifically, the manuscript provides a very detailed description of experiments and observations. However, in several parts it is difficult to follow and the the reader needs more guidance about the logic involved in reaching conclusion. Specifically, several aspects of the paper are written for experts in Candida (yeast) metabolism. Here, explaining the rationale for some of the experiments, and providing more background information that is not obvious to a non-expert, is required.

In particular, writing a clear and measured summary sentence at the end of each paragraph and a conclusion paragraph that summarizes key findings in simple terms would help make the manuscript more digestible for readers.

In addition, the impressive microscopy and broad range of in vivo experiments is comprehensive but only adds incremental information relevant to proline metabolism-that filamentous growth in vivo and virulence is reduced in cells carrying some mutations in one or more put genes. However, this broad sweep of model systems and the development of the in vivo imagining system might have more impact in a separate paper focused on the real-time in vivo visualization of kidney invasion.

Major comments:

1. The main finding that impressed this reviewer is that "removing the ability to catabolize proline, in an organism that evolved to catabolize it, leads to (growth) defects". This point could be better highlighted throughout the manuscript.
2. The authors show that deletion strains for proline metabolism have defects that are important for in vivo pathogenicity. This is an important finding. However, as the manuscript reads now, it suggests that the main findings are that the ability to use proline in the respective host niche is key. Mechanistically, the manuscript revolves primarily around defects that arise when deleting PUT1 and/or PUT2 (i.e., an "unknown" toxicity of proline in the case of put1-/- (or put1-/- put2-/-) and the additional P5C-dependent toxicity for put2-/- mutants; see below).
3. In order to claim that catabolizing prolines promotes pathogenicity (as opposed to the alternative hypothesis that the inability to catabolize proline leads to the observed defects), additional experiments would be required. For example, the put mutants would need to be compared with mutants that significantly reduce/impair proline uptake, such as the referenced gnp2 mutant (Garbe et al 2022). While the finding that less pathogenic yeast species are unable to catabolize proline is both intriguing and important, it also remains as is presented as a loose, non-quantitative correlation that only tangentially address the question of whether "proline catabolism is key for pathogenicity".
4. 238 onwards: The conclusion that "the primary growth inhibitory effect of proline is linked to catabolic intermediates formed by Put1 and that are metabolized further by Put2"does not appear to be fully supported by the evidence. Addition of proline to put1 mutants already reduced OD600 by ~50% (Figure 2); and is further reduced to ~10% when put2 is deleted. This implies that there are two inhibitory effects of proline, not one primary one. At the least, this option should be discussed, including why deletion of PUT1 leads to proline toxicity. The latter is not clear-is it that too much proline accumulates in the cell and this accumulation is toxic? If this is the case, the effect would be expected to be proline concentration dependent. Performing a relatively simple experiment as performed for the put2 mutant (Fig. 3 / S3F) may clarify this issue. Particularly, if the experiment would be coupled with intracellular quantification of proline.
5. The caption "P5C mediates a respiratory block" is misleading, as the evidence is not that compelling: Although P5C increases in put2, but not in put1 mutants, and given that both single mutants experience a proline-dependent respiratory defect (Fig. 3E), the results suggest a more complex relationship.
6. The virulence assays and in vivo experiments do not present a unifying view: in Drosophila put2∆∆ is less virulent than put1∆∆, which appears similar to put3∆∆. Given that put2 mutants grow slowly, likely because of P5C inhibition, this seems logical. However, in mice, put3∆∆ remains highly virulent while put1∆∆ and put2∆∆ results for survival are mixed. Furthermore, in 4 mouse organs, put1∆∆ and put2∆∆ are not significantly different from one another but are different from wt, while put3∆∆ has no significant reduction in CFU. Kidney histology shows very little invasion by put1 and put2 and more by put3, but visually put3 appears to invade much less than the WT, and the human neutrophil experiment shows effects of put2 or put3 but not put1. This leaves the reader rather confused. It may be worth discussing the reasons for different results in different models. Is the availability of proline in each of the organisms and organs similar?
7. The ex vivo and in vivo analysis of the dynamics of C. albicans growth in the host is visually impressive, but it distracts from the focus of the paper and the metabolic findings. Showing that put mutant cells do not form filaments in vivo (as in vitro) does not add much conceptually to the paper. Furthermore, this lovely advance in in vivo visualization is lost at the end of this paper and the authors should consider whether it might fit better in manuscript that could really highlight the in vivo visualization approach.
8. The discussion of cells stained with FITC and expressing yEmRFP does not clearly point out that the FITC is only an indicator for those cells that were used to innoculate the tissue and that finding cells without FITC indicates that they are mitotic progeny, indicating that they have been dividing. The authors clearly understand this, but a naive reader may miss this important point if it is not stated explicitly.

Minor comments:

1. Throughout: what is the distinction between utilization of proline for C or for energy? These terms seem to be used interchangeably.
2. Introducing the schematic in Fig. 2A at the beginning of Figure 1, would help explain proline catabolism before delving into the growth experiments that rely upon this framework. This should include an explanation, for readers less familiar with the metabolic issues, of the main limitations to catabolizing proline, and the key issues for being able to use proline for nitrogen, carbon, and energy (potentially indicated in the overview figure, e.g. pointing towards gluconeogenesis etc.).
3. Saccharomyces can only grow on proline as a nitrogen source, but not as energy/carbon source. Could the authors briefly mention or discuss why this is the case? This is not clearly apparent after reading the manuscript and it leaves the reader confused and trying to understand if the fact that proline is required for carbon utilization is a new finding of this paper or was already known. Do the authors think this is tied to the presence of complex 1 components in C. albicans that are not found in S. cerevisiae. Is this consistent for the pathogenic, but not the non-pathogenic yeasts analyzed in figure 1?
4. 100: While Gdh2 is apparently an important enzyme for generating ammonium, why is it not necessary for macrophage escape and virulence as shown in reference 18? A recent paper from Garbe et al (ref 12) suggests that Gnp2 is the major proline permease in C. albicans and what is known, and not known, about proline uptake would be good to mention, given that PUT gene functions require that proline enters the cells.
5. 116: Is the "low sugar environment of the host" referring to a specific niche, such as the GI tract, or human blood? Compared to most natural environments, glucose is abundant in the host, e.g., at ~5 mM, it is the most abundant metabolite in blood, and similarly, in the GI tract, levels can go beyond 50 mM glucose (see e.g. PMIDs 34371983, 21359215). Or is this comment indicating that the in vivo sugar concentration is lower than that in common lab growth media? Please spell out the niche/concentration for clarification - and compare that to other niches that are considered "high sugar environments".
6. 123: "proline as sole energy source" - suggest "is the source of carbon, nitrogen, and energy"
7. 142: it is worth noting to readers that C. neoformans is a basidiomycete and thus VERY distant from the other yeasts studied here-it is in a different major phylum of fungi.
8. 143: Here it is implied that put1 and put2 mutant strains do not grow on SPD, but this is not stated explicitly.
9. 151: The abbreviation SPG is not explained in main text.
10. Paragraph 156 onwards: this section is particularly hard to read and very dense. Also, it is difficult to understand the significance of these experiments for the overall findings of the paper. Please at least provide a small conclusion / summary at the end of the paragraph that puts the findings into perspective.
11. Figure 2 C: simplifying the scheme (e.g. lots of redundant information, P2 and Mito - just give it one name) would help. This figure may be better in the supplementary material.
12. Figure 2B: It is not directly apparent from the micrographs that Put1-RFP localisation is mitochondrial. Co-localisation of the RFP with a mitochondrial dye (e.g., mitotracker) or something similar is required to validate it.
13. Throughout the manuscript (figure legends): Suggest using "mean" instead of "Ave."
14. 175: According to the 'Yeasttract' and 'Pathoyeasttract' databases, Put1 regulates at least 36 and 22 genes, in S. cerev. and C. alb., respectively (based on DNA binding and/or regulatory changes). The only gene in common between these two lists of genes is PUT1. Thus, it is quite likely that Put3 regulates many other processes that explain its function and that its major function may not be only to regulate Put1.
15. 175: Is it clear whether the Put3-independent mechanisms are positive or negative with respect to Put1?
16. 218: Suggestion: "growth was indistinguishable".Unless growth curves or growth rates are provided and if one time-point data are the basis for this point, than "rates" is not a relevant term.
17. 256 onwards: did the authors test if the ROS scavenging effectively reduced ROS? i.e. does the luminol-HRP assay yield less ROS in +proline +scavenger treatment? This is necessary to effectively conclude that the growth inhibitory effect of proline is due to blocking respiration.
18. The Figure captions are extremely lengthy and detailed, making it cumbersome to find the relevant information. Suggest moving some of the information, such as additional experimental details, into the methods section.
19. 277-301: Phloxine is not exclusively a live/dead cell indicator-it is an indicator of metabolic activity. In Scerev. and Calb. it also indicates slower growth, opaque growth, and it has been used as an indicator of aneuploidy in C. glabrata (https://journals.asm.org/doi/10.1128/msphere.00260-22) and of diploids vs haploids in S. pombe. The colonies illustrated aer made up of many live cells, and thus the section "Defective proline utilization is linked to cell death" needs to be presented more carefully. In addition, it appears that this section shifts from using defined medium to using rich medium and 37C instead of 30C. Why was this shift necessary?
20. 295-301: Related to the point above, these results are hard to interpret due to the switch from defined medium in all prior experiments to rich growth medium here. Also, it is not clear why a 48h old YPD culture was chosen to show that the degree of PI staining correlates with mitochondrial activity - is this due to the culture age? It would be more clear to image cells grown on glucose vs. glycerol/lactate, or under repressive / de-repressive glucose concentrations (e.g., as shown in Fig. S4C where a PI+ difference is apparent for 0.2% glucose vs. 2% glucose at 30{degree sign}C).
21. 313-14: The statement 'the invasion process was dependent on the ability of cells to catabolize proline' doesn't take into account that put mutant cells are defective in filamentous growth irrespective of their utilization of proline...and like the efg1 cph1 double mutant.
22. 316-327: The results of the experiment described can only be interpreted as an effect of proline catabolism if the three strains (efg1 cph1; put1; put2) have similar growth rates as yeast cells in vitro. Why weren't the cells competed directly (efg1 cph1 vs put cells)?
23. Fig 6: The logical order of the experiments, and in the text, is: 1) 4 h window, 2) 26 h window and then 3) ex vivo. The cartoon in 6B should be in this order as well.
24. 337: it is not clear what the 'direct exposure...' is trying to tell us. Can this be made more explicit?
25. 340-346: Here proteins with high proline content were used to ask if they could be induce transcription of PUT1 or PUT2 RNA and protein. This experiment is designed only to test the role of these proteins to induce utilization of nitrogen, as glucose is included in the medium. Given that these proline-rich proteins need to be lysed by proteases before they can be imported, and since no import pathways were tested, the results appear to tell us that mucin is more readily digested to peptides that contain proline-but why that is the case is not clear and how it relates to proline utilization is also not clear.
26. 363-369 An alternative is that Put3 induces different proteins important for growth.
27. 379-380-the conclusion for this paragraph is somewhat of an overstatement as there is no analysis of the degree to which proline utilization is a predictor of virulence. It simply shows that put mutants affect the ability to survive in neutrophils.
28. Discussion: The statement that "S. cerevisiae" evolved in high sugar environments is debatable. The natural niche could well be forest soil and tree bark, or insect/wasp guts with arguably little glucose around.
29. 469-470-how strong is the 'correlation' between the ability to utilize proline and virulence? Given that different mutants had different effects in different models, this seems like a very loose 'correlation'; it would be good to have some quantitative measures to make this claim.
30. 500: Was the experiment was done in larvae, and not in adult Drosophila? Fig 5 legend says flies and shows a picture of a fly and larvae are only mentioned much later in the text..
31. 512:Why is it presumed that proline accumulates in the mitochondria in put1 mutants? How strong is the presumption?
32. 539: why are MMPs important for digestion of collagen? This is not clear at this point of the Discussion.
33. 574: Concluding sentence of this paragraph seems unsubstantiated. There are at least two defects in put2 strains-hyphal growth and growth in general, presumably because of P5C accumulation.
34. Fewer abbreviations would make the manuscript easier for non-experts to read. For example, P5C is not defined in the abstract. Furthermore, if an abbreviation is not used more than 3 times, it is not necessary to provide it (e.g., mammalian proteins in the last paragraph).

Typos: 1. 82: should read 'is restricted to the mitoch...' 2. 102-103: should read 'to evade macrophages' 3. Fig. S4F is mislabelled as Fig. S4G.

Referees cross-commenting

Overall, we stand by our initial assessment of the study. However, we were not aware of previous studies that investigated proline utilization in yeasts, as noted by Rev # 2 (https://onlinelibrary.wiley.com/doi/epdf/10.1002/yea.1845). The current study suggests that using proline as an energy/carbon source is more wide-spread, beyond pathogenic yeasts. Further, the C. albicans strain they used for this study (ATCC 10231) was apparently unable to grow on proline in the quoted paper. In light of this, we think the authors should reference this study, tone down the claims about the clear correlation of pathogenicity and proline utilization, and address this apparent discrepancy with the indicated Candida albicans isolate. We note that our review considered this a paper mostly of interest to specialists.

Significance

1. The advance in this paper is conceptual for the proline utilization connection to virulence in a range of species and technical for the in vivo microscopy. Limitations are that the conceptual advance is based only on qualitative work in figure 1 and that the animal studies do not provide a conceptual advance, although the technical advance of in vivo visualization of kidney tissue is impressive and (to the knowledge of this reviewer) quite new as the only prior work was in mouse ears.
2. The work fits well as an extension of the body of work from the corresponding author's lab with additions from the labs with expertise in models of infection.
3. People interested in yeast metabolism and pathogenic yeast virulence will be the audience for this paper and as written it is for a specialized audience interested in pathogenic yeast metabolism and, perhaps, (although not mentioned at all in the text) for those who want to try PUT gene products as new drug targets.
4. Reviewer expertise is in pathogenic yeast biology and yeast metabolism. Little expertise in high tech microscopy.

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

Comment 1.

The impact of this study would be greatly enhanced if the authors could provide electrophysiological results validating that intra-VTA infusion of Grin2c siRNA affects the excitability or discharge regularity of different populations of VTA neurons (at least dopaminergic and GABA neurons).

*Reply. The reviewer is suggesting an additional experiment that constitutes a next logical step to further determining the role of NMDA receptors (NMDARs) containing the GluN2C subunit(s) in induction of EPSP in VTA TH+ and TH- neurons; such an experiment could be performed in vivo in behaviorally tested animals, but the recordings would have to be done in anesthetized animals 24h after VTA siRNAs microinjections; an alternative would be an in vitro experiment with slices that would maintain a functional link between dorsal raphe (DR) glutamatergic reward neurons and VTA neurons; we agree that this would constitute an important step forward, and that can be performed once our findings are published. *

Comment 2.

The strategy of targeting Grin2c transcripts with siRNAs is appropriate and the authors implemented a western blot validation to control the effectiveness of their approach. However, due to the large volume of siRNA injected into the VTA (500 nl), the authors should consider extending their western blot validation experiments to structures anatomically close to the VTA and rich in GluN2C subunit such as RMTg. The DRN would also be an interesting structure to show that GluN2C subunit expression is not affected by the siRNA approach.

*Reply. Optogenetic studies have shown that the reward signal initiated by glutamatergic neurons in the dorsal raphe is transmitted to VTA neurons (Liu et al., 2014; McDevitt et al., 2014 Qi et al., 2014;), findings that are consistent with, and that were predicited by, our previous findings (Rompré and Miliaressis, 1985; Boye and Rompré, 2000; Ducrot et al 2013). Moreover, a large body of research carried out with local (VTA) drug injections generated data confirming the role of VTA neurons in reward (See for instance Wise and McDevittt, 2018). Thus, it is reasonable to hypothesizes that it is the reduction of GluN2c (NMDARs that contain this subunit) in VTA neurons that is responsible of the attenuation of reward. Such a hypothesis is further reinforced by our findings that GluN2c and its gene are expressed in VTA neurons. Why should it be within the RMTg? The reviewer likely knows many different studies that have shown that RMTg neurons play a key role in aversion; they provide a strong tonic inhibitory input to reward-relevant VTA neurons and more specifically to VTA DA neurons (see Zhou et al). RMTg neurons receives a strong excitatory input from the lateral habenula; activation of these neurons strongly inhibit reward. Consequently, a logical hypothesis is that a reduction of the glutamatergic excitation of RMTg caused by a reduction of NMDARs that contain GluN2Csubunits would have produce an enhancement of the reward signal, not an attenuation. *

*The DR, site at which the electrical stimulation was delivered is located near 2 mm behind and 3 mm above the injection sites, it is very unlikely that the small volume injected hit the DR. *

Comment 3.

The authors should consider discussing or re-evaluating their findings from their 2015 study which suggested that the effect on reward induced by DRN stimulation was controlled by GluN2A-containing NMDARs most likely located on afferent terminals.

Reply. We cannot understand why we should re-evaluate or re-discuss these findings. It has been repeatedly shown that blockade of GluN2A containing NMDARs by local VTA PPPA and R-CPP microinjections enhances the reward signal initiated by DR electrical stimulation (Bergeron and Rompré, 2013; Ducrot et al., 2013; Hernandez et al., 2015; and the present study (control group); this enhancement effect was not attenuated by a decrease in NMDARs that contain GluN2A subunit(s) (Hernandez et al 2015). Our discussion is reasonable in view of these findings, and the hypothesis that these GluN2A containing NMDARs are located on afferent terminal explains the results.

Comment 4.

In Figure 2, the authors should quantify the results of the colocalization levels of Grin2c and TH in dopaminergic neurons of the substantia nigra pars compacta.

Reply. As mentioned in our reply to comment 2, the quantification of VTA neurons was highly justified by a large body of the literature which is not the case for the substantia nigra pars compacta neurons.

Comment 5.

The results should be presented differently in figure 5 in order to be able to compare on the same graph and with the appropriate statistical analyses (two-way ANOVA), the impact of PPPA infusion or solvent in the SCRGluN2C or siRNA groups.

Reply. Unfortunately, this is not possible because not all subjects were tested with PPPA.

Comment 6. The authors should clarify the N=12 per condition in Figure 3, especially since 11 values for the control conditions are plotted on the histogram.

*Reply. The reviewer is correct there are 11 Subjects in the control group and 12 in the active sirna group. We made the changes to the methods section. *

Comment 7.

Authors should standardize the way they cite literature throughout the manuscript (number or authors).

Reply. Thanks for the suggestion changes have been made to standardize the citations.

Comment 8.

The authors should clarify sentences 102-105 of their introduction, which seem to conflict but ultimately describe similar results.

Reply. Reviewer is correct the following sentences at the end the paragraph were deleted:

*This hypothesis predicts that activation of a given subtype(s) potentiates DA burst firing and DA release, whereas activation of different subtype(s) increases the inhibitory drive to DA neurons. This idea is supported by data mentioned above and others showing that both activation and blockade of VTA NMDARs increase DA burst firing (French et al., 1993), accumbens DA release (Karreman et al., 1996; Westerink et al., 1996; Mathé et al., 1998; Kretschmer, 1999), and stimulate forward locomotion (Kretschmer, 1999; Cornish et al., 2001). Rodents also readily learn to directly self-administer the non-selective NMDAR antagonists, AP-7, into the VTA (David et al., 1998), showing that NMDAR blockade can have positive rewarding properties on its own. *

Comment 9.

The expression of different GluN2 subunits across different regions of the brain has been known since the early 90's as the authors acknowledged. In the abstract, the authors state that GluN2C is "the most abundant subunit of the NMDA receptor expressed in the VTA" (line 67). This idea that GluN2C is "the most abundantly expressed in DR and VTA compared to other Grin2 subunit transcripts" (line 425), is repeated throughout the paper. However, in the Results section they state that GluN2C is present at the same level as GluN2B; something that is also clearly visible in figure 1, where is also clear that GluN2A is also present almost at the same level. The emphasis that GluN2C has a larger representation over 2A and 2B in VTA is not necessary and misleading.

*Reply. The point here was to bring attention to the reader to the expression of GLuN2C, the main target of the current study. As shown in Figure 1, GluN2c is indeed the most abundant in the VTA. *

Comment 10

3) Performing an immunoblot in tissue obtained with a tissue punch of the VTA, the authors confirmed that the GluN2C mRNA detected is translated into protein. Unfortunately, this important data is not showed, and it should be shown. Moreover, immunocytochemistry of GluN2C could help to identify the cellular type where the protein is expressed, something that could be key to better understand the role of NMDARs in the reward pathway. Are 2A/2B expressed in different cells that 2C? What type of cells express 2C? These are just a few of the question that a better and more detailed analysis of 2C expression could provide. Without this, the interpretation of results presented here, as well as previous results, regarding the role of NMDARs continuous being confusing.

Reply. Because we measured GluN2C proteins within the VTA, we infer that some Grin2C detected in VTA TH+ and TH- neurons is translated into proteins, and reduction of the protein expression resulted into a selective attenuation of reward We added a supplementary fig showing the GluN2c protein in different brain regions.

Comment 11.

4) The largest number of 2C positive cells do not express TH complicating the interpretation that 2C is necessary to convey reward information in the DR-VTA circuit. Other effects due to downregulation of 2C could be responsible of the behavior changes observed. Although the authors offer an explanation for this, is not enough. They suggest that 2C maybe involved in a reduction of excitatory inputs into inhibitory interneurons that when downregulated should produce an opposite effect to what is observed. However, without knowing the identity of those GluN2C expressing cells this comment is only speculation and does not rule out a role for other GluN2C expressing cells that are not TH positive.

*Reply. We do consider the hypothesis that the attenuation of reward is due to a reduction of GluN2C in TH- neurons, in fact we discuss both hypothesis, TH+ and TH-. Characterization of the reward-relevant neuronal pathway has been an important aim since Olds and Milner discovery. Our findings constitute, as mentioned in reply to comment 1, and important step forward, and indeed identification of the specific VTA cells that convey the reward signal is another important question that should be addressed. Our findings provide a strong ground to focus on GluN2C but not the other subunits. *

Comment 12.

5) In this line, there is no good explanation why treatment of animals with GluN2A blocker enhances the reward pathway only in animals treated with control siRNA. Two possibilities could explain this. 1) there is some sort of relationship between 2C and 2A that when 2C is absent, PPPA has no effect. Again, it could be important to know if 2C and 2A are expressed in the same cellular type; 2) the control siRNAs are not completely innocuous and may produce unknown effects that alter the functionality of VTA.

*Reply. We believe that our data are strong and valid because the methods we used have been validated and the results with PPPA in control group are similar to those previously published. Previously we have shown that a reduction of VTA GluN2A proteins has no impact on reward per se nor on the enhancement of reward by PPPA (Hernandez et al., 2015). The hypothesis raised by the reviewer that 2C and 2A interact is incompatible with the findings that we obtained. Could it be a non-specific effects like tissue damage. In such a case however we would have observed a decrease in 2A subunits as well which is not the case. *

Comment 13- 16

6) Downregulating 2C suggests that this subunit is vital to relay a reward signal in VTA neurons. The following are comments regarding the analysis of the data of Fig 4 and 5. 7) It is not clear how the maximum and minimum are estimated in order to fit a sigmoidal curve to the data. Are they average of the stable part? Where the error bars on each data point come from? What is the actual value and standard deviation of M50 values?

Reply: As described in the self-stimulation training in the materials and method section

“The data relating to the rate-frequency was fitted to a sigmoid described by the following equation y=Min+((Max-Min) )/(1+[10]^((x50-x)*p) ) where Min is the lower asymptote, Max is the upper asymptote, x50 is the position parameter denoting the frequency at which the slope of the curve is maximal, and p determines the steepness of the sigmoid curve. The resulting fit was used to derive an index of reward defined as the pulse-frequency sustaining a half-maximal rate of responding (M50). Self-stimulation behavior was considered stable when the M50 values varied less than 0.1 log unit for three consecutive days”

The Max, Min and all the free parameters of the equation are the determine by the best-fit parameters by minimizing a chosen merit function. A merit function, also known as a figure-of-merit function, is a function that measures the agreement between data and the fitting model for a particular choice of the parameters. By convention, the merit function is small when the agreement is good. To optimize the merit function, it is necessary to select a set of initial parameter estimates and then iteratively refine the merit parameters until the merit function does not change significantly between iterations. The Levenberg-Marquardt algorithm has been used for nonlinear least squares calculations in the current implementation.

As described in the self-stimulation training and material section.

“.. Four stimulation sweeps were run daily, and the first sweep was considered a warm-up and discarded from the analysis”. The remaining 3 sweeps were fitted to a sigmoid and the parameters were obtained. The error bars correspond to the difference across each sweep”.

• What is the actual value of the M50 SD? Don’t understand why this is relevant? We already provide the SEM.*

8) This type of data is better analyzed by nonlinear regression analysis followed by ANOVA and some post hoc multiple comparison test.

Reply: We totally agree with the reviewer that is why the analysis was done fitting a sigmoid line. In fact, the fitting uses non-linear regression to compare the data points to the function, which in this case is a sigmoid function defined by the equation y=Min+((Max-Min) )/(1+[10]^((x50-x)*p) ) where Min is the lower asymptote, Max is the upper asymptote, x50 is the position parameter denoting the frequency at which the slope of the curve is maximal, and p determines the steepness of the sigmoid curve. In fact, intracranial self-stimulation data has been analysed using non-linear regression models since the seminal work by Coulombe and Miliaresis 1986 [1]

Indeed, after obtaining the results of the parameters, we follow it up with traditional statistics like t test and Anovas

9) Given the large variance of individual data points in Figure 4 and 5, a stricter statistical analysis than a t-test is necessary.

*Reply: Thanks for the suggestion, but we do not fully understand what the reviewer is suggesting. In general, the condition to apply or not a specific statistical test assumes about the underlying distribution the conditions required to conduct a t-test include: the measured values are in ratio scale or interval scale, simple random extraction, homogeneity of variance (i.e., the variability of the data in each group is similar), and normal distribution of data. The normality assumption means that the collected data follows a normal distribution, which is essential for parametric assumption. In all the data presented in figure 4-5 the assumptions are respected and checked (the data is measure in a continues scale, the group assignation was performed randomly, the data does not violate the normality assumption and the variance between the groups is similar. In the only case where the variance assumption was not held the Welsh correction was applied. *

Comment 17

10) Minor comments include the need to refer to figure panels in ascending order in the same sequence as they are described in the text.

Reply: Thanks for the suggestion we made the required changes. Now the figure panels are in the same sequence as they are described.

Comment 18

The role of NMDARs in VTA are explained in a rather confusing manner in the introduction. Lines 104 to 106 need some rewording since it conveys that blockade of NMDARs stimulates reward and that an opposite effect is observed following the blockade of NMDARs.

Reply. We simply report data from the literature. Each statement is supported by the relevant literature.

[1] Coulombe and Miliaressis, “Fitting Intracranial Self-Stimulation Data with Growth Models.”

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

Given the importance of glutamatergic synaptic transmission in the reward pathway from the dorsal raphe to VTA, the authors set to study the role of GluN2 subunits in a reward behavior. First, using qRT-PCR they quantify the amount of GluN2 subunits in different brain regions. They highlight the presence of GluN2C as the most abundant species of GluN2 subunits in VTA. They also identify that ~50% of TH positive neurons are also positive for GluN2C mRNA. Finally, they downregulate GluN2C in VTA using a commercially available siRNA. Knockdown of GluN2C reduces reward-seeking behavior as it increases the M50 and reduces the maximum response. The authors conclude that "VTA glutamate neurotransmission relays a reward signal initiated by DR stimulation by acting on GluN2C NMDA receptors".

Major comments

1. The expression of different GluN2 subunits across different regions of the brain has been known since the early 90's as the authors acknowledged. In the abstract, the authors state that GluN2C is "the most abundant subunit of the NMDA receptor expressed in the VTA" (line 67). This idea that GluN2C is "the most abundantly expressed in DR and VTA compared to other Grin2 subunit transcripts" (line 425), is repeated throughout the paper. However, in the Results section they state that GluN2C is present at the same level as GluN2B; something that is also clearly visible in figure 1, where is also clear that GluN2A is also present almost at the same level. The emphasis that GluN2C has a larger representation over 2A and 2B in VTA is not necessary and misleading.
2. Previous reports showed that pharmacological blockade on GluN2B or downregulation of GluN2A, the most common GluN2 subunits in the brain, do not affect the nose-poke behavior the authors use here. The biophysical properties of GluN2C are very different from those of 2B and 2A, therefore the fact that 2C downregulation does affect the behavior observed makes an interesting case for the subunit.
3. Performing an immunoblot in tissue obtained with a tissue punch of the VTA, the authors confirmed that the GluN2C mRNA detected is translated into protein. Unfortunately, this important data is not showed, and it should be shown. Moreover, immunocytochemistry of GluN2C could help to identify the cellular type where the protein is expressed, something that could be key to better understand the role of NMDARs in the reward pathway. Are 2A/2B expressed in different cells that 2C? What type of cells express 2C? These are just a few of the question that a better and more detailed analysis of 2C expression could provide. Without this, the interpretation of results presented here, as well as previous results, regarding the role of NMDARs continuous being confusing.
4. The largest number of 2C positive cells do not express TH complicating the interpretation that 2C is necessary to convey reward information in the DR-VTA circuit. Other effects due to downregulation of 2C could be responsible of the behavior changes observed. Although the authors offer an explanation for this, is not enough. They suggest that 2C maybe involved in a reduction of excitatory inputs into inhibitory interneurons that when downregulated should produce an opposite effect to what is observed. However, without knowing the identity of those GluN2C expressing cells this comment is only speculation and does not rule out a role for other GluN2C expressing cells that are not TH positive.
5. In this line, there is no good explanation why treatment of animals with GluN2A blocker enhances the reward pathway only in animals treated with control siRNA. Two possibilities could explain this. 1) there is some sort of relationship between 2C and 2A that when 2C is absent, PPPA has no effect. Again, it could be important to know if 2C and 2A are expressed in the same cellular type; 2) the control siRNAs are not completely innocuous and may produce unknown effects that alter the functionality of VTA.
6. Downregulating 2C suggests that this subunit is vital to relay a reward signal in VTA neurons. The following are comments regarding the analysis of the data of Fig 4 and 5.
7. It is not clear how the maximum and minimum are estimated in order to fit a sigmoidal curve to the data. Are they average of the stable part? Where the error bars on each data point come from? What is the actual value and standard deviation of M50 values?
8. This type of data is better analyzed by nonlinear regression analysis followed by ANOVA and some post hoc multiple comparison test.
9. Given the large variance of individual data points in Figure 4 and 5, a stricter statistical analysis than a t-test is necessary.
10. Minor comments include the need to refer to figure panels in ascending order in the same sequence as they are described in the text.
11. The role of NMDARs in VTA are explained in a rather confusing manner in the introduction. Lines 104 to 106 need some rewording since it conveys that blockade of NMDARs stimulates reward and that an opposite effect is observed following the blockade of NMDARs.

Overall, the data and analysis still leave too many open questions and the role of GluN2C, vs the other subunits, is not clearly established.

Significance

The study is limited in its scope and possible interpretations. The role of GluN2 subunits in the relay of reward information is only incrementally advanced and still continuous to be confusing.

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 by Hernandez et al provides a series of molecular, anatomical, and behavioral experiments exploring the distribution and the function of the NMDA Glun2C subunit in the ventral tegmental area (VTA). The authors demonstrate that reward signals originating from the dorsal raphe (DR) are carried to VTA neurons through the activation of GluN2C NMDA receptors. this study did not address the specific role of VTA dopaminergic neurons in mediating the reward signal. The major strengths of this paper are the quality of the FISH experiments and the well-established model of Brain Stimulation Reward targeting the Dorsal Raphe. This study is a follow-up of a previous study published by the same group (2015), which explored the expression of GluN2A/2D subunits in the VTA and their role in reward induced by dorsal raphe stimulation. This is an interesting and original manuscript. However, several issues, concerning the design of the experiment and the interpretation of the data, reduce my enthusiasm.

The impact of this study would be greatly enhanced if the authors could provide electrophysiological results validating that intra-VTA infusion of Grin2c siRNA affects the excitability or discharge regularity of different populations of VTA neurons (at least dopaminergic and GABA neurons).

The strategy of targeting Grin2c transcripts with siRNAs is appropriate and the authors implemented a western blot validation to control the effectiveness of their approach. However, due to the large volume of siRNA injected into the VTA (500 nl), the authors should consider extending their western blot validation experiments to structures anatomically close to the VTA and rich in GluN2C subunit such as RMTg. The DRN would also be an interesting structure to show that GluN2C subunit expression is not affected by the siRNA approach.

The authors should consider discussing or re-evaluating their findings from their 2015 study which suggested that the effect on reward induced by DRN stimulation was controlled by GluN2A-containing NMDARs most likely located on afferent terminals.

In Figure 2, the authors should quantify the results of the colocalization levels of Grin2c and TH in dopaminergic neurons of the substantia nigra pars compacta.

The results should be presented differently in figure 5 in order to be able to compare on the same graph and with the appropriate statistical analyses (two-way ANOVA), the impact of PPPA infusion or solvent in the SCRGluN2C or siRNA groups.

The authors should clarify the N=12 per condition in Figure 3, especially since 11 values for the control conditions are plotted on the histogram.

Authors should standardize the way they cite literature throughout the manuscript (number or authors).

The authors should clarify sentences 102-105 of their introduction, which seem to conflict but ultimately describe similar results.

Significance

This study is a follow-up of a previous study published by the same group (2015), which explored the expression of GluN2A/2D subunits in the VTA and their role in reward induced by dorsal raphe stimulation. This is an interesting and original manuscript.

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

We thank the reviewers for their insights and comments on this manuscript. Specific responses to reviewer concerns are detailed below. We made a couple of significant changes based on the feedback. First, we performed more experiments to increase biologic replicates and then quantified image data for multiple figures. The new quantitative information added to Figure 3 fully supports our original conclusions about changes to the ONH in Hes-TKO mutants. The quantification of Atoh7, Otx2, Rbpms and Crx expressing cells among the different genotypes revealed interesting differences in Notch intracellular gene requirements for both RGC and cone development. The most startling outcome is that changes in both cell types correlate with significant changes in Otx2, but not Atoh7. This singular finding suggests interesting future work is needed, well beyond the scope of this paper about the molecular mechanisms underlying these cell fates. Second, our data presentation was reorganized with new information added to Fig 1 that clarifies the relationships between Hes1, Hes5, Foxg1 and Pax2; old Figs 6 & 7 about neurogenesis were merged; and some data moved to new Suppl Figs 2 and 5. The numbering for multiple figures changed and a new summary model (now Fig 8) is provided. In addition, the manuscript was completely rewritten to improve clarity. We hope this revised manuscript is acceptable for publication.

Reviewer #1 Summary:

In this study, the authors employed an impressive set of mouse mutant or Cre lines to investigate the complexity of Notch signaling across different stages of retinal development. These comprehensive analyses led to two main findings: 1. Sustained hes1 in the OHS/OS is Notch-independent; 2. Rbpj and Hes1 exhibited opposing roles in cone photoreceptor development. Although the study is potentially interesting, the current manuscript needs the essential research background and quantification, a lack of which significantly reduced the clarity of the manuscript and the credibility of the major conclusions. Also, how the authors organized the results is quite confusing, making the manuscript very difficult to follow.

Response: We agree with all reviewers concerning incomplete quantification of the data. We directly addressed this shortcoming in revised Figs 3 and 6 (the latter combines old Figs 6 +7). To do this, we repeated some IHC experiments to add more replicates and reorganized all of the neurogenesis phenotypic data figures. Our quantifications uncovered several surprising outcomes that clarify our model. For these reasons, the manuscript was exhaustively rewritten. We merged E13 neurogenesis data into revised Figure 6 and moved the most relevant E16 analyses to new supplemental data Fig 5. All changes made should make the paper easier to understand for retinal development, neurogenesis, and Notch pathway aficionados, in addition to readers lacking such expertise.

Major comments: 1. The authors needed to make the quantification for many analyses to strengthen the conclusions, such as Fig. 1F, 1G, and etc.

Response: We quantified optic nerve head (ONL) immunohistochemistry data in the revised Fig 3. We also quantified neurogenesis markers Atoh7, Otx2, Rbpms (RGCs), and Crx at E13 in revised Fig 6 (former Figs 6 and 7). Older stages were moved to a new Suppl Fig 5.

Respectfully, Hes5 mRNA expression in old Fig 1F and 1G shows that Hes5, like other retinal progenitor cell (RPC) markers, expanded in Rax-Cre deletion but not Chx10-Cre deletion conditions. This is analogous to Pax6 and Rax expansion in Rax-Cre;Hes1 CKO eyes and Pax2 mutants (doi: 10.1523/JNEUROSCI.2327-19.2020) (1). In revised Fig 1, we now show analogous expansion of Hes5 mRNA in Pax2 mutant retinas (compare Figs 1F-1I). Because Hes5 RNA in situ hybridization experiments are nonquantitative, we do not discuss the possibility of Hes5 mRNA level changes in labeled cells.

The authors reported many exciting results. However, further mechanistic insights are largely missing. They may focus on one of these exciting findings and give some mechanistic insights. For example, hes1 suppresses hes5 expression as the ONH boundary forms; hes1 expression in the ONH is Notch independent; differential influences of Rbpj and Hes1 on cone development. It is better for the authors to select one of these exciting findings and provide a deeper mechanistic study.

Response: This revision brings fresh focus to Notch regulation of RGC and photoreceptor development, particularly differential influences for Rbpj versus Hes1. We also better support our interpretation of image data in Fig 1. We include new data about the spatial relationships between Hes5-GFP/Pax2 and Hes5-GFP/Foxg1. In summary, we find that as Pax2 becomes restricted to the nasal optic cup prior to the onset of RGC genesis, it becomes mutually exclusive with Hes5-GFP, at the same time that Hes5-GFP+ cells coexpress Hes1. This is consistent with Hes1 indirectly regulating Hes5-GFP as a marker of neurogenic RPCs at the forming ONH. Furthermore, it emphasizes the importance of genetically teasing apart the separate and potentially compensatory roles for Hes1 versus Hes5 undertaken here. These relationships remain poorly resolved during vertebrate CNS development.

Some analyses lack an explanation of the rationale. For example, "To understand if the loss of multiple Hes genes is more catastrophic than Hes1 alone..."(PAGE 7). Please explain its significance.

Response: We assume the reviewer is referring to the first sentence of the last paragraph on this page. We analyzed Hes triple mutant mice (TKO) to understand if removing multiple Hes genes reveals redundant functions. This is an open question, given that Hes1 is expressed in the ONH/OS, which is normally devoid of Hes5 by the time retinal neurogenesis begins. These questions have only been explored in a handful of tissues throughout the body. Also see response to point 2 above. In general, we have expanded the rationale for all of the experiments throughout the revised manuscript.

Significance: In general, many results are quite interesting. However, the significance of these findings is largely hampered in the following aspects: 1. The authors were unable to provide the sufficient research contexts that are essential for understanding many results.2. Many conclusions were solely based on descriptive images but lacked statistical quantification, which significantly weakened many conclusions. 3. Many interesting findings are quite descriptive, and some mechanistic understandings of one of these exciting findings will be beneficial to improve the focus and significance of the study. Current format of the manuscript fits more specialized audience.

Response: During in vivo development, we wished to understand which particular Notch pathway genes can interact in a Notch-dependent versus a Notch-independent manner. Genetic (phenotypic) studies produce extremely rigorous datasets, in our opinion. This revision now extensively quantifies key findings. Here we dissected the "receipt" of a Notch signal by identically testing the functional requirements of particular pathway members. For Mastermind (Maml), there are 3 paralogues, double mutants for Maml1 and Maml3 are early lethal, and no floxed alleles exist, so it was logical to employ the ROSA-dnMaml mouse strain, particularly since it has been discussed throughout the Notch literature as "analogous" to removing either a Notch receptor or Rbpj. Our finding that the dnMAML allele does not function like a Rbpj null in the retina is important for researchers in the broad Notch field to consider when designing and interpreting experiments.

Reviewer #2: Hes genes are effectors of the Notch signaling pathway but can also act down-stream of other signaling cascades. In this manuscript the authors attempt to address the complexity of Hes effectors during optic cup development and retinal neurogenesis. To do so, they compared optic cup patterning and retinal neurogenesis in seven germline or conditional mutant mouse embryos generated with two spatio-temporally distinct Cre drivers. These lines allowed for the analysis of the consequences of perturbing the Notch ternary complex and multiple Hes genes alone or in combination. The authors show that the optic disc/nerve head is regulated by Notch independent Hes1 function. They also confirm that perturbation of Notch signaling interferes with cell proliferation enhancing the production of differentiated ganglion cells, whereas photoreceptor genesis requires both Rbpj and Hes1 with Notch dependent and independent mechanisms. This is a rather complex study that dissects further the role of the Notch pathway and Hes proteins during eye development, a topic that has been addressed in many previous studies but perhaps not with the details that the authors have used here. In this respect, this study adds to current literature but will likely be of interest to retina aficionados. The manuscript reads well and the figures are of very good quality. However, many of the statements are based on qualitative rather than on quantitative analysis. This should be, at least in some cases, remediated, despite the effort that this may require given the number of mouse lines used in the study.

Response: As described in the response to Reviewer 1, we agree and present considerably more quantification data. We extensively reorganized and rewrote this manuscript to emphasize that Hes1 in the ONH/OS is fully Notch-independent and highlight branchpoints in Notch-dependent signaling, for Rbpj versus Hes,1 during early retinal neurogenesis. It is too simplistic that the ternary complex (Rbpj-NICD-Maml) simply activates Hes1 (and/or multiple Hes genes) to regulate downstream signaling targets. This paradigm has been portrayed in the literature numerous times for many processes throughout vertebrate development, homeostasis or relative to particular diseases. By focusing on one tissue and a narrow window of development, our phenotypic studies delved more deeply to show the greater complexity and molecular cross-talk that we think underlie the modulation of signaling levels with in vivo context. Thus, our results are of broad interest and impact to the greater Notch field.

1. The title is somewhat misleading. The authors have explored mostly the role of Hes1, 3 and5. Although these are Notch effectors, there is already evidence that they participate in other pathways This is confirmed by the data present here. I would suggest to eliminate Notch from the title and use instead "Hes" to better reflect the findings. Furthermore, it is unclear why there is a reference to "mutations" or what are the Notch branchpoints to which the authors refer at the beginning of the discussion.

Response: We appreciate the reviewer’s viewpoint but disagree this paper is mostly about Hes genes, as there is a critical direct, comparable evaluation with Rbpj and dn-Maml. Direct comparison of 7 genotypes highlights where each pathway member exhibits idiosyncratic phenotypes. We are striving for a clear, simple title about a very complex topic, involving the in vivo genetic dissection of a signaling pathway. We modified the title to: "Notch pathway mutations do not equivalently perturb mouse embryonic retinal development "

1. "Although the Pax6-Pax2 boundary is intact in Rax-Cre;RbpjCKO/CKO eyes, ONH shape was attenuated compared to controls (Fig 3I)". This statement is arguable as the difference seems subtle. Perhaps some kind of quantification would help.

Response: We quantified Pax2+ cells (ONH domain) using the adjacent proximal terminus of the retinal pigmented epithelium (RPE) to indicate a transition from ONH to optic stalk (OS). We also quantified the number of Pax2+Pax6+ double positive cells where the 2 domains abut (boundary cells). Some higher magnification examples are now provided in Fig 3H';3K';3N'. Grossly, the imaging data support that the Pax2+ ONH is expanded in Chx10-Cre;TKO eyes, while boundary cells are most affected in Rax-Cre;HesTKO eyes, due to an expansion of retinal tissue. This is supported by our quantitative data (Fig 3O,3P). We observed even in controls that Pax2-expressing cells show some numerical variability. We attributed this to the position of the section through the ONH, which is a 3-dimsenional ring (torus). Therefore, we quantified additional wild-type controls and mutant samples in the new Fig 3O,3P graphs, improving statistical power, and allowing us to detect quantitative differences.

Page 12 first paragraph. "....but all other genotypes were unaffected". This statement is unclear. All lines in which the Rax-Cre has been used seem to have an increased number of apoptotic cells. This should be better explained

Response: Respectfully, only one genotype, Rax-Cre;Rbpj mutants contain a statistically significant increase in apoptotic cells (Fig 5P). This is demonstrated by one-way ANOVA analyses that included all pairwise comparisons. To ensure that the quantification was not misleading due to changes in tissue morphology, data in Figs 5, 6, and 7 were normalized to optic cup area. The area was traced in FIJI, creating a polygon whose area was determined in square microns. For every section image, the marker+ cells were divided by the square micron area of the retina (excluding the opening for the optic nerve). Such a method is critical for comparison across this allelic series, given the morphologic changes, differences in cell clustering where rosettes form, and reduced proliferation whenever Notch signaling is lost or reduced.

Page 12, end of second paragraph: "E13.5 Chx10-Cre;HesTKO eyes had a milder RGC phenotype (Figs 6G, 6N, 6U), but all other mutants were unaffected (Figs 6E, 6F, 6L, 6M, 6S, 6T). This statement is also rather subjective. The phenotype of Chx10-Cre;HesTKO is quite strong and the other mutants seem to have a phenotype. Some quantifications here will help.

Response: We agree and provide quantification for both Atoh7 and Rbpms positive cells in the revised Figure 6. This is now in the same figure with quantification of Otx2+, Otx2+Atoh7+ and Crx+ cells. The reviewer is correct that both ROSA-dnMaml and both HesTKO mutants have a statistically significant increase in RGCs. Surprisingly, neither of the Rbpj CKO mutants have this outcome (Fig 6Y).

1. Page 13, toward the bottom..."...but noted that Chx10-Cre RbpjCKO/CKO eyes were not different from controls (Figs 7E, 7AA)". Again, this statement is questionable as staining for both CRX and Rbpms seem reduced as compared to controls as quantifications in 7AA seems also to indicate (about half?). Did the authors calculate whether there is a statistical difference between controls and Chx10-Cre RbpjCKO/CKO ?

Response: Rbpms+ RGCs and Crx+ photoreceptor precursors were colabeled and quantified on sections for all genotypes. All counts were normalized to area as described above. Upon quantification and ANOVA with pairwise comparisons, there was no statistical difference in Crx+ or Rbpms+ cells between control and Chx10-Cre;Rbpj mutants (new Fig 6Y and Z).

In Fig 7CC the authors should make the effort of including at least one additional sample, 2 biological replicates seem insufficient to draw a conclusion.

Response: The Rax-Cre;Hes1CKO/+ X Hes1CKO/CKO matings stopped producing litters in late 2022. While this manuscript was out for review, we obtained younger mice, from which new control and Rax-Cre; Hes1 mutant littermates were collected, stained, imaged and quantified. Upon adding samples, we found that the outcome was unchanged, but the data better support the lack of a statistical difference in rods between genotypes at E17. These data were moved to revised Suppl Fig 5.

Significance: This is a rather complex study that dissects further the role of the Notch pathway and Hes proteins during eye development, a topic that has been addressed in many previous studies but perhaps not with the details that the authors have used here. In this respect, this study adds to current literature but will likely be of interest to retina aficionados. The manuscript reads well and the figures are of very good quality. However, many of the statements are based on qualitative rather than on quantitative analysis. This should be, at least in some cases, remediated, despite the effort that this may require given the number of mouse lines used in the study.

Response: To increase the impact of our manuscript, we quantified all markers except Tubb3, since its localization in cell bodies and axons make it impossible to assign to individual cells. We feel that this additional quantification strongly improves the quality of our findings and allowed us to make well-supported and novel conclusions. While we certainly believe that the retinal development community will find this paper of interest, it will also be of value to the broader Notch pathway scientific community. In this manuscript, we simultaneously compared phenotypes for Notch pathway genes in signal receiving cells. We could find essentially no studies like this for the mouse CNS and only a few from the Kopan lab about the kidney and immune system. Interestingly, one of us (NLB) is a coauthor on a recent paper about Notch signaling in the cortex, in which ROSA-dnMaml behaves analogously to Notch1CKO or RbpjCKO. This emphasizes that findings in one organ may not recapitulate the "rules" for this pathway for other cell types or tissues (doi: 10.1242/dev.201408)(2). Deeper understanding of how the Notch pathway in the retina functions, analogously or differently, is important. We feel our revised study advances when and where there are "branchpoints" in canonical signaling that may be overlooked in other developing tissues and organs.

Reviewer #3: I have reviewed a manuscript submitted by Bosze et al., which is entitled "Not all Notch pathway mutations are equal in the embryonic mouse retina". The authors focused on Notch signaling pathway. Notch signaling is deeply conserved across vertebrate and invertebrate animal species: in general, two transmembrane proteins, Delta and Notch, interact as a ligand and a receptor, respectively, which induces proteolytic cleavage of Notch receptors to generate Notch intracellular domain (NICD). NICD is translocated into nucleus, then forms the transcription factor complex including Rbpj (also referred to as CBF1) and Mastermind-like (Maml), and activates the transcription of Hes family transcription factors. Three Hes proteins, Hes1, 3, and 5, are important for nervous system development. In the vertebrate developing retina, these Hes proteins inhibit neurogenesis to maintain a pool of neural progenitor cells. In addition to their primary role in neurogenesis, the authors recently reported that Hes1 promotes cone photoreceptor differentiation. In the later stages of development, Hes proteins also promote Müller glial differentiation. In addition, Hes1 is highly expressed in the boundary between the neural retina and optic stalk and required for this boundary maintenance. To understand precise regulation of Notch component-mediated signaling network for retinal neurogenesis and cell differentiation, the authors compared retinal phenotypes in the knockdown of three Notch pathway components, that is (1) Hes1/3/5 cTKO, (2) Rbpj KO, and (3) dominant-negative Maml (dnMaml) overexpression, under the control of two Cre derivers; Rax-Cre and Chx10-Cre. First, the authors found that Hes1 expression in the boundary between optic stalk and neural retina is lost in Rax-Cre; Hes1/3/5 cTKO, but still retained in Rax-Cre; Rbpj KO and Rax-Cre; dnMaml overexpression, suggesting that Delta-Notch interaction is not required for Hes1 expression in the boundary between optic stalk and neural retina. Furthermore, Hes1 expressing boundary region expands distally at the expense of the neural retina in Chx10-Cre; Hes1/3/5 cTKO. Maintenance of ccd2 expression in this expanded boundary area suggests that Hes1 normally maintains a proliferative state in the optic stalk, which may allow these cells to differentiate into astrocyte in later stages. Second, in addition to precocious RGC differentiation in all the Notch component KO, the authors found that, as compared with wild-type, cone and rod photoreceptor genesis is highly enhanced in Rax-Cre; Rbpj KO and Rax-Cre; dnMaml overexpression and mildly enhanced in Chx10-Cre; dnMaml overexpression. On the other hand, in Rax-Cre; Hes1/3/5 cTKO, cone and rod photoreceptor genesis is not enhanced but similar to wild-type level. Since the authors previously reported that cone genesis is reduced in Rax-Cre; Hes1 cKO and Chx10-Cre; Hes1 cKO, so Rax-Cre; Hes1/3/5 cTKO may rescue decrease in cone genesis in single Hes1 cKO. The authors raise the possibility that elevated Hes5 expression in single Hes1 cKO may suppress cone photoreceptor genesis. The authors also found that amacrine cell genesis is significantly suppressed in Rax-Cre; Rbpj KO but not changed in Rax-Cre; dnMaml overexpression and Rax-Cre; Hes1/3/5 cTKO, suggesting that Rbpj is specifically required for amacrine cell genesis. From these observations, the authors propose that there are at least two branchpoints for photoreceptor and amacrine cell genesis in Notch component-mediated signaling network. Their findings are very interesting and provide some new insight on how Notch signaling components are integrated into other signaling pathways and promote to generate diverse but well-balanced retinal cell-types during retinal neurogenesis and cell differentiation, in addition to conventional classic view of Notch signaling pathway. However, one weak point is that, although the authors figured out what kinds of phenotypic difference appear in the KO retinas between these Notch components, the research result is descriptive and less analytical. Most of their conclusions may be supported by their previous works or others; it is still hypothetical. So, it is important to show more analytical data to support their interpretation and more clearly show what is new conceptual advance for Notch signaling pathways.

For example, sustained Hes1 expression in the boundary region between optic stalk and neural retina may be reminiscent to brain isthmus situation. I would like to request the authors to show more direct evidence that Hes1 regulation in optic stalk/retina boundary is independent of Delta-Notch interaction. One possible experiment is whether DAPT treatment phenocopies Rax-Cre; Rbpj KO and Rax-Cre; dnMaml overexpression (Hes1 in optic stalk boundary is normal?).

Response: Usage of the gamma secretase inhibitor DAPT is an interesting experiment as it can phenocopy the loss of Notch signaling in developing tissues. However, the reviewer's proposed DAPT experiment is problematic for two major reasons. First, DAPT blocks the gamma secretase complex, which has more than 90 protein targets in the cell membrane (3). Therefore, DAPT may not be informative for Hes1 regulation given the myriad of expected off-target effects. Second, it would be difficult to treat embryos at the relevant stages with DAPT. Injections into pregnant mice are lethal and we cannot localize drug to the relevant area during in vivo development. Our direct phenotypic comparisons with two Cre drivers strongly indicate that Hes1 is independent of canonical Notch signaling in the developing optic stalk.

We include an extra related data figure (Reviewer Fig 1) showing anti-Hes1 immunolabeling of E13.5 Rax-Cre;Notch1CKO/CKO (n=2) and E13.5 Rax-Cre;Notch2CKO/CKO eyes (n=3). The Notch1 mutant lost oscillating Hes1 expression in retinal progenitors, but the uniform Hes1 ONH domain remains. Interestingly, the Notch2 mutant had essentially no effect on Hes1 (oscillating or sustained), or Hes5 mRNA expression. A Notch2 RNA in situ hybridization demonstrates that Notch2 mRNA was lost in the E13 optic cup and RPE (Rax-Cre expressing tissues). These data emphasize: A) the Notch1-specific dependency of oscillating Hes1 expression in retinal progenitors is absent from the ONH; B) although coexpressed in the same tissue, Notch receptors have unequal activities.

Does Rax-Cre; Rbpj KO; Hes1-cKO phenocopy Rax-Cre; Hes1-cKO (or Rax-Cre; Hes1/3/5 cTKO)?

Response: This is a good question! The first author tried very hard to produce Rax-Cre; Rbpj CKO;Hes1 CKO double mutant embryos. However, these progeny could not be recovered from E10-E13 embryos, despite collecting more than 10 litters. Thus, it is likely that this genotype is lethal before eye formation.

Could the authors identify an enhancer element that drives Hes1 transcription in optic stalk/retina boundary, which should be not overlapped with that of NICD/ Rbpj binding motif? Such additional evidence will make their conclusion more convincing.

Response: Another interesting question. We have been working for >3 years on Hes1 cis regulatory enhancers, but the pandemic greatly delayed progress. The proximal Hes1 600bp upstream region is a generic enhancer that contains Hes1 binding sites for repressing its own expression (4) and has a pair of Rbpj consensus sites for Notch ternary complex activation of Hes1 expression (5,6). Nearby is a binding site occupied by Gli2 in the E16 mouse retina (7). Recently, it was shown that Ikzf4 binds slightly farther away (8). The upstream 1.8 kb region (including the 600bp just described) can drive destabilized GFP or dsRed reporters in early postnatal retinal explants (9). However, this sequence was used to make and analyze a classic Hes1-GFP transgenic reporter mouse, in which GFP was not expressed in the early embryonic mouse optic vesicle or cup (10). Therefore, any early eye-specific enhancer(s) are located farther upstream, in an intron, or downstream (or combination thereof). Public domain epigenetic and chromatin accessibility datasets support this idea. Identifying the gene regulatory logic for Hes1 expression in the eye will be an exciting future story, well beyond this manuscript. We are excited to use live imaging of enhancer reporters to discern oscillating versus sustained activity patterns during early ocular development.

Regarding the conclusion on new branchpoints on photoreceptor and amacrine cell genesis, a model shown in Figure 9 is still hypothetical. Figure 9B indicate a model in which the increase of Otx2+ cells and Crx+ cells in Rax-Cre; Rbpj KO is mediated by Hes1, which is presumed to be activated in Notch-independent signaling. However, Hes1 expression in the neural retina is markedly reduced in Rax-Cre; Rbpj KO (Fig. 2I), which does not fit in with the model.

Response: We removed Fig 9B and now present new models about the Notch-dependent versus -independent roles for both Rbpj and Hes1. The new summary is Fig 8.

So, I would like to request the authors to examine whether the increase of Otx2+ cells and Crx+ cells in Rax-Cre; Rbpj KO, (or Rax-Cre; dnMaml overexpression and Chx10-Cre; dnMaml overexpression) is inhibited by Hes1 KO.

Response: If we understand this correctly, it would mean generating double mutants, some of which we determined are not viable (see the response above, and Suppl Table 2). Given there is only a partial knockdown of Hes1 or Hes5 in either dnMaml mutant we do not believe repeating this in the Hes1 CKO genetic background to be informative and it would take 3 generations to perform.

Second, the authors concluded that both cone and rod genesis are enhanced in Rax-Cre; Rbpj KO by showing the data on Crx/Nr2e3 labeling in Rax-Cre; Hes1 cKO in Fig. 7BB. However, as the authors mentioned in the manuscript, Hes5 expression is elevated in Rax-Cre; Hes1 cKO (Fig. 1G). So, since Rax-Cre; Hes1 cKO has residual Hes activity in the retina, Fig. 7BB should be replaced with labeling of Crx/Nr2e3 in Rax-Cre; Hes1/3/5 cTKO.

Response: Unfortunately, Rax-Cre;HesTKO embryos do not live past E13 (Suppl Table 2). Thus, we cannot evaluate rods, whose genesis starts around E13.5. Revised Fig 1G shows the Hes5 domain is shifted with the expansion of retinal tissue in E13.5 Hes1 single mutants, but importantly, also analogously shifted in Pax2 mutants (Fig 1H). We do not conclude that mRNA levels are "elevated" since mRNA in situ hybridization is not a quantitative technique. Our initial examination of rods in E17 Rax-Cre;Hes1 CKO mutants tested the idea of a fate shift from cones to rods. However, deeper quantification (Suppl Fig 5) do not support such a fate change.

Furthermore, possibly, it is best to examine labeling of the retinas of Rax-Cre; Rbpj KO with rod and cone-specific markers and confirm that the number of both rods and cones is significantly increased. Third, as for defects in amacrine cells genesis in Rax-Cre; Rbpj KO, I would like to request the authors to show the data on Crx10-Cre; Rbpj KO. Although Rbpj KO is mosaic in Crx10-Cre; Rbpj KO, we can distinct Rbpj KO cells by GFP expression (Fig. S2C, C', C'). So, the authors can confirm that amacrine cell genesis is inhibited in a cell-autonomous manner in Crx10-Cre; Rbpj KO retinas but not in Crx10-Cre; dnMaml overexpression. Addition of such data will make the authors' conclusion is more convincing.

Response: Suppl Table 1 lists multiple references (two from the NLB lab) that demonstrated both a rod and cone increase in Rbpj loss-of-function conditions. Chx10;Rbpj CKO animals were evaluated by Zheng et al., who showed an amacrine loss phenotype in these mutants (11). This is equivalent to what we see in our Rax-Cre;Rbpj CKO data, but without the complications of Chx10 mosaic Cre expression upon Rbpj deletion.

Other comments: 1) Title of this manuscript is "Not all Notch pathway mutations are equal in the embryonic mouse retina". However, this title is quite obscure in what is research advancement of their findings. I suggest the authors to include more concrete and conclusive sentence in the title, for example "Hes and Rbpj differentially promotes retina/optic stalk boundary maintenance and photoreceptor genesis, in parallel with neurogenic inhibition by Notch signaling pathway".

Response: We appreciate the reviewer's perspective. We are striving for a relatively simple title about a very complex topic, involving the in vivo genetic dissection of a signaling pathway. We modified the title to "Notch pathway mutations do not equivalently perturb mouse embryonic retinal development ".

2) The "Results" section is a bit difficult to follow logics without detailed knowledge on roles of Notch signaling in mouse retinal development. I suggest the authors to improve a writing style of "Results" section for readers without such detailed knowledge on mouse Notch mutant phenotypes to follow logical flow more easily. There are many additional descriptions on research background before start to mention results. Such introductory sentences should be moved to the "Introduction" section, by which logical flow in the Results section should be simpler. In addition, the authors should show a concrete question at the beginning of each result subsection. Furthermore, the authors sometimes jump over from one result subsection and suddenly move to cite another figure panel in a far ahead subsection whose data has not been explained. Such a back-and-forth citation of figure data generally makes it difficult to follow logical flow.

Response: We now present a considerable amount of new quantified data, reorganized multiple figures, and extensively rewrote the paper. We significantly revised the summary figure to improve clarity. In addition, Suppl Table 1 provides a wealth of background information to orient the reader on this topic. We feel that this extensive revision has greatly improved the quality, logical flow, and readability of the manuscript.

3) In addition, figure configuration is not well organized. Each figure compared some particular marker expression in wild-type, Rax-Cre; HesTKO, Rax-Cre; Rbpj cKO, Rax-Cre; dn-Maml-GFP, Chx10-Cre; HesTKO, Chx10-Cre; Rbpj cKO, Chx10-Cre; dn-Maml-GFP. For example, Fig. 2 shows Hes1 for inhibition of neurogenesis, Fig. 3 shows Vsx2; Mitf and Pax2; Pax6 for retinal pigmented epithelium and optic stalk, Fig. 6 shows Atoh7, Rbpms, and Tubb3 for retinal ganglion cells. Fig. 7 shows Crx, Otx2, and Thrb2 for photoreceptor differentiation. Fig. 8 shows Prdm1, and Ptf1a for photoreceptors and amacrine cells. Although this figure configuration is convenient to show phenotypic difference between different genetic mutations, it is difficult to know how each differentiation steps are spatially and temporally coordinated during development. At least, I recommend the authors to show one summary figure, which shows spatio-temporal expression profile of retinal markers in wild-type mouse retinas.

Response: We recognize this point and completely reorganized and combined Figs 6 and 7 to improve clarity. New Figure 6 presents E13 quantification for Atoh7, Otx2, Atoh7/Otx2, Rbpms and Crx expressing retinal populations. E16-E17 data were condensed and moved to a new Suppl Fig 5.

4a) Page 7, line 7-10 "With earlier deletion using Rax-Cre, hes5 mRNA abnormally extended into the optic stalk": I wonder how the authors define the optic stalk. It is likely that optic stalk area (Pax2+, Vax1+ area) is shifted to more proximal (depart from the optic cup and move toward the brain), and neural retina is expanded accordingly (Fig. 4B, 4F), resulting in expansion of hes5 expression. Thus, it may be better to mention that optic stalk/neural retina boundary is abnormally shifted toward the brain.

Response: The retina, including the optic nerve head, ends where the adjacent RPE terminates. This is conspicuous morphologically in our sections. We also defined this by colabeling for Pax2 and Pax6, which is now quantified in revised Fig 3. To clarify this further, we added the words " in all panels the brain is to the right" in the Fig 4 legend.

4b) Page 8, line 14-15, "ONH/OS cells still express it (Hes1), demonstrating that sustained Hes1 is independent of Notch": I presume that Cre-Rax drives Cre in neural retina as well as optic stalk and pigmented epithelium. However, it is likely that Rbpj is not expressed in optic stalk/neural retina boundary area in wild type (Fig. S2A). No expression of Rbpj in optic stalk/neural retina boundary may support that Hes1 expression in this boundary area is Notch-independent. However, Rbpj expression is retained in some vitreal cells near optic nerve head in Rax-Cre; Rbpj-CKO retinas (Fig. S2B). What are these Rbpj+ cells? I would like to request the authors to confirm that Rbpj expression is completely absent in both neural retina and optic stalk in Rax-Cre; Rbpj-CKO mice. Otherwise, this conclusion is still not fully supported.

Response: We show the Rax-Cre lineage in Suppl Fig 2 via the Ai9 (tomato) reporter. The results are striking, with all of the optic cup derivatives (retina, RPE, ONH, optic stalk, and presumptive ciliary tissue and iris) being tomato positive, while the well-described population of vascular cells in the hyaloid space lack tomato expression. Furthermore, our figure shows that Rbpj expression is only absent from the optic cup derivates, rather than the vascular structures in the vitreous. Vascular cells also depend on the Notch pathway and express Rbpj. Based on considerable evidence from the literature and our lineage experiments, the population of cells the reviewer highlights represents the hyaloid vasculature and associated cell types. It does not represent any population that derives from neuroectoderm.

4c) Page 9, line 16-18, "Foxg1 had spread into the nasal optic stalk": Is Foxg1 expanded nasal area really "OS" rather than expanded retina? I suggest the authors to confirm molecular markers Pax2 expression is overlapped with Foxg1. Otherwise, it is difficult to conclude that foxg1 is expanded into the optic stalk territory, because foxg1 is normally a marker of retina. Indeed, Fig. 3K shows pax2 expression is shifted into more inside towards the brain, suggesting that neural retina is expanded. Please explain the situation.

Response: Foxg1 (BF-1) mRNA and protein are found in the nasal retina and are expressed in other brain tissues. Multiple studies show Foxg1 in the nasal side of the E10 optic cup/retina/optic stalk and developing hypothalamus (See extra data figure Reviewer Fig 2; top row figure is data from Smith et al., 2017 (12) with Foxg1 mRNA in purple. Also see our new manuscript panel Fig 1C. We include here for reviewers (extra data Reviewer Fig 2 showing E13 ocular cryosections colabeled for Foxg1 and Pax2, highlighting their relationship in the retina, optic stalk and adjacent forming hypothalamus. On page 9 the text now reads "At E13.5 Rax-Cre;HesTKO eyes, the Foxg1 nasal retinal domain was contiguous with the nasal optic stalk (Suppl Fig 4D). This is reminiscent of younger stages (Fig 1C), since normally at E13.5, Foxg1 in the nasal optic cup/retina is separated from expression in the ONH/OS (Suppl Fig 4A). Based on the expansion of Pax6, Vsx2 and Hes5 RPC domains into the optic stalk, we conclude that the change in Foxg1 similarly reflects an extension of retinal tissue."

4d) Page 10, line 4-5, In Rax-Cre; Hes1/3/5 cTKO eye, this tissue (RPE) extended into the optic stalk": This description seems to be incorrect. A part of Pax2 area, which is adjacent to the neural retina, contacts with RPE in wild type (Fig. 3AH), so most of RPE covers the neural retina even in Fig. 3DK.

Response: We disagree with the reviewer’s interpretation. Fig 3D shows Mitf labeling of RPE nuclei. Figure 3K shows the adjacent section labeled with Pax2 and Pax6 (labels both retina and RPE). As the retina extended "towards the brain", the RPE analogously extends and surrounds the retinal domain. We also added higher magnification data panels 3H, 3K and 3N, showing merged and single channels.

4e) Page 10, line 22-23, "For Chk10-Cre; Hes1/3/5 cTKO, there was a unique presence of ectopic Pax2 within the retinal territories": I wonder if this description is correct. I suspect that proliferative Pax2+ cells expand into regressing territory of Hes KO retinal cells, which undergo precocious neurogenesis and lose proliferative activity, in Chk10-Cre; HesTKO. In this case, it is possible that the Pax2/Pax6 interface may be maintained. Please show red and green channel panels for Fig. 3N to confirm that there is ectopic pax2 and pax6 double positive cells.

Response: New quantification in revised Fig 3 (see panels O,P) fully supports our original conclusion. Only Chx10-Cre;HesTKO mutants have a statistically significant increase in Pax2+ cells. There are not more Pax2+Pax6+ double labeled cells. Only this particular genotype has an increase in Pax2+ single labeled cells.

5a) Page 11, line 20-25. There seems to be inconsistency between result description and image data of Fig. 5A-G, and histogram Fig. 5O. Authors mentioned that a modest loss of pH3+ cell fraction in Chx10-Cre; Hes1/3/5 cTKO but not in Rax-Cre; Hes1/3/5 cTKO. However, Fig. 5D indicates severe reduction of pH3+ cell fraction in Rax-Cre; Hes1/3/5/ cTKO, which is similar to reduction of pH3+ cell fraction in Rex-Cre; Rbpj (Fig. 5B), but histogram data is different (Fig. 5O). Furthermore, pH3+ cell fraction is severely reduced in Chx10-Cre; ROSA(dn-Maml-GFP) (Fig. 5F) and modestly reduced in Chx10-Cre; Hes1/3/5 cTKO (Fig. 5G). However, pH3+ cell fraction seems to be normal in Chx10-Cre; Rbpj (Fig. 5E). These Chx10-Cre image data do not match the histogram of Fig. 5O. Please check their situation.

Response: Images in old Figs 5-8 were normalized using area measurements, see methods and above comments (note: old Figs 6&7 were combined into new Fig 6). One-way ANOVA with pairwise comparisons for each mutant genotype compared to control were calculated using Prism. All genotypes except two have a statistically significant loss of M phase cells and we discuss possibilities for this outcome (Fig 5O). A normalization method for the sampled area is an essential component of these studies since morphologic differences are apparent for particular genotypes. The quantitative data are consistent with our original conclusions.

5b) Fig. 5H-N, P: I wonder if the stage E13 is appropriate to evaluate cell death and survival because optic cup already becomes smaller in Rax-Cre; Rbpj, Hes1/3/5 cTKO, or ROSA(dn-MAML-GFP) than in wild-type control. I suggest the authors examine more earlier stage.

Response: While an earlier effect is possible, we only observed size differences in a subset of the genotypes. Thus, E13 serves as a critical timepoint to examine early developmental phenotypes across the totality of our mutant conditions. It is also first age when the ONH is fully formed.

5c) Page 12, line 19-20, "all other mutants (Chx10-Cre; Rbpj, and Chx10-Cre; ROSA(dn-MAML-GFP) were unaffected (Fig. 6EF, LM, ST)": It is likely that atoh7 expressing cells are mildly decreased and neuronal marker, Tubb3 and Rbpms-expressing cells are increased in Chx10-Cre; Rbpj, and Chx10-Cre; ROSA(dn-MAML-GFP). I requested the authors to evaluate the fraction of these markers in retinal area statistically in all the cases.

Response: As described above, we quantified Atoh7 and Rbpms nuclear expression by immunohistochemistry. We do not believe that Tubb3+ cells can be reliably quantified. Nonetheless, it is useful to qualitatively show the extent of excess neuron formation. Importantly, we observed that it is not the Atoh7 status that matters for RGC formation, rather it is the Otx2 expression status. This is in good agreement with single cell-RNA transcriptomics data from Wu et al 2021 showing that Atoh7 mRNA in all early transitional RPCs remains fairly constant and its loss does not block the formation of early RGC cell states (13). By contrast Otx2 fluctuates but remains expressed in transitional RPCs that progress to photoreceptor lineages.

6a) Page 7, line 19 "Ectopic blood vessels protruded from the ONH (Fig. 1K, 1L)": It is difficult to see blood vessel structures in these panels (Fig. 1I-L). Please show some molecular marker of blood vessels to confirm how blood vessel is organized in Hes1/3/5 cTKO.

Response: These vascular structures are highly conspicuous by morphology in the H&E insets. Nonetheless, we used adjacent P21 sections to immunolabel for Endomuscin (14) and Tubb3 antibodies. This colabeling confirms the morphology and position of ectopic blood vessels in the abnormal tissue masses in Chx10-Cre;HesTKO mutant eyes. Ectopic tissue contains only rare Tubb3+ cells or cell processes suggesting it is overwhelmingly nonneural. All P21 data were moved to a new Suppl Fig 2. A full detailing of vascular phenotypes is beyond the scope of this manuscript and, interestingly, would be potentially attributable to non-autonomous effects of perturbing the Hes genes in the adjacent retina.

6b) Fig. 5: Increase of pH3 fraction indicates several possibilities, for example (1) increased fraction of mitotic cells due to precocious neurogenesis, (2) increased fraction of mitotic cells due to activated cell proliferation of retinal progenitor cells, (3) increased cell-cycle arrest in M phase due to some stress response of progenitor cells. So, I suggest the authors to examine (1) BrdU percentage of retinal section area, (2) the percentage of pH3+ cells in PCNA+ retinal cells.

Response: The data listed in Suppl Table 1 presents a unified picture that disrupting Notch signaling reduced proliferation. This paradigm extends to other model organisms (e.g., Drosophila, chick, frog, zebrafish and even to nonneural tissues). We included the phospho-histone H3 staining so readers would see how the six mutants evaluated in this study align with this paradigm, providing confidence for the novel findings in other figures. A full evaluation of cell cycle kinetics is interesting, but beyond the scope and focus of this manuscript.

6c) Fig. 5: It is better that cell death fraction will be evaluated by TUNEL and labeling with anti-activated caspase 3 antibody.

Response: We disagree. The DNA repair enzyme PARP is inactivated upon cleavage by activated caspase 3. There are currently ~3,600 citations that use it as a marker of apoptosis. PARP also has a separate and very specific role in maintaining the integrity of sperm DNA. This antibody works on all metazoans and is amenable to many tissue preparations and fixatives, making it easy to use, robust and quantifiable.

7a) Please show red channel (Hes1) image in Fig1BC.

Response: This was added to Revised Fig 1 (Fig 1A).

7b) Fig. 1DH should be shown in neighbor. Fig. 1H should be assigned as Fig. 1E.

Response: The new Fig 1 layout addresses this point.

7c) Fig. S2D, F, H, J: Please show GFP green channel as well. Otherwise, it is difficult to see non-overlapping expression in optic stalk area.

Response: In the revision, this is Suppl Fig 3. Chx-10-Cre is not expressed by ONH-OS cells (1). The green and fuchsia overlap (coexpression) in RPCs is white, we feel this is fairly clear. If needed, all readers can turn on and off the green channel in the final PDF version of this figure to compare GFP with Hes1 expression for those panels.

7d) Fig. 9B: It is better to show Rax-Cre: Hes1/3/5 TKO rather than Rax-Cre: Hes1 cKO. 7e) Fig. 9B: Lettering "Rbpj mutant" should be revised as "Rax-Cre: Rbpj KO".

Response: Fig 9B was removed so these terms are now irrelevant. Our models are presented in new Fig 8.

Significance: The senior author of this manuscript, Dr. Nadean Brown, is an expert scientist who has investigate the role of Notch signaling pathway in vertebrate ocular tissue, including the neural retina and lens. In general, Notch signaling pathway consists of signaling stream from the interaction of Delta and Notch, Notch receptor activation by proteolytic cleavage, translocation of Notch intracellular domain (NICD) into nucleus, formation of transcription factor complex consisting of NICD/Rbpj/Maml, to the transcriptional activation of Notch target genes, Hes family transcription factors. Finally, Hes suppresses neurogenic program and maintain a pool of neural progenitor cells. Therefore, Notch is a key factor to regulate the balance between neurogenesis and progenitor proliferation. In this manuscript, the authors investigated retinal phenotypes in the knockout mice of different Notch signaling components, including Rbpj, Maml, and Hes. They found that functions of these three factors are not always equal in retinal cell differentiation; rather, they specifically regulate a particular step of retinal development. The authors propose the possibility that each of Notch signaling components may be modified by other signaling pathways and achieve some new roles beyond the conventional frame of classic Notch signaling pathway. In this point, this work has a potential to provide a new conceptual advance in the field of developmental and cell biology.

We fully agree this work is a significant advance for the fields of developmental and cell biology. Our findings provide new information and stimulate fresh ideas for anyone working on signal transduction and signal integration.

References cited:

1. Bosze et al., 2020 Journal of Neuroscience Vol 40:1501-13; Bosze et al. 2021 Dev Biol Vol 472:18-29.
2. Han et al., 2023 Development Vol 150 dev201408.
3. Kopan and Ilagan, 2004 Nat Rev Cell Biol. Vol 5:499-504
4. Hirata et al., 2002 Science Vol 298:840-3
5. Friedmann and Kovall, 2010 Protein Sci. Vol 19:34-46
6. Ong et al., 2006 JBC Voll24:5106-19
7. Wall et al., 2009 J Cell Biol. Vo 184: 101-12.
8. Javed et al., 2023 Development Vol 150:dev200436
9. Matuda and Cepko 2007 PNAS Vol 104: 1027-1032
10. Ohtsuka et al., 2006 Mol. Cell Neurosci. Vol 31:109-22
11. Zheng et al., 2009 Molecular Brain Vol 2:38
12. Smith et al., 2017 Journal of Neuroscience Vol 37:7975-93.
13. Wu et al., 2021 Nature Communications Vol 12:1465: doi 10.1038/s41467-021-21704-4
14. Saint-Geniez et al., 2009 IOVS Vol 50: 311-21.

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

I have reviewed a manuscript submitted by Bosze et al., which is entitled "Not all Notch pathway mutations are equal in the embryonic mouse retina". The authors focused on Notch signaling pathway. Notch signaling is deeply conserved across vertebrate and invertebrate animal species: in general, two transmembrane proteins, Delta and Notch, interact as a ligand and a receptor, respectively, which induces proteolytic cleavage of Notch receptors to generate Notch intracellular domain (NICD). NICD is translocated into nucleus, then forms the transcription factor complex including Rbpj (also referred to as CBF1) and Mastermind-like (Maml), and activates the transcription of Hes family transcription factors. Three Hes proteins, Hes1, 3, and 5, are important for nervous system development. In the vertebrate developing retina, these Hes proteins inhibit neurogenesis to maintain a pool of neural progenitor cells. In addition to their primary role in neurogenesis, the authors recently reported that Hes1 promotes cone photoreceptor differentiation. In the later stages of development, Hes proteins also promote Müller glial differentiation. In addition, Hes1 is highly expressed in the boundary between the neural retina and optic stalk and required for this boundary maintenance.

To understand precise regulation of Notch component-mediated signaling network for retinal neurogenesis and cell differentiation, the authors compared retinal phenotypes in the knockdown of three Notch pathway components, that is (1) Hes1/3/5 cTKO, (2) Rbpj KO, and (3) dominant-negative Maml (dnMaml) overexpression, under the control of two Cre derivers; Rax-Cre and Chx10-Cre.

First, the authors found that Hes1 expression in the boundary between optic stalk and neural retina is lost in Rax-Cre; Hes1/3/5 cTKO, but still retained in Rax-Cre; Rbpj KO and Rax-Cre; dnMaml overexpression, suggesting that Delta-Notch interaction is not required for Hes1 expression in the boundary between optic stalk and neural retina. Furthermore, Hes1 expressing boundary region expands distally at the expense of the neural retina in Chx10-Cre; Hes1/3/5 cTKO. Maintenance of ccd2 expression in this expanded boundary area suggests that Hes1 normally maintains a proliferative state in the optic stalk, which may allow these cells to differentiate into astrocyte in later stages.

Second, in addition to precocious RGC differentiation in all the Notch component KO, the authors found that, as compared with wild-type, cone and rod photoreceptor genesis is highly enhanced in Rax-Cre; Rbpj KO and Rax-Cre; dnMaml overexpression and mildly enhanced in Chx10-Cre; dnMaml overexpression. On the other hand, in Rax-Cre; Hes1/3/5 cTKO, cone and rod photoreceptor genesis is not enhanced but similar to wild-type level. Since the authors previously reported that cone genesis is reduced in Rax-Cre; Hes1 cKO and Chx10-Cre; Hes1 cKO, so Rax-Cre; Hes1/3/5 cTKO may rescue decrease in cone genesis in single Hes1 cKO. The authors raise the possibility that elevated Hes5 expression in single Hes1 cKO may suppress cone photoreceptor genesis. The authors also found that amacrine cell genesis is significantly suppressed in Rax-Cre; Rbpj KO but not changed in Rax-Cre; dnMaml overexpression and Rax-Cre; Hes1/3/5 cTKO, suggesting that Rbpj is specifically required for amacrine cell genesis. From these observations, the authors propose that there are at least two branchpoints for photoreceptor and amacrine cell genesis in Notch component-mediated signaling network.

Their findings are very interesting and provide some new insight on how Notch signaling components are integrated into other signaling pathways and promote to generate diverse but well-balanced retinal cell-types during retinal neurogenesis and cell differentiation, in addition to conventional classic view of Notch signaling pathway. However, one weak point is that, although the authors figured out what kinds of phenotypic difference appear in the KO retinas between these Notch components, the research result is descriptive and less analytical. Most of their conclusions may be supported by their previous works or others; it is still hypothetical. So, it is important to show more analytical data to support their interpretation and more clearly show what is new conceptual advance for Notch signaling pathways.

For example, sustained Hes1 expression in the boundary region between optic stalk and neural retina may be reminiscent to brain isthmus situation. I would like to request the authors to show more direct evidence that Hes1 regulation in optic stalk/retina boundary is independent of Delta-Notch interaction. One possible experiment is whether DAPT treatment phenocopies Rax-Cre; Rbpj KO and Rax-Cre; dnMaml overexpression (Hes1 in optic stalk boundary is normal?). Does Rax-Cre; Rbpj KO; Hes1-cKO phenocopy Rax-Cre; Hes1-cKO (or Rax-Cre; Hes1/3/5 cTKO)? Could the authors identify an enhancer element that drives Hes1 transcription in optic stalk/retina boundary, which should be not overlapped with that of NICD/ Rbpj binding motif? Such additional evidence will make their conclusion more convincing.

Regarding the conclusion on new branchpoints on photoreceptor and amacrine cell genesis, a model shown in Figure 9 is still hypothetical. Figure 9B indicate a model in which the increase of Otx2+ cells and Crx+ cells in Rax-Cre; Rbpj KO is mediated by Hes1, which is presumed to be activated in Notch-independent signaling. However, Hes1 expression in the neural retina is markedly reduced in Rax-Cre; Rbpj KO (Fig. 2I), which does not fit in with the model. So, I would like to request the authors to examine whether the increase of Otx2+ cells and Crx+ cells in Rax-Cre; Rbpj KO, (or Rax-Cre; dnMaml overexpression and Chx10-Cre; dnMaml overexpression) is inhibited by Hes1 KO. Second, the authors concluded that both cone and rod genesis are enhanced in Rax-Cre; Rbpj KO by showing the data on Crx/Nr2e3 labeling in Rax-Cre; Hes1 cKO in Fig. 7BB. However, as the authors mentioned in the manuscript, Hes5 expression is elevated in Rax-Cre; Hes1 cKO (Fig. 1G). So, since Rax-Cre; Hes1 cKO has residual Hes activity in the retina, Fig. 7BB should be replaced with labeling of Crx/Nr2e3 in Rax-Cre; Hes1/3/5 cTKO. Furthermore, possibly, it is best to examine labeling of the retinas of Rax-Cre; Rbpj KO with rod and cone-specific markers and confirm that the number of both rods and cones is significantly increased. Third, as for defects in amacrine cells genesis in Rax-Cre; Rbpj KO, I would like to request the authors to show the data on Crx10-Cre; Rbpj KO. Although Rbpj KO is mosaic in Crx10-Cre; Rbpj KO, we can distinct Rbpj KO cells by GFP expression (Fig. S2C, C', C'). So, the authors can confirm that amacrine cell genesis is inhibited in a cell-autonomous manner in Crx10-Cre; Rbpj KO retinas but not in Crx10-Cre; dnMaml overexpression. Addition of such data will make the authors' conclusion is more convincing.

Other comments are shown below.

1. Title of this manuscript is "Not all Notch pathway mutations are equal in the embryonic mouse retina". However, this title is quite obscure in what is research advancement of their findings. I suggest the authors to include more concrete and conclusive sentence in the title, for example "Hes and Rbpj differentially promotes retina/optic stalk boundary maintenance and photoreceptor genesis, in parallel with neurogenic inhibition by Notch signaling pathway".
2. The "Results" section is a bit difficult to follow logics without detailed knowledge on roles of Notch signaling in mouse retinal development. I suggest the authors to improve a writing style of "Results" section for readers without such detailed knowledge on mouse Notch mutant phenotypes to follow logical flow more easily. There are many additional descriptions on research background before start to mention results. Such introductory sentences should be moved to the "Introduction" section, by which logical flow in the Results section should be simpler. In addition, the authors should show a concrete question at the beginning of each result subsection. Furthermore, the authors sometimes jump over from one result subsection and suddenly move to cite another figure panel in a far ahead subsection whose data has not been explained. Such a back-and-forth citation of figure data generally makes it difficult to follow logical flow.
3. In addition, figure configuration is not well organized. Each figure compared some particular marker expression in wild-type, Rax-Cre; HesTKO, Rax-Cre; Rbpj cKO, Rax-Cre; dn-Maml-GFP, Chx10-Cre; HesTKO, Chx10-Cre; Rbpj cKO, Chx10-Cre; dn-Maml-GFP. For example, Fig. 2 shows Hes1 for inhibition of neurogenesis, Fig. 3 shows Vsx2; Mitf and Pax2; Pax6 for retinal pigmented epithelium and optic stalk, Fig. 6 shows Atoh7, Rbpms, and Tubb3 for retinal ganglion cells. Fig. 7 shows Crx, Otx2, and Thrb2 for photoreceptor differentiation. Fig. 8 shows Prdm1, and Ptf1a for photoreceptors and amacrine cells. Although this figure configuration is convenient to show phenotypic difference between different genetic mutations, it is difficult to know how each differentiation steps are spatially and temporally coordinated during development. At least, I recommend the authors to show one summary figure, which shows spatio-temporal expression profile of retinal markers in wild-type mouse retinas.
4. There are several logically incorrect sentences or inconsistent sentences in the results section. Please respond my comment below.
   • a) Page 7, line 7-10 "With earlier deletion using Rax-Cre, hes5 mRNA abnormally extended into the optic stalk": I wonder how the authors define the optic stalk. It is likely that optic stalk area (Pax2+, Vax1+ area) is shifted to more proximal (depart from the optic cup and move toward the brain), and neural retina is expanded accordingly (Fig. 4B, 4F), resulting in expansion of hes5 expression. Thus, it may be better to mention that optic stalk/neural retina boundary is abnormally shifted toward the brain.
   • b) Page 8, line 14-15, "ONH/OS cells still express it (Hes1), demonstrating that sustained Hes1 is independent of Notch": I presume that Cre-Rax drives Cre in neural retina as well as optic stalk and pigmented epithelium. However, it is likely that Rbpj is not expressed in optic stalk/neural retina boundary area in wild type (Fig. S2A). No expression of Rbpj in optic stalk/neural retina boundary may support that Hes1 expression in this boundary area is Notch-independent. However, Rbpj expression is retained in some vitrial cells near optic nerve head in Rax-Cre; Rbpj-CKO retinas (Fig. S2B). What are these Rbpj+ cells? I would like to request the authors to confirm that Rbpj expression is completely absent in both neural retina and optic stalk in Rax-Cre; Rbpj-CKO mice. Otherwise, this conclusion is still not fully supported.
   • c) Page 9, line 16-18, "Foxg1 had spread into the nasal optic stalk": Is Foxg1 expanded nasal area really "OS" rather than expanded retina? I suggest the authors to confirm molecular markers Pax2 expression is overlapped with Foxg1. Otherwise, it is difficult to conclude that foxg1 is expanded into the optic stalk territory, because foxg1 is normally a marker of retina. Indeed, Fig. 3K shows pax2 expression is shifted into more inside towards the brain, suggesting that neural retina is expanded. Please explain the situation.
   • d) Page 10, line 4-5, In Rax-Cre; Hes1/3/5 cTKO eye, this tissue (RPE) extended into the optic stalk": This description seems to be incorrect. A part of Pax2 area, which is adjacent to the neural retina, contacts with RPE in wild type (Fig. 3AH), so most of RPE covers the neural retina even in Fig. 3DK.
   • e) Page 10, line 22-23, "For Chk10-Cre; Hes1/3/5 cTKO, there was a unique presence of ectopic Pax2 within the retinal territories": I wonder if this description is correct. I suspect that proliferative Pax2+ cells expand into regressing territory of Hes KO retinal cells, which undergo precocious neurogenesis and lose proliferative activity, in Chk10-Cre; HesTKO. In this case, it is possible that the Pax2/Pax6 interface may be maintained. Please show red and green channel panels for Fig. 3N to confirm that there is ectopic pax2 and pax6 double positive cells.
5. There seems to be some mismatch descriptions between image data and histogram (or text in the result section). Please respond my comments below.
   • a) Page 11, line 20-25. There seems to be inconsistency between result description and image data of Fig. 5A-G, and histogram Fig. 5O. Authors mentioned that a modest loss of pH3+ cell fraction in Chx10-Cre; Hes1/3/5 cTKO but not in Rax-Cre; Hes1/3/5 cTKO. However, Fig. 5D indicates severe reduction of pH3+ cell fraction in Rax-Cre; Hes1/3/5/ cTKO, which is similar to reduction of pH3+ cell fraction in Rex-Cre; Rbpj (Fig. 5B), but histogram data is different (Fig. 5O). Furthermore, pH3+ cell fraction is severely reduced in Chx10-Cre; ROSA(dn-Maml-GFP) (Fig. 5F) and modestly reduced in Chx10-Cre; Hes1/3/5 cTKO (Fig. 5G). However, pH3+ cell fraction seems to be normal in Chx10-Cre; Rbpj (Fig. 5E). These Chx10-Cre image data do not match the histogram of Fig. 5O. Please check their situation.
   • b) Fig. 5H-N, P: I wonder if the stage E13 is appropriate to evaluate cell death and survival because optic cup already becomes smaller in Rax-Cre; Rbpj, Hes1/3/5 cTKO, or ROSA(dn-MAML-GFP) than in wild-type control. I suggest the authors examine more earlier stage.
   • c) Page 12, line 19-20, "all other mutants (Chx10-Cre; Rbpj, and Chx10-Cre; ROSA(dn-MAML-GFP) were unaffected (Fig. 6EF, LM, ST)": It is likely that atoh7 expressing cells are mildly decreased and neuronal marker, Tubb3 and Rbpms-expressing cells are increased in Chx10-Cre; Rbpj, and Chx10-Cre; ROSA(dn-MAML-GFP). I requested the authors to evaluate the fraction of these markers in retinal area statistically in all the cases.
6. Some experiments are necessary to improve their design. Please respond my comments below.
   • a) Page 7, line 19 "Ectopic blood vessels protruded from the ONH (Fig. 1K, 1L)": It is difficult to see blood vessel structures in these panels (Fig. 1I-L). Please show some molecular marker of blood vessels to confirm how blood vessel is organized in Hes1/3/5 cTKO.
   • b) Fig. 5: Increase of pH3 fraction indicates several possibilities, for example (1) increased fraction of mitotic cells due to precocious neurogenesis, (2) increased fraction of mitotic cells due to activated cell proliferation of retinal progenitor cells, (3) increased cell-cycle arrest in M phase due to some stress response of progenitor cells. So, I suggest the authors to examine (1) BrdU percentage of retinal section area, (2) the percentage of pH3+ cells in PCNA+ retinal cells.
   • c) Fig. 5: It is better that cell death fraction will be evaluated by TUNEL and labeling with anti-activated caspase 3 antibody.
7. Panel configuration of Figures should be revised as below.
   • a) Please show red channel (Hes1) image in Fig1BC.
   • b) Fig. 1DH should be shown in neighbor. Fig. 1H should be assigned as Fig. 1E.
   • c) Fig. S2D, F, H, J: Please show GFP green channel as well. Otherwise, it is difficult to see non-overlapping expression in optic stalk area.
   • d) Fig. 9B: It is better to show Rax-Cre: Hes1/3/5 TKO rather than Rax-Cre: Hes1 cKO.
   • e) Fig. 9B: Lettering "Rbpj mutant" should be revised as "Rax-Cre: Rbpj KO".

Significance

The senior author of this manuscript, Dr. Nadean Brown, is an expert scientist who has investigate the role of Notch signaling pathway in vertebrate ocular tissue, including the neural retina and lens. In general, Notch signaling pathway consists of signaling stream from the interaction of Delta and Notch, Notch receptor activation by proteolytic cleavage, translocation of Notch intracellular domain (NICD) into nucleus, formation of transcription factor complex consisting of NICD/Rbpj/Maml, to the transcriptional activation of Notch target genes, Hes family transcription factors. Finally, Hes suppresses neurogenic program and maintain a pool of neural progenitor cells. Therefore, Notch is a key factor to regulate the balance between neurogenesis and progenitor proliferation. In this manuscript, the authors investigated retinal phenotypes in the knockout mice of different Notch signaling components, including Rbpj, Maml, and Hes. They found that functions of these three factors are not always equal in retinal cell differentiation; rather, they specifically regulate a particular step of retinal development. The authors propose the possibility that each of Notch signaling components may be modified by other signaling pathways and achieve some new roles beyond the conventional frame of classic Notch signaling pathway. In this point, this work has a potential to provide a new conceptual advance in the field of developmental and cell biology.

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

Hes genes are effectors of the Notch signaling pathway but can also act down-stream of other signaling cascades. In this manuscript the authors attempt to address the complexity of Hes effectors during optic cup development and retinal neurogenesis. To do so, they compared optic cup patterning and retinal neurogenesis in seven germline or conditional mutant mouse embryos generated with two spatio-temporally distinct Cre drivers. These lines allowed for the analysis of the consequences of perturbing the Notch ternary complex and multiple Hes genes alone or in combination. The authors show that the optic disc/nerve head is regulated by Notch independent Hes1 function. They also confirm that perturbation of Notch signaling interferes with cell proliferation enhancing the production of differentiated ganglion cells, whereas photoreceptor genesis requires both Rbpj and Hes1 with Notch dependent and independent mechanisms.

This is a rather complex study that dissects further the role of the Notch pathway and Hes proteins during eye development, a topic that has been addressed in many previous studies but perhaps not with the details that the authors have used here. In this respect, this study adds to current literature but will likely be of interest to retina aficionados. The manuscript reads well and the figures are of very good quality. However, many of the statements are based on qualitative rather than on quantitative analysis. This should be, at least in some cases, remediated, despite the effort that this may require given the number of mouse lines used in the study. Specific comments are listed below:

1. The title is somewhat misleading. The authors have explored mostly the role of Hes1, 3 and5. Although these are Notch effectors, there is already evidence that they participate in other pathways This is confirmed by the data present here. I would suggest to eliminate Notch from the title and use instead "Hes" to better reflect the findings. Furthermore, it is unclear why there is a reference to "mutations" or what are the Notch branchpoints to which the authors refer at the beginning of the discussion.
2. "Although the Pax6-Pax2 boundary is intact in Rax-Cre;RbpjCKO/CKO eyes, ONH shape was attenuated compared to controls (Fig 3I)". This statement is arguable as the difference seems subtle. Perhaps some kind of quantification would help.
3. Page 12 first paragraph. "....but all other genotypes were unaffected". This statement is unclear. All lines in which the Rax-cre has been used seem to have an increased number of apoptotic cells. This should be better explained
4. Page 12, end of second paragraph: "E13.5 Chx10-Cre;HesTKO eyes had a milder RGC phenotype (Figs 6G, 6N, 6U), but all other mutants were unaffected (Figs 6E, 6F, 6L, 6M, 6S, 6T). This statement is also rather subjective. The phenotype of Chx10-Cre;HesTKO is quite strong and the other mutants seem to have a phenotype. Some quantifications here will help.
5. Page 13, toward the bottom..."...but noted that Chx10-Cre RbpjCKO/CKO eyes were not different from controls (Figs 7E, 7AA)". Again, this statement is questionable as staining for both CRX and Rbpms seem reduced as compared to controls as quantifications in 7AA seems also to indicate (about half?). Did the authors calculate whether there is a statistical difference between controls and Chx10-Cre RbpjCKO/CKO ?
6. In Fig 7CC the authors should make the effort of including at least one additional sample, 2 biological replicates seem insufficient to draw a conclusion.

Significance

This is a rather complex study that dissects further the role of the Notch pathway and Hes proteins during eye development, a topic that has been addressed in many previous studies but perhaps not with the details that the authors have used here. In this respect, this study adds to current literature but will likely be of interest to retina aficionados. The manuscript reads well and the figures are of very good quality. However, many of the statements are based on qualitative rather than on quantitative analysis. This should be, at least in some cases, remediated, despite the effort that this may require given the number of mouse lines used in the 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

Summary: In this study, the authors employed an impressive set of mouse mutant or Cre lines to investigate the complexity of Notch signaling across different stages of retinal development. These comprehensive analyses led to two main findings: 1. Sustained hes1 in the OHS/OS is Notch-independent; 2. Rbpj and Hes1 exhibited opposing roles in cone photoreceptor development. Although the study is potentially interesting, the current manuscript needs the essential research background and quantification, a lack of which significantly reduced the clarity of the manuscript and the credibility of the major conclusions. Also, how the authors organized the results is quite confusing, making the manuscript very difficult to follow.

Major comments:

1. The authors needed to make the quantification for many analyses to strengthen the conclusions, such as Fig. 1F, 1G, and etc.
2. The authors reported many exciting results. However, further mechanistic insights are largely missing. They may focus on one of these exciting findings and give some mechanistic insights. For example, hes1 suppresses hes5 expression as the ONH boundary forms; hes1 expression in the ONH is Notch independent; differential influences of Rbpj and Hes1 on cone development. It is better for the authors to select one of these exciting findings and provide a deeper mechanistic study.
3. Some analyses lack an explanation of the rationale. For example, "To understand if the loss of multiple Hes genes is more catastrophic than Hes1 alone..."(PAGE 7). Please explain its significance.

Significance

In general, many results are quite interesting. However, the significance of these findings is largely hampered in the following aspects: 1. The authors were unable to provide the sufficient research contexts that are essential for understanding many results.2. Many conclusions were solely based on descriptive images but lacked statistical quantification, which significantly weakened many conclusions. 3. Many interesting findings are quite descriptive, and some mechanistic understandings of one of these exciting findings will be beneficial to improve the focus and significance of the study.

Current format of the manuscript fits more specialized audience.

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

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

Schiano and colleagues present data on two mouse knock-in models with a missense mutation in uromodulin (C171Y and R186S). A strength of the paper is that the mutations are found in patients with autosomal dominant tubulointerstitial kidney disease (ADTKD) but lead to divergent disease progression. The mouse models are characterized in detail examining changes in uromodulin processing, plasma and urine biochemistry and transcript levels by RNA-sequencing. These findings combined with studies in collecting duct lines provide evidence that the extent of uromodulin aggregate formation is related to the severity of the disease and mechanisms are provided to explain these findings including clearance pathway which might be targeted in the future. Overall, there is a large quantity of good data in the manuscript which moves our understanding of uromodulin mutations forward. However, there are some issues that need to be addressed as outlined below.

Major Comments

1. In the Introduction, the authors state that the current mouse models have only provided limited information warranting this new study. More information is required here to provide a stronger rationale. What are the specific weaknesses of the prior approaches and what precise questions remain unanswered and how is this hindering therapeutic development. Subsequently, how does this study fill these gaps in our knowledge? This narrative of highlighting the new aspects of this study should also run through the Abstract of the paper more prominently.
2. The authors have selected two missense mutations from the Belgo-Swiss ADTKD Registry to subsequently model in mice. Are these mutations also present at a high prevalence in other genetic studies of ADTKD? The authors indicate that the patients with a Arg185Ser mutation have a faster progression than Cys170Tyr. One caveat here is that in Supplementary Table 1-2, the patients with Arg185Ser are predominately male and those with Cys170Tyr predominately female. Therefore, is gender playing a role here with males more susceptible to renal disease. Taking this concept forward, if the generated mice are separated by gender are comparable results seen in pathology and renal function parameters than if the animals are grouped together as presented in the paper.
3. In Figure 1D, an examination of kidney biopsies is undertaken. Can the authors provide any quantification across multiple samples/sections/cells to strengthen this data? The authors measure CD3+ cells in their mouse models - any evidence of these cells in the human biopsies.
4. In Figure 2C, the quantification presented does not seem to fully reflect the pattern of the blot shown, for example, increase in total signal seen in homozygous mice versus heterozygous C171Y mice. As one of the focuses of the paper is the formation of uromodulin aggregates, perhaps there is a rationale for the core and HMW proteins to be quantified separately, rather than the ratio between them.
5. The authors use electron microscopy (Figure 2F) to conclude that expansion and hyperplasia of the ER occurs in their mutant mice. A representative snapshot is shown, but can quantification be provided to strengthen this data.
6. A detailed assessment of plasma and urine biochemistry has been made. As highlighted above, separating this data by sex could be helpful. It is stated that the C171Y mice have a progressive increase in BUN at 4 months, but this statement requires clarification. Are the authors referring to a progressive change over time or with respect to gene dosage? An additional measurement of creatinine clearance might also be useful here. Are there any changes in glomerular function? Significant changes are also found in the urine of C171 heterozygous mice (in sodium and creatinine) but not in the homozygous animals. Any explanation for these findings which are not mentioned in the text? Some of the data is not reported corrected, for example it is stated that uric acid excretion is reduced at 1 month, but this has not been measured then. The conclusion that there are strong gene-dosage effects in both models seems strong. The reviewer agrees this holds for BUN but is not so clear cut for other parameters such as diuresis and osmolarity in C171Y mice. This should be refined.
7. An interesting analysis is presented on the effect of partial and total denaturation treatments of uromodulin. The reproducibility of these experiments is unclear. Please clarify. Do the authors have any information on how the protein structure of uromodulin might change due to these mutations, for example by structural modelling?
8. Next, the authors delete a wild-type allele in the R186S mice and examine the severity of disease. In Figure 4D and E it would be more informative to also present the specific changes in HMW and core proteins separately. Is there really a pronounced reduction in premature uromodulin in Figure 4E? Why have the authors focused on CD3+ cells as a marker of inflammation, how about other cell types such as macrophages? The rationale needs to be provided here. Are there changes in fibrosis by histology? Importantly, there appears to be no changes in clinical parameters when the wild-type allele is deleted, so is the main conclusion of this part that the deletion of the wild-type allele has no effect on disease severity, despite some of the gene changes observed.
9. In Figure 5, the relationship between the amount of uromodulin aggregates and the UPR pathway, fibrosis and inflammation is examined. As highlighted above, the methodology to determine the number of uromodulin aggregates needs to be considered. It is unclear in Figure 5C how this parameter has been generated. Can the authors present the data in this panel as individual mice of all six groups rather than the grouped analysis currently done. This would distinguish if the individual mice with greatest uromodulin aggregates also had the most fibrosis and inflammation and strengthen the presentation of this data.
10. In your RNA-sequencing data, please clarify if the mice were of the same sex. Interesting changes are found, but the final conclusion is that the transcription signals recapitulate severe ADTMD. This seems an overinterpretation and to strengthen this section the authors could go back to their biopsy samples and examine some of the expression patterns of the novel genes they have identified. Similarly, can any of the novel transcripts identified in the RNA-seq be examined (and/or) altered in the cell lines they have generated with the same mutations in uromodulin.
11. Using their cells the authors show the autophagy may be involved in the clearance of uromodulin in R185S mutants. However, this pathway is not explored in vivo, an assessment of autophagy in these mice would strengthen this connection.

Minor

1. The authors should present full Western blots in their Supplementary data
2. Figure 2C (and others). Please clarify and label clearly the blots from 1 month and 4-month-old mice.

Significance

Schiano and colleagues present data on two mouse knock-in models with a missense mutation in uromodulin (C171Y and R186S). A strength of the paper is that the mutations are found in patients with autosomal dominant tubulointerstitial kidney disease (ADTKD) but lead to divergent disease progression. The mouse models are characterized in detail examining changes in uromodulin processing, plasma and urine biochemistry and transcript levels by RNA-sequencing. These findings combined with studies in collecting duct lines provide evidence that the extent of uromodulin aggregate formation is related to the severity of the disease and mechanisms are provided to explain these findings including clearance pathway which might be targeted in the future. Overall, there is a large quantity of good data in the manuscript which moves our understanding of uromodulin mutations forward. However, there are some issues that need to be addressed; in particular the authors should (i) precisely outline the novelty of their study compared with the prior literature; (ii) clarify the reproducibility of their experiments; (iii) refine areas of overinterpretation in the manuscript; (iv) consider the potential role of gender in their findings and (v) complete the circle in some of their findings, for example examining the novel genes identified in their RNA-sequencing in their human biopsy samples and examining autophagy in their mouse models. These changes will considerably strengthen their article.

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

UAKD, a subtype of ADTKD, is extensively studied, although it is an rare inherit kidney disease. Using a knock-in strategy, the authors raised a novel concept that the differences in allelic and gene dosage of Umod mutation triggered distinct protein catabolic pathways, yielding distinct phenotypes and prognosis. The functional mechanisms include that UmodR186S mutation caused insoluble uromodulin aggregates resulting in activation of autophagy, and UmodC171Y mutation led to uromodulin misfolding and touched off ubiquitin-dependent ERAD pathway. Accordingly, the authors tested whether enhancing autophagy attenuates the accumulation of UmodR186S protein in cell cultures. Based on these observations, the authors suggested a strategy to improve clearance of mutant uromodulin. This study was carried out by a team with strong reputation in this area. However, the story appears to be incomplete and in vivo testing of their therapeutic strategy is needed to improve this research.

Specific comments

1. Figure 1D: Images at low magnification do not show DAPI, therefore there is no information on the total number of cells in the selected field. Nephron loss (represented by glomeruli) did not appear to differ between UMOD p C170Y and UMOD p R185S, which is inconsistent with the overall conclusions. In addition, PAS staining should be added in Figure 1D.
2. Figure 2E: in image of C171Y/+, this is no corresponding tubules which is represented by the insert. Figure 2F lower panel, the bars in EM fields are same, indicating a hypertrophy of nuclei in R186S? Figure 2G: how about serum creatinine in these mice? In addition, signs of catabolism (e.g., loss of body weight) are associated with these KI mice?
3. Figure 3C: what is rationale of using two high speed centrifuges. Please state briefly in method.
4. Figure 4: histologic assessment of progression is missing here, please add images of PAS, Masson staining at low magnification
5. Figure 5: Can the authors provide low magnification images (40X) for each condition? A histological evaluation of kidney damage is critical to support the conclusion.
6. Figure 6: Why are no ubiquitin-related catabolic processes or pathways enriched in C171Y? The authors should perform GSEA analysis to determine whether defined gene sets have significant differences between C171Y and R186S.
7. Following the experiments in Figures 7 and 8, the authors should assess whether administration of autophagy agonists could improve kidney injury and function in R186S mice.

Significance

Although ADTKD is an rare inherit kidney disease, the authors provide new insight into its pathogenesis. As nephrologist, I agreed with the observations and conclusions provided by the study. However, sufficient histological assessment and in vivo validation of the proposed therapeutic strategy would significantly improve 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

Uromodulin (Tamm-Horsfall protein) is the most abundant protein excreted in human urine.<br /> It plays role in protection against urinary tract infections and renal stones. Mutations in UMOD gene encoding uromodulin cause Autosomal Dominant Tubulointerstitial Disease (ADTKD) that slowly progresses to chronic kidney disease.

In this manuscript, Schiano et al. isolate 12 missense UMOD mutations, which they classify into two groups by age occurrence. They then proceed to study two of these mutations: one from the earlier-onset - Arg185Ser - and the second from the later-onset - Cys170Tyr.

The authors generate UmodC171Y and UmodR186S knock-in mice with distinct dynamic pathways impacting on ADTKD progression. These mutations are equivalent with UMOD mutations (C170Y and R185S) in patients. UmodC171Y and UmodR186S knock-in mice show impaired uromodulin biogenesis, with strong allelic and gene-dosage effects. The trafficking problem of ADTKD-UMOD mutants, involving ER retention, ER stress, and activation of the UPR is recapitulated in mIMCD-3 cells, where the R185S mutant reveals more aggregates that are triggering PERK and IRE1 pathways and ER stress responses.

The manuscript is well written, experiments are in general well described and performed, results offer important insights on cellular events eventually leading to organ damage in ADTKD resulting from missense mutation in the UMOD gene.<br /> The part of the work investigating the degradation mode of two different UMOD mutants, one relying on proteasomal and one relying on lysosomal clearance, is the most interesting for a general audience. Unfortunately, this last part of the work is too preliminary to be accepted as it is.

Comments/Suggestions:

• Selection of the UMOD variants, page 5: "R185S and C170Y are the most prevalent mutants in the clusters" please document/add reference.
• Fig. 1D: please show the position of the insets in the UMOD and BiP panels. Please separate the IF panels from the Picrosirius red panels (these are not the same samples that are shown),<br /> Formally, the BiP panels in Fig. 1D reveal that there is more BiP in cells expressing R185S. That this correlates with UPR induction (as confirmed in Fig. x) should be written at the end of page 5 to make this issue clear for non-experts.<br /> In Fig. 1D, the signal of BiP is not visible in WT and C170Y tissue/cells, which is odd because BiP is abundant protein. Moreover, the differences in BiP levels quantified in WB (semi-quantitative analyses) are not that dramatic in the mouse model (SFig. 3). Which panel in SFig. 3 (mouse) should be representative of the IF shown in Fig. 1D (patients)?<br /> Fig. 1D: Magnification of these images is not sufficient to conclude that R185S accumulates in the ER, and that WT and C170Y are at the apical cell's membrane as written (page 5). Authors should refer to Suppl Fig 1C, where individual cells are visible.<br /> Authors should briefly explain at the end of page 5 how the P. red staining in Fig. 1D informs on fibrosis.
• In the analyses of misfolded UMOD mutants (e.g., Fig. 2, 3, 4, ...) one would expect a test showing that BiP associates with R185S>C170Y>WT.
• Fig. 2F: in R186S there is a dramatic enlargement (at least 2x) of nuclei. Can the authors comment on that?
• Fig. 7E: Shouldn't one expects apical signal for C170Y?

• Fig. 7F: Why there is apical signal for R185S (and not for C170Y)?

• The part covering the degradation of the two UMOD variants would be of great interest for a wide audience of cell biologists. However, these data are too preliminary and, in this form, inconclusive.<br /> Few examples: MG132 is a non-specific inhibitor of the proteasome, which may enhance endogenous and trans-gene expression (check in Pubmed "mg132 promoter" for relevant literature). Thus, an increase in the intracellular level of C170Y on MG132 treatment does not necessarily indicate inhibition of the protein's proteasomal turnover. It could also, at least in part, be caused by an increased synthesis of UMOD. The authors should show that MG132 does not increase synthesis of mutant UMOD (or use the more selective proteasome inhibitor PS-341 in their experiments); similarly, the data on R185S do not prove that this protein is client of autophagy. They rather show that autophagy removes the protein when cells are under nutrient restriction (note that starvation activates bulk autophagy, the non-selective lysosomal clearance of cellular components). To show that misfolded R185S is removed from cells by misfolded protein-induced ER-phagy (i.e., ER-to-lysosome-associated degradation), the authors should monitor in WB the accumulation of R185S in the presence of BafA1 and/or in IF the accumulation of R185S within lysosomes in the presence of BafA1.

Minor comments

• Figure 1B: dotted lines should be defined in the legend.
• Figure 1C: "phenotypes are denoted as indicated". The color-code used for the phenotype is unclear to me. For example, what is the phenotype of the V.2 (grey square)?
• The meaning of "Unlike in UMOD R185S cells, higher SQSTM1 puncta colocalizing with uromodulin were initially present in C170Y mutant cells and further accumulated in MG132-treated cells (Supplementary Figures 10A, B). These data suggest that mutant cells respond differently to UPS inhibition, with C170Y mutant uromodulin being mainly targeted to this pathway." (page 14) and the interpretation of the results shown in 10A and 10B is unclear to me.
• Page 7: "The UmodC171Y mice showed a progressive increase in BUN at 4 months" please define BUN.
• Please, provide a complete list of primary antibodies used for immunoblotting, immunohistochemistry, and immunofluorescence staining.

Significance

The manuscript is well written, experiments are in general well described and performed, results offer important insights on cellular events eventually leading to organ damage in ADTKD resulting from missense mutation in the UMOD gene.<br /> The part of the work investigating the degradation mode of two different UMOD mutants, one relying on proteasomal and one relying on lysosomal clearance, is the most interesting for a general audience. Unfortunately, this last part of the work is too preliminary to be accepted as it is.

My expertise: protein quality control, ER-phagy

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]

This study represents a detailed analysis of the mechanistic bases of Atg15 function in autophagy, which relates back to our initial studies of autophagy published by our group 30 years ago in JCB (10.1083/jcb.119.2.301), where we reported autophagy by disrupting vacuolar protease function. We report that Atg15 is the sole vacuolar lipase in yeast, and that it exhibits broad activity on a range of lipids. Following submission of our study to Review Commons, we have received favorable feedback from all three reviewers. We plan to perform the revisions below within one month, following which we will submit a full revision to JCB with complete point-by-point responses to the reviewers. We are confident that this study will be of interest to the broad readership of JCB and trust that you will find it worthy of further consideration for publication.

2. Description of the planned revisions

Reviewers 1 and 3 suggested that we should confirm whether Atg15 is indeed the sole vacuolar lipase using lipids other than NBD-PE. While we have already shown that the kinase-dead S332A variant is non-function in vacuolar lysates, we will further address this comment by determining whether vacuolar lysates isolated from _atg15_Δ cells are able to process other lipid species. We will also collect replicates and quantify data for all figures to address comments made by the reviewers.

We also received a comment from reviewer 2 asking us to determine the function and expression level of vector-borne ATG15 and ATG15-Flag expressed in the _atg15_Δ background strain. We will provide these data, along with a comparison with results from WT cells, in our revision.

Reviewer 1 indicated that we need to address the localization of Atg15 in more detail. We plan to better explain the results of our initial analyses, as well as collecting more detailed data using super-resolution microscopy, and these data will be analyzed and discussed in further detail.

Regarding the text, we will update the results, discussion and methods to make these easier to follow, as pointed out by reviewer 1.

A complete point-by-point response to reviewer comments will be provided in the full revision.

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

No revisions yet carried out.

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

Reviewer 3 suggested that when considering Atg15 lipase activity we provide information about the lipid makeup of autophagic body membranes. While we agree that this is an interesting suggestion, our pilot experiments have indicated to us that this analysis is complex and will generate a very large amount of additional data and technical details that would need to be supplemented, clearly exceeding the scope of this study. Further, while we are currently performing these analyses, it will take significant additional time to bring these experiments to a conclusion in line with the reviewer’s comment.

On a related note, reviewer 3 suggested that we provide more detailed analyses of the degradation products arising from Atg15 lipase activity to provide some context into our finding that Atg15 acts on purified autophagic body membranes. We have collected initial lipidomic data for vacuolar extracts following the induction of autophagy (see attached figure), and have confirmed that we detect lysophospholipids in a manner that quantitatively depends on the amount of Atg15. However, as we feel that these data require further careful and time-consuming lipidomic analyses, as well as a nuanced discussion of results arising, we plan to publish these data in a separate, detailed paper and not in the present study.

With regard to reviewer 1’s comment about the vulnerability of Atg15 to the presence of detergent, we agree that this is an important point, but we can not eliminate the possibility that a very small amount of Atg15 exhibits lipase activity that is detected by this assay. The key message of our study is that Atg15 is activated by proteases in the vacuole to function as a broad-activity lipase; we feel that a detailed investigation of the active fragment of Atg15 is secondary to this finding and would unfortunately be very difficult to determine using currently available techniques, especially when considering the expression of Atg15. While it would be very nice to have these data, we therefore cannot provide these data in the full revision.

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 manuscript by Kagohashi and colleagues provide evidence of the enzymatic activity of ATG15 as a phospholipase B with a broad substrate specificity and dissect the mechanisms by which this protein is activated in the vacuole to promote the hydrolysis of autophagic bodies. The manuscript is very clear and well written, the scientific and experimental concepts are clearly presented and easy to follow. This is excellent biochemistry with very well-thought concepts and well-designed experiments. I am convinced by the authors' results and conclusions which, for the most part, are justified by appropriate experiments and conclusive data.

Main points:

• The authors claim that they show that ATG15 is the sole vacuolar phospholipase (see abstract). This conclusion is based on the in vitro analysis of the hydrolysis of NBD-PE (a fluorescent, thus not physiological, phospholipid). Fig. 1 indeed show the absence of NBB-PE hydrolysis in cells lacking ATG15 or expressing a catalytic dead version of the protein, which supports the authors' conclusion. But (1) this does not reflect the hydrolysis of physiological phospholipids in the vacuole, (2) this is only based on the analysis of one phospholipid (PE), (3) this is not consistent with results presented in Fig.5 C-F, in which, in the absence of ATG15 or when the catalytic dead version of the protein is expressed, there is clearly still some hydrolysis of PC, PG and PI. Concerning Fig. 4, the authors state that 'the commercially available NBD-PC and NBD-PG had been somewhat decomposed' and it is unclear to me if the lanes marked by an oblique line correspond to the lysate of atg15D cells or to NBD-PL alone. If this corresponds to the NBD-PL alone, I suggest that the authors perform the experiment presented in Fig.1 (at least Fig. 1C), with additional NBD-PL to actually test for residual phospholipases activity in the absence of ATG15.

I also suggest that the authors perform lipid analyses of purified vacuoles including engulfed organelles (autophagic bodies, MVBs etc.) to detect changes when put in contact with vacuolar lysates from WT, ATG15D cells and ATG15S332A, which will be more physiological that the use of NBD-PL, and could therefore support their conclusion.<br /> - The use of NBD-PL is very powerful and support the authors conclusions, but the paper lacks from physiological results as stated above. For instance, concerning the efficiency and substrate specificity of ATG15: what is the actual lipid composition of autophagic bodies ? How efficient is ATG15 in regard to the lipids that mainly compose the membrane of autophagic bodies ? Can the authors quantitatively compare the activity of ATG15 from one phospholipid to the other ? Here the experiments are performed on lipids in solution, what about hydrolysis activity on lipids in membranes ? As mentioned in my comment above, I suggest that the authors perform tests with purified autophagy bodies, or, at least, on reconstituted vesicles with a composition similar to what is found in autophagic bodies to assess the activity of ATG15 in physiological conditions.<br /> - The results and data presented by the authors are clear and seem unequivocal for the main parts but none of the results are quantified and there is no statistical analyses provided. How many times were the experiments repeated and how consistent are the results ? The authors must provide quantitative information and stats for all the figures.

Significance

Although it has been long known that ATG15 is required for the degradation of autophagic bodies, how this protein which transits to the vacuole through the MVB pathway can be activated in the vacuole to specifically target autophagic bodies and/or its cargo, remained completely unknown. The results presented here, pending their confirmation with additional experiments, thus fill an important gap in knowledge to understand the last crucial steps of the autophagy pathway which had remained largely elusive across organisms. Autophagy is critical for the physiology and development of all eucaryotes with major implication in human diseases. This manuscript will thus be of interest not only for the autophagy community but also for a broader general scientific community with potential applications in medical sciences. The results presented here also provide crucial elements to understand lipid hydrolysis in the vacuole and how this is finely regulated to ensure proper disruption of autophagic bodies, and thus, to the support the finality of autophagy degradation, while maintaining the integrity of the vacuolar membrane. In that context, this paper will influence all cell biologists by providing knowledge of the function, activities and homeostasis of the vacuole. This work raises the question of how ATG15 is specifically addressed to the membrane of autophagic bodies in the vacuole which will be the subject for future exciting research.

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

Atg15, a membrane-bound phospholipase targeted to the vacuole via the MVB pathway, has been studied in recent years. It is synthesized as an inactive proenzyme activated by vacuolar enzymes and is responsible for the degradation of the autophagic bodies' membranes. While its physiological function is well characterized, the biochemical details of its activation and overall activity remain largely unknown.

In the present study, Kagohashi et al conducted a comprehensive investigation to understand the mechanisms involved in the disruption of autophagic bodies (ABs) membranes. They focused on the activities of Atg15 and Pep4/Prb1 and employed primarily in vitro methods to elucidate the functional mechanism of Atg15. Purified Atg15 and ABs were used in the experiments. According to the proposed model, Pep4/Prb1 processes and activates Atg15 during its localization to the AB membrane, and this activation is necessary for Atg15's lipase activity. To support this model, the authors performed in vitro assays using purified proteins, vacuole and ABs purification, genetics, mutations, lipase activity assays, and morphological examination of autophagic bodies using a super-resolution fluorescence microscope. Their findings demonstrated that the activity of Pep4/Prb1 is required for Atg15's lipase activity, and Atg15 functions as a vacuolar lipase. The importance of the lipase motif for Atg15's activity was confirmed through the use of hydrolase mutants and purified Atg15 from both wild-type and mutant samples. Overall, the manuscript provides a comprehensive and solid analysis of Atg15 that will most likely interest the cell biology community. The experiments are well-controlled, and the conclusions are based on solid experimental data.

There are only a couple of minor issues that deserve the authors' attention:

Figure 1c shows a lower level of NBD-LPE in cells expressing ATG15 from a plasmid compared to the wild type, indicating that the plasmid did not fully restore the lipase activity. Additionally, the exact details for Atg15 expression should be explicitly described.

To ensure transparency and reproducibility, the authors should provide information about the specific expression vector(s) used for the plasmids in the study.

Significance

As indicated above, the study provides elaborate and solid characterization of an important enzyme that has been previously mainly functionally characterized.

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:

The rupture of single membrane-bound autophagic bodies is essential to release and catabolize contents of autophagosomes deposited in the vacuole. The phospholipase Atg15 has been thought to play an important role in this process. This study establishes methods to analyze phospholipase activity in isolated Saccharomyces cerevisiae vacuoles and elucidates the mechanisms that activate Atg15. Using an elegant cell-free assay the authors demonstrate that vacuolar extracts can cleave phosphatidyl ethanolamine in an Atg15 and Pep4/Prb-dependent manner. Atg15 is cleaved in the presence of Pep4/Prb, likely causing the release of Atg15 cytosolic domain in the vacuole. An Atg15 construct lacking the transmembrane anchor retains its lipase activity and when artificially targeted to vacuole using CPY tag localizes to autophagic bodies. The authors also establish the minimum construct of Atg15 that is sufficient to execute lipase function. The authors then isolate Atg15 from vacuolar extracts using a FLAG tag-based pulldown and show that the FLAG eluate is sufficient to cleave a range of phospholipids. Finally, using a protease-protection assay the authors show that Atg15 isolated using FLAG resin can cause disruption of isolated autophagic bodies.

Major comments:

1. Throughout the manuscript, TLC data and Ape1 maturation data are not quantified. The authors should include data on replicates and quantitation for all TLC and Ape1 processing data.
2. The conclusion that Atg15 is the sole source of phospholipase activity is based on cleavage of NBD-PE alone. It is not clear why specifically PE was chosen to test lipase activity of Atg15. It is possible that Atg15 has a higher preference for PE as has been shown previously (Ramya and Rajsekaran 2016). Have the authors tested to see if other phospholipids can be cleaved by vacuolar lysates derived from Atg15 knockout cells? This should be investigated further before concluding that Atg15 is the sole source of all lipase activity in vacuolar extracts.
3. Atg15 overexpressed and purified from Saccharomyces cerevisiae is shown to be sufficient to catalyze the cleavage of PE (among other phospholipids). How do the authors reconcile this finding with their observations on the requirement of Pep4 and Prb? This information should be included in the discussion.
4. Regarding Figure 3 and movie EV3, especially the lower panel, the overlap of cherry-Atg8 (autophagic bodies) and CPY(1-50)-Atg15(DN35)-mNG is not very clear. There appear to be several CPY(1-50)-Atg15(DN35)-mNG rings that do not surround Atg8.
5. a. Are these images from a single stack or represent the entire volume of the cell? This result could be better represented as a line profile and through a correlation analysis.
6. b. The finding that CPY(1-50)-Atg15(DN35) binds autophagic bodies is interesting, but it should be demonstrated with native/wild type protein. This can be achieved by expressing lipase deficient Atg15-mNG in rapamycin-treated cells, which should have intact accumulated autophagic bodies.
7. c. Atg15-mNG also localizes to a ring-like structure outside the vacuole. The authors should comment on the potential impact of this finding.
8. The rationale for using detergent solubilized and FLAG-eluted Atg15 to test lipase activity with other phospholipids (LPC, PI, PC and PG) is not clear. Detergent solubilized and FLAG-eluted Atg15 is degraded (Figure4C). Does this mean that degraded forms of Atg15 exhibit broader lipase activity? The authors should test for breakdown of other phospholipids with whole vacuolar extracts or vacuolar pellet fraction that has intact membrane bound Atg15. If only degraded forms of Atg15 show broad phospholipid lipase activity, then this will be informative about regulation of Atg15 function.
9. Figure6B: ProteinaseK is a broad-spectrum protease. It is unclear why it would specifically cleave GST-GFP and prApe1 to produce single bands (and not a smear) corresponding to free-GFP and dApe1. This result can be explained better.

Minor comments:

1. Fig1E legend states, "Each vacuolar lysates were added at a volume ratio of 1:5:25". It's not clear what this means or what this ratio is for. In general figure legends need to be more descriptive on how the experiment was performed.
2. It's not clear what processed Atg15 (pcrAtg15) refers to in Figure4C. Is it indicating the smear around the 75kDa band? This should be explained clearly in the figure legend and the results section.

Significance

The phospholipase Atg15 is known to play a crucial role in the degradation of autophagic bodies within the vacuole. However, the regulatory mechanisms that prevent detrimental lipase activity of Atg15 have remained unclear. This study shows that proteolytic processing and membrane binding could activate Atg15, thereby providing important insights into the mechanism of Atg15 regulation.<br /> Using isolated autophagic bodies and vacuolar extract, the results here show direct disruption of autophagic bodies by Atg15. The cell-free assay to assess lipase activity can be further utilized to analyze vacuolar function. These finding will be of interest to a audience interested in various forms of autophagy and vacuolar degradation.

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]

Reply to general assessment of referee #2:

1. General assessments: The current study adds some to these observations…some of these observations are incremental…biological significance is limited. While this reviewer does not suggest additional experimentation, this manuscript would be suitable as a resource paper.

Reply: It appears we were not clear enough in explaining the novel aspects of our study.

The starting points are two published studies from our lab demonstrating a global increase of ISGF3 association with ISG promoters in IFNγ-treated cells and a remarkable similarity of IFN-γ and type I IFN-induced early transcriptome changes. These findings challenge the notion in the field (as mentioned by the referee) that IFNγ specificity is produced by the predominant deployment of STAT1 homodimers. We thus tested the hypothesis that the specificity of the IFNγ-induced transcriptome is generated over time, rather than during the early response, and relies on secondary responses to transcription factors such as IRF1. In contrast, IRF1 plays no or only a small role in the type I IFN response that utilises ISGF3 and/or unknown secondary factors in the delayed response. We tested this hypothesis with PRO-seq technology to rule out confounding effects of mRNA processing over a 48h period. The data are clear in showing that many genes associated with the antibacterial or anti parasite profile of activated macrophages are indeed much more abundant in late-stage rather than briefly IFNγ-treated macrophages and these delayed changes are to a large extent dependent on IRF1. Our findings are based on the best available technologies, a combination of nascent transcript analysis with genetics and protein interaction studies. In addition, our findings rule out alternative models of sustained or secondary ISG transcription, such as the employment of alternative ISGF3 complexes (such as STAT2-IRF9) or of ISGF3 complexes formed with unphosphorylated STAT1 and STAT2. We provide evidence for higher order waves of transcription caused by unknow transcription factors that are produced by transcriptional activation of ISGF3 or IRF1 target genes and identify candidates among the AP1 and Ets transcription factor families. We agree that some of the data are confirmatory rather than novel (i.e. some of the genes we describe were known from previous literature to be IRF1 targets), but it is the systems approach of our study, and particularly the delineation of conditions under which the largely neglected delayed response diverts the IFNβ and IFNγ-induced transcriptomes, that generates a comprehensive and conclusive view of IFNγ acting predominantly as a macrophage activating factor, and IFNβ being an essential antiviral cytokine. We do think this main outcome is immunologically meaningful and not incremental. For this reason, we would prefer to publish the paper as a relevant contribution to innate immunology rather than a resource. Emphasizing our point, a paper appeared in ‘Cell’ while our study was under review, showing that human IRF1 mutations cause mendelian susceptibility to mycobacterial disease (MSMD), a term coined by JL Casanova and colleagues for immunological defects that reduce the ability of macrophages to cope with intracellular bacteria (new ref. 65). This important study emphasizes the main conclusions of our study about the relevance of IRF1 for macrophage activation. We discuss this paper on p. 14 lines 9-14.

Revision: We tried to better explain the scientific motivation for this study and the significance of the results (p. 4, lines, lines 12-25).

Revision plan: n. a.

2. Description of the planned revisions

Referee #3; major comment 1:

In Fig. 1d is difficult to interpret and misleading for many reasons. First, the cluster numbering is disconnected from the cluster order; why not numbering them based on the hierarchical clustering and writing the cluster number besides the cluster itself? Second, having a 2-color gradient is misleading; negative values shouldn't be in the same color tone than the positive values. Third, the authors did not provide adequate rationale behind using only the top 1,000 most expressed gene? Why not using all the differentially expressed genes in at least one of the condition to provide a comprehensive analysis? Could this potentially lead to bias in the data, and is there any information lost by not using the - lower - expressed genes fraction? Fourth, it is not clear what the color scale is representing and how the data was transformed. Was a mean centering of the expression values of the log2FC applied to the RNA-seq data to facilitate clustering? Mean centering and z-scoring is a common technique used to adjust expression data, but it can potentially exaggerate differences between samples. More information about the data and analysis should be provided, as it is difficult to determine whether this was a valid approach or not.

Reply:

• To create the heatmap, we used the pheatmap package from R and the cutree_rows option to separate 11 clusters with strikingly different patterns of gene expression based on visual exploration. The numbering was autogenerated by the program.
• The data is now shown in red-blue.
• We restricted our list to only 1000 genes from each comparison as we aimed to analyze the prominent patterns of gene expression across timepoints. Considering all differentially expressed genes based on a padj value would also include genes expressed at very low levels as evident from the low baseMean values obtained from DESeq2. Hence, we applied a selection of 1000 genes which effectively represented the major patterns of gene expression across timepoints.
• Variance stabilized transformation was applied on read counts obtained from PRO-seq using the DESeq2 package. The transformed reads were z-score normalized and used for performing hierarchical clustering by the “Ward.D2” method using the pheatmap package in R. A total of 3126 genes were used for this analysis. 11 distinct clusters were defined using cutree_rows option. The color scale represents z-score normalized counts. The genes represented in the heatmap were selected based on the following criteria: each timepoint of interferon treatment was compared to the homeostatic condition (untreated sample) in wildtype BMDMs. The differentially expressed genes from each comparison were selected based on the filtering criteria: absolute log2FoldChange >=1 and adjusted p value <0.01 by Wald test. Following the differential analysis, the first 1000 differentially expressed genes in each treatment condition (ordered based on adjusted p values) were selected for both IFN types and combined and selected for creating a list which consisted of 3126 unique genes. The scale in the heatmap represents z-scores of variance-stabilized reads, calculated across all genotype and treatment conditions, separately for each IFN type.

Revision plan: We will label the clusters with the cluster number next to it in addition to the color codes.

Referee #3; major comment 3:

The large standard deviation bars in the claim that ChIP data confirmed the binding of ISGF3 components to the promoter of Mx2 cast doubt on the validity of the results and conclusions. The authors should consider additional experiments or complementary analyses to validate their findings. Or alternative, to adjust their claims accordingly.

Reply: To demonstrate sufficient quality of the data the ratio of Stat1/ Stat2 was calculated for early (1.5hrs) and late (48h) separately. The unpaired two-tailed t test comparing this ratio between 1.5 hrs and 48hs, shows that they are not significantly different. This indicates that all ISGF3 components are associated with ISG during both early and delayed responses, i. e., that STAT2/IRF9 complexes are unlikely to contribute to delayed ISG control. However, we agree with the referee that the standard deviations of the kinetic ChIP experiment are high and that it would be good to generate additional data.

Revision plan: We will perform additional ChIP experiments to improve the statistical power of the results in fig. S2c.

Referee #3, major comment 6:

The authors interpret their ATAC-seq and ChIP-seq results based on a 2kb window to the TSS of genes, not considering relatively close enhancers or longer range cis-regulatory interactions in their interpretation. For example, they mention on p.7 "Contrasting the strong binding of IRF9 and IRF1 to the Mx2 (cluster 2) and Gbp2 (cluster 9) promoters, respectively, we saw no evidence for direct binding to Lrp11 (cluster 3) and Ptgs2 (cluster 10)", but on Fig 3d they show only the proximal regions. No scale bars are shown either. Moreover, exploring the same published IRF1 ChIP-seq dataset, there is a clear IRF1 binding site at the promoter of Ptgs2, while the authors report none.

Reply:

• According to the literature (e. g. refs. 11, 27), most IFN-induced accessibility changes occur in the vicinity of the TSS of ISG. This is further strengthened by the data shown in this manuscript. In addition, most functionally validated GAS and ISRE sequences are in the DNA interval chosen for our analysis. While distal ISG enhancers have been reported (e. g. DOI: 10.26508/lsa.202201823), an analysis beyond the placement of most control regions increases the risk of wrong assignments between ISG and their regulatory elements, hence the causality between transcription factor binding and accessibility changes.
• We extended the regions for the analysis of the Lrp11 and Ptgs2 regulatory regions and found no evidence for the binding of ISGF3 or IRF1. We find no evidence for a clear peak in the Ptgs2 promoter. There is a peak called by the Macs2 algorithm, but visual inspection of the track (bigwig file) shows it consists of a minor increase in reads above background that does not suggest a bona fide IRF1 binding site (see below). This view is supported by our inability to find an IRF binding site in the vicinity of the peak.

IRF1 binding indicated by bigWig browser tracks and corresponding peakfiles detected at the locus. We identified the peakfile from Langlais et al., 2016 and identified peaks using MACS2, however using mm10 genome as the analysis in the original paper was done with mm9 genome. The peak identified here appears to be an artefact of the MACS2 program as there is no evident enrichment at the gene promoter region upon inspection of the bigWig files.

Revision plan: Scales will be added to the browser tracks as requested.

Referee #3, major comment 7:

Lack of statistical analysis on chromatin accessibility claims: The authors claim that ATAC-seq data in BMDMs stimulated with IFNβ or IFNγ for a short (1.5 hours) or long (48 hours) period reveals a striking similarity between transcription and the general trends of chromatin accessibility at regions up to 1000 bp upstream of the TSS (Fig. 2a), suggesting continuous chromatin remodeling during the transcriptional response. However, I would like to know if this conclusion is well-supported by the correlation between the chromatin accessibility from ATAC-seq data from only one sample and the PRO-seq data.

Reply: See revision plan.

Revision plan: We will analyze single experiments whether they support the conclusions derived from the z-score of the triplicate samples.

Referee #3, major comment 8:

The need for additional experiments to verify claims such as the dependence of Ifi44 on IRF1 for gaining ATAC signal, as stated in the claim, "Expression required IRF1 for both, but accessibility of the Ifi44 regulatory region depended upon IRF1 whereas that of Gbp2 acquired an open structure independently of IRF1 (Fig. 5c).

Reply: We think the lack of clarity might be related to the size of figures 5a and 5b and the density of the dots in some areas of the plot. We agree it is very difficult to assign our gene labels unambiguously to a single dot.

Fig. 5a combines ATACseq data in wt and IRF1 knockout cells with the expression data from the Pro-seq experiment, Fig. 5b is the same set-up, but IRF9-deficient macrophages are analyzed.

Blue dots show ATACseq signals induced by IFN treatment. Violet dots represent genes that require IRF1 (Fig. 5a) or IRF9 (Fig. 5b) for transcriptional induction. Yellow dots mark genes such as IFI44 requiring IRF1 (Fig. 5a) or IRF9 (Fig. 5b) for both expression and the accessibility change in the promoter region. Fig. 5c visualizes representative examples of genes whose accessibility is coupled to the transcription factor dependence of the transcriptional induction (IFI44), or not (Gbp2). Thus Fig. 5c must be interpreted based on the dot color code in fig. 5a and we admit this has been difficult with the figure in its present form.

Revision plan: We will improve the clarity of figs 5a and 5b in several ways:

• We will label the panels to better indicate the intersected data sets.
• We will increase the size of the panels and figure legends and make sure that the correspondence between gene names and dots are unambiguous.
• We will include trend lines of the Ifi44 and Gbp2 genes to visualize their induction and IRF1 dependence.

Referee #3, major comment 13 (see also section 3):

The authors have not adequately addressed the methodological limitations in their discussion, which extends beyond the aforementioned comments. It is suggested they include a comprehensive discussion of the claims made pertaining to the necessity of IRF1 for accessibility and the potential biases in the interactomes, along with their associated consequences.

Reply: The contribution of IRF1 to the accessibility of ISG promoters emerges from the data in figures 5a, whose clarity will be improved (see reply to point 8). We do not interpret the impact of IRF1 beyond the data, in fact we state a relatively minor effect of IRF1 in the control of promoter accessibility (p. 10, lines 20-22) and we have added a reference in agreement with an impact of IRF1 on basal expression of antiviral genes (ref. 39, as suggested by the referee).

We have added discussion on potential limitations of the TurboID approach (p. 11, lines 22-24 and p. 15, lines 3-11).

Revision plan: Improvement of fig 5a (see ref. #3, point 8).

Referee #3, minor comment 2

Fig 1e. The color scales on the GO enrichment graphs are misleading since they use the same blue-to-red gradient for adj p-values ranging from 10-25 to 10-49 and 0.008 to 0.016, which could be considered non significant.

Reply: We agree that this is confusing. It results from automated assignments of the color gradients by the software.

Revision plan: We will investigate possibilities to change color codes for different ranges of p values.

Referee #3, minor comment 4

The incomplete schema in Figure 1a, which only focuses on PRO-seq and does not include the ATAC-seq element.

Reply: We will add a new figure to visualize the set-up of the ATAC seq experiments and their intersection with the Pro-seq data.

Revision plan: We will add a new figure in accordance with the referee’s request.

Referee #3, minor comment 6

The clearer labeling of Figure 5a and 5b.

Reply: Please refer to our reply to major point 8.

Referee #3, minor comment 10

Fig S1b, S3b. The PRO-seq was generated in triplicates, hence these graphs should include the Log2FC for the individual data points.

Reply: The Log2FC from DESeq2 were calculated from the triplicates, the software does not compute Log2FC from individual replicates.

Revision plan: We mention the p-values for the Log2FC to show the degree of consistency (figure legends). We will provide a table with log2FC and corresponding padj values of the genes represented at each timepoint (table_showing_padj_values_and_log2fc).

Referee #3, minor comment 12

In the genomic snapshot shown, only bars or fading triangles are shown in place of the gene body. The authors should provide an accurate gene structure; i.e., exons and introns.

Reply: We will try to include the exon-intron structure wherever the size of the figure allows this.

Revision: n. a.

Revision plan: If figure size permits, we will add the exon-intron structure of the genes in browser tracks as requested.

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

Referee #1, major comment 1

Figure 2. Difficult to interpret data as it is presented. Consider quantifying figure 2C in order to make "changes in Pol II pausing were more pronounced during IFNb signaling" statement more apparent.

Reply: We presented the pausing data in two different graphic representations (figures 2c and S2) to make the understanding of the information content easier. In hindsight we may have generated more confusion than clarity.

Revision: We removed the original figure 2c and replaced it with original figure S2. This representation is quite intuitive as the graphs represent a direct quantitative logarithmic display whether and how much the relative amount of paused polymerase changes when comparing IFN-treated and untreated cells. The calculation of these ratios is now explained better in the legend to figure 2.

Referee #1, major comment 2

How are you distinguishing autocrine signaling in the BMDMs driven by IFN treatment from late transcripts (for example, at 48 hours are differential genes due to autocrine cytokine signaling or are they truly late transcripts)?

Reply: We do not exclude autocrine effects. In case of ISG, the most likely autocrine factor would be secreted interferon. According to our Proseq data, the differentially expressed genes do not include any interferon genes. That being said, it is possible that the transcription factors from the AP1 family we hypothesize as drivers of secondary or tertiary waves of transcription are activated by non-IFN cytokines secreted from IFN-treated cells (see also reply to comment 3).

Revision: We now mention that enhanced IFN production is not sustaining ISG responses (p.5 lines 18/20). We mention the possibility that secreted factors may drive secondary or tertiary waves of ISG transcription (p. 8, lines 21/23).

Referee #1, major comment 3

Figure 3D. Authors choose Gbp2 (as positive control for IFNg driven gene), but don't show that Gbp2 is a IFNb independent gene. Consider using IRF1 KO BMDMs in this data as well.

Reply: This is a misunderstanding. Gbp2 is not shown as an IFNγ-specific gene (it’s induction by both IFN types has been shown previously and emerges from our Pro-seq analysis, see also response to minor issue no. 2). It represents the cluster of genes that are sustained specifically after IFNγ treatment in an IRF1-dependent manner. The purpose of fig. 3D is to show that not all ISGF3/IRF9-dependent genes have promoter binding sites for ISGF3 and not all IRF1-dependent genes have binding sites for IRF1. This suggests indirect effects of both transcription factors in sustaining IFN-induced transcription (in line with the referee’s comment 1).

Previous figure S3e (now S2f) confirms binding of IRF1 to the GBP2 promoter by ChIP with kinetics correlating to its transcriptional effect. This experiment is normalized with an IgG control. IRF1 knockout cells did not produce a ChIP signal with IRF1 antibody, as expected (data not shown).

Revision: We better explain the rationale behind the experiments shown in figure 3D (text on p8, lines 12-16). In addition, we show the trend line of Gbp2 expression in WT vs IRF1KO as well as that of additional genes showing delayed/sustained responses in the new Figure S3.

Referee #1, minor comment 2

Define known IFNg and IFNb driven genes when they are introduced in figure 2 rather than in discussion.

Reply: Following the referee’s suggestion we provide the examples of IFNβ and IFNγ-controlled genes and the characteristics of their regulation in the context of our description of the results displayed by fig. 2 (p.6 lines 15-21). This includes Gbp2 (see major issue no. 3).

Revision: The text on p. 6 lines 15-21 has been modified in accordance with the request.

Referee #1, minor comment 4

Unclear whether IRF1 expression in figure 3A is from whole cell lysate or nuclear fraction.

Reply: We indicate in the figure legend that whole cell lysates were used.

Revision: We added a sentence with the relevant information in the legend of figure 3.

Referee #1, minor comment 5

Authors suggest IFNb treatment induces less IRF1 at later time points, however loading control also seems slightly lower than other considerations. Is it possible that IFNb treated cells are dying at later time points, given that type I IFN signaling can be pro-apoptotic.

Reply: The graph below the blot represents quantified IRF1 signals, normalized to the loading control. It shows that the differences are not generated by unequal loading of the blotted gel. We and others have shown that IFNβ may indeed enhance macrophage death, however only when the cells are simultaneously infected with an intracellular pathogen (e.g. new ref. 25). These studies also show that treatment with IFNβ alone over periods used in the present study does not affect macrophage viability.<br /> Revision: We added a sentence about the viability of IFN-treated macrophages (p. 4, lines 31-32).

Revision plan: n. a.

Referee #2, major comment 3

The sequencing and BioID data are not submitted to public databases.

Reply: An accession number has been added.

Revision: The accession number was added on p.29, line 25.

Referee #3, major comment 1 (see also revision plan, section 2):

Revision: The rationale for using the top 1.000 genes is explained (p.5, lines 7-9). The description of the pro-seq read count processing has been extended in accordance with our reply to the referee in the legend of figure 1d and in the methods section (p. 33, lines following line 10.)

Referee #3, major comment 2

Fig 2c. The authors claim that RNA Pol II pausing is a major factor in controlling the dynamics of ISG transcription. However, they did not provide sufficient explanation of the results, and in all fairness there is not much variation between the clusters to sustain the claim that this is a major factor in ISG transcriptional control.

Reply: We agree with the referee that we cannot posit RNA pol II pausing as a major factor for the differences of transcriptional control of ISG in individual clusters. We have made sure to remove any statements suggesting this possibility. We also try to better integrate our findings with RNA pol II pausing into the existing literature.

Revision: We added relevant literature on p. 6 lines 28-30 and p. 7, lines 4-6.

Referee #3, major comment 4

On p.5, the authors mention "Representative browser tracks from the Gbp2 and Slfn1 genes further validate this observation" but they are simply referring to genome browser snapshot, i.e., specific genomic examples, extracting from the same single dataset. Without using an independent dataset, this can not "further validate" the initial findings.

Reply: We agree the wording is incorrect.

Revision: We changed the paragraph describing this experiment (p. 6, lines 15-21).

Referee #3, major comment 5

IRF1 was successfully pulled down with STAT1 bait but not in the reciprocal experiment. The author should discuss this point as it is important for the conclusions. Could it potentially indicate issues with the technique used, and if this could introduce any bias into the results. The statement, "In contrast, interactors of the IRF1 bait did not include STAT1. This discrepancy could result from steric constraints of the tagged proteins due to the limitation of the 10nm distance reached by the biotin ligase," does not seem to be sufficient to explain this discrepancy.

Reply: STAT1 was present in the IRF1 pull-down and the interaction increased significantly after IFN treatment but after normalization to the NLS control it did not conform to our criterium of a 95% confidence interval for the FDR. To be consistent we did not include it in the list of IRF1 interactors. We have observed on several occasions that the significance of proximity is not reciprocal, even for well- documented physical interactions. A prime example for this is the interaction between STAT1 and IRF9 in IFN-treated cells which is recorded in the STAT1 pull-down, but not that with IRF9 (ref. 10). Apart from steric reasons the lack of reciprocity may result from different signal/noise ratios in pull downs with different baits.

Revision: We mention that IRF1 was a STAT1 interactor below the statistical cut-off (p. 11, lines 26-28) as well as the possibility of different signal/noise ratios in the IRF1 and STAT1 pull-downs on p.11, lines 22-24.

Referee #3, major comment 9

In the figure legends, there is missing information about the number of times experiments were replicated, suggesting that some were done a single time. Moreover, some graphs are missing statistical analysis, e.g., in Fig S3cS3e, S3f, the ChIP-qPCR experiments were done on biological triplicates, there is no mention of statistical test performed, it is not mentioned what the error bars represents (SD, SEM, etc.) and the variance is large, but the authors still interpret these results as significant enrichment of the transcription factors to the Mx2 promoter.

Reply: Where missing the relevant information has been added to figure legends. In brief, all experiments represent at least three biological replicates. The only exception is the western blot shown in figure S3a, (no S2a) which represents two independent replicates. Here, the clarity of the difference of IRF1 expression and the fact that the only purpose is to show that Raw264.7 macrophages behave like bone marrow-derived macrophages in fig. 3a justifies the omission of another replicate (please see also answer to point 3).

Revision: The relevant information has been added to figure legends where necessary (figs. 1, a, 3a, 6a-f, S1, S4, S5).

Referee #3, major comment 10

Another example are the RNA Pol II pausing index ratios, which show minor variations and not are supported by statistics to support a possible significance. Proper description, replication and statistical analyses of the results are critical.

Reply: We agree.

Revision: Statistics underlying the RNA Pol II pausing data are included in supplementary data 2.

Referee #3, major comment 11

The authors used CRISPR-Cas9 genome editing to generate knockout cell lines. However, they did not verify the knockouts at the protein level. Further experiments could confirm that the targeted proteins are not expressed in the knockout cell lines.

Reply: We included a western blot showing the lack of IRF1 and STAT1 expression in the respective cell lines.

Revision: New figure S6.

Referee #3, major comment 12

On p.9, it is mentioned "IRF1 affects chromatin structure ...". Here chromatin structure is related to minor changes in chromatin accessibility, this can not be qualified as changes in chromatin structure.

Reply: ‘structure’ has been changed in accordance with the request.

Revision: ‚structure‘ has been replaced with ‘accessibility’. (p. 10, lines 19 and 21).

Referee #3, major comment 13 (see also section 2, revision plan, major comment 8)

The authors have not adequately addressed the methodological limitations in their discussion, which extends beyond the aforementioned comments. It is suggested they include a comprehensive discussion of the claims made pertaining to the necessity of IRF1 for accessibility and the potential biases in the interactomes, along with their associated consequences.

Reply: The contribution of IRF1 to the accessibility of ISG promoters emerges from the data in figures 5a, whose clarity will be improved (see reply to point 8). We do not interpret the impact of IRF1 beyond the data, in fact we state a relatively minor effect of IRF1 in the control of promoter accessibility (p. 10, lines 20-22) and we have added a reference in agreement with an impact of IRF1 on basal expression of antiviral genes (ref. 39, as suggested by the referee).

We have added discussion on potential limitations of the TurboID approach (p. 11, lines 22-24 and p. 15, lines 3-11).

Revision: Change of the discussion section (p. 11, lines 22-24 and p. 15, lines 3-11).

Revision plan: Improvement of fig 5a (see ref. #3, point 8).

Referee #3, major comment 15

The work should be discussed in the context of the demonstrated physiopathological evidence of the IRF1 and IRF9 functions. IRF9 (Hernandez et al., JEM 2018) and more recently IRF1 (Rosain et al Cell, 2023) were identified as causing non overlapping phenotypes in human patients carrying loss-of-function mutations for these genes. The authors must interpret their results in this context.

Reply: We thank the referee for reminding us about the importance of these papers for our work.

Revision: The papers have been mentioned and discussed (p. 13 lines 19-28 and p.14, lines 9-14).

Referee #3, minor comment 3

The inconsistency in the title referring to IFNb as Type 1 but using IFNg instead of Type 2 nomenclature, perhaps consistency is best.

Reply: We agree about the importance of consistency but find ourselves in yet another quandary. While the use of ‘type I IFN’ is clearly indicated and widely used as a collective name for this group of cytokines, the use of ‘type II IFN’ for IFNγ is rare because it is the only member of this type. Hence, we decided for sticking with convention at the expense of a bit of consistency. We agree about the title, though, and have changed type I IFN to IFNβ.

Revision: We adapted the title in agreement with the referee’s comment.

Referee #3, minor comment 5

Figure 6d includes a color scale of -1 to +3, but it is unclear what these values represent and how they were calculated per interactor. The figure legend should be revised to clarify this information.

Reply: We agree. The relevant information has been added to the figure legend.

Revision: We added information (log2FC with regard to the NLS control) to the legend of fig. 6d.

Referee #3, minor comment 9

Fig 1e, S1c. Graphs having circles of varying sizes in function of a value are named "bubble plots" and not "dot plots".

Reply: Thank you for pointing this out, we corrected our mistake.

Revision: We changed dot plot to bubble plot in legend to figure S1c.

Referee #3, minor comment 11

Fig S3c legend. It is mentioned "Graph represents RT-qPCR of genomic Mx2". RT-qPCR usually stands for reverse transcription quantitative PCR, hence we suggest to change to "ChIP-qPCR" or qPCR. Confusingly, in the literature the term "RT-PCR" is used for real-time PCR and "qPCR" for quantitative PCR. Also, the authors should be specific about the "genomic" region targeted; the graphs mention "promoter", hence it would be appropriate to use the same designation in the legend.

Reply: We agree and thank the referee for correction of the terminology.

Revision: We changed RT-PCR to qPCR throughout the manuscript. Moreover, we specifically refer to ‘promoter region’ as the amplified DNA.

Referee #3, minor comment 12

Fig S3e. The y-axis names are missing.

Reply: Thanks for spotting this.

Revision: The y axis in the figure received its proper label.

Referee #3, minor comment 14

Raw cells are sometimes spelled as "Raw" and other times as "RAW". Please choose one for consistency.

Revision: This inconsistency has been corrected

Referee #3, minor comment 15

In p.10 l.20, the figure number is missing.

Revision: We corrected this mistake.

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

Referee #1, minor comment 1

Simplify figure 4B- consider focusing on most differentially expressed genes between clusters

Reply: The purpose of fig. 4B is to provide a visual overview of the kinetics of eRNA transcription in response to both IFN types and of the effects of IRF9 and IRF1 knockouts. This information needs to be given to demonstrate the similarities and differences between the control of eRNA and the corresponding ISG transcripts in the different regulatory clusters (as shown in figs. 1d and 2a).

Simplifying the figure would mean to separate it according to time point, IFN type treatment or knock-out effect. We think this would require to mentally reassemble the figure to understand the interrelationships between these parameters. To our opinion the visual display of the data interrelationship in fig. 4B facilitates the impropriation of the information content.

Revision: n. a. - we hope our reasoning has become sufficiently clear.

Revision plan: n. a.

Referee #1, minor comment 3

Clarify which cell types (IRF1 KO vs IRF9 KO) are used in figure 5 A/B.

Reply: The cell type (bone marrow-derived macrophages) is mentioned in the first sentence of the figure legend. Since all experiments except the Bio-ID experiment were performed with this cell type we decided not to label each figure.

Revision: n. a.

Revision plan: n. a.

Referee #2, major comment 2 and referee #3, major comment 14

Ref #2: Biological significance is limited as this study is largely descriptive and they do not test the hits obtained from BioID.

Ref #3: Although the TurboID experiments identify known STAT1 and IRF1 interactors, the proposed new interactors are numerous, and none are validated through independent co-IP experiments. Moreover, the results are very noisy, with little differences between untreated BMDMs (where IRF1 is barely expressed) and IFN-treated conditions.

Reply: The big advantage of BioID or TurboID is the ability to score proximity and very transient interactions. Validating BioID hits with technologies such as coIP is not particularly useful as the two technologies will obviously produce different interactomes. In fact, we show in this manuscript that IRF1 and STAT1 show proximity, but they do not form a stable complex under co-IP conditions. This leaves genetic approaches (LOF or GOF) as alternatives. However, apart from the workload (> 100 genes would have to be knocked out or their products overexpressed), most of our hits are expected to produce very broad effects in such experiments, hard to interpret regarding ISGF3 and IRF1 activities.

In view of this situation, we publish exclusively the high confidence nuclear interactors identified in our screen: biological replicates were performed in triplicate, a stringent internal control (TurboID-NLS) was used, and a stringent statistical cut-off for high-confidence interactors (95% FDR between groups) was applied. We further account for the experimental situation by limiting interpretation of the data to confirmed molecular events. For example, STAT1 dimers and the ISGF3 complex are required for histone acetylation in response to IFN, and ISGF3 is known to contribute to the exchange of the H2AZ histone variant (refs 11, 14, 71, 72). Our data show that IRF1 contributes to promoter accessibility changes and this is in line with its proximity to a remodelling complex. Thus, the BioID data indeed validate previous findings. However, in agreement with the referee’s comment, some of the data remain descriptive (such as the intriguing proximity of both STAT1 and IRF1 to nuclear products of ISG). To determine the importance of this molecular proximity is a major undertaking and beyond the scope of this study.

Revision: We added discussion to state the difficulty of validating TurboID-based interactions and the limitations of the TurboID experiments (p.15 lines 3-11).

Referee #3, minor comment 1

In most graphs the expression values or log2FC are shown separately for IFNb and IFNg, however in the heatmaps (Fig 1d, S1d) the IFNb and IFNg results are intercalated keeping them side-by-side for each time point, which makes them more difficult to interpret.

Reply: We are in a quandary about the design of the figure. On the one hand our goal is to visualize gene clusters with distinct behaviors for each IFN type. For this purpose, it would be advantageous to separate the IFN types. On the other hand, we aim at showing similarities and differences between genes induced by each IFN type, for this purpose it is better to maintain the current sample order. While understanding the referee’s point, we prefer to keep the figure as it is, because the suggested change will not increase its overall clarity.

Revision: n. a.

Revision plan: n. a.

Referee #3, minor comment 7

The statement that "IFN-I are the more important mediators of antiviral immunity" is not entirely accurate and may be an oversimplification, as there are certainly articles which suggest a larger role for type ll IFN elements than type l (ref: Yamane D et al., 2019 Nature microbiology). While yes, IFN-I plays a critical role in the innate immune response to viral infections, IFNγ also has antiviral activity and is involved in the adaptive immune response to viral infections, and in some instances to a larger extent than IFN l.

Reply: The Yamane et al study (now mentioned on p 10, lines 22-25 and referenced) agrees with our findings because it shows that IRF1 contributes to the basal expression of an ISRE-driven ISG subset. Our statement about the predominant role of type I IFN versus IFNγ refers to genetic data in both humans (mainly Casanova’s work including effects of autoantibodies against type I IFN, see also the paper about human STAT2 deficiency in the June 15th issue of the JCI, https://doi.org/10.1172/JCI168321) and mice (hundreds of papers) showing that disruption of type I IFN synthesis or response causes profound effects of antiviral immunity (i.e. resulting susceptibilities are first and foremost to viral pathogens) whereas susceptibilities as a consequence of disrupting the IFNγ pathway are first and foremost to intracellular nonviral pathogens such a mycobacteria. In fact, the term mendelian susceptibility to mycobacterial disease (MSMD) was coined by Casanova and colleagues to describe a variety of human mutations that include those of the IFNγ, but not the type I IFN pathway.

Maybe more importantly, the Rosain et al. paper mentioned by the referee which appeared in ‘Cell’ while our study was under review, shows that human IRF1 mutations also fall into the MSMD category (new ref. 65). In contrast, the authors did not observe diminished antiviral immunity. This emphasizes the main conclusions of our study about the relevance of IRF1 for macrophage activation. We discuss this paper on p 14. lines 9-14.

Obviously, this does not exclude a role of type I IFN in nonviral infection or of IFNγ in viral infection, in fact much of our own work has been dedicated to a role of type I IFN in infections with L. monocytogenes. Nevertheless, we think that in a generic statement about the difference between type I IFN and IFNγ it is correct to label the former as predominantly antiviral and the latter predominantly as a macrophage activating factor against nonviral, intracellular pathogens.

Revision: We added discussion of Rosain et al. (ref. 65) on p 14. lines 9-14.

Referee #3, minor comment 8

The authors claim that a significant portion of ISG promoters is associated with ISGF3 upon IFNγ receptor engagement and that the transcriptomes of macrophages treated briefly with IFNβ or IFNγ exhibit remarkable similarity and sensitivity to Irf9 deletion. However, I am uncertain about the extent of consensus on this claim.

Reply: The data were surprising but supported by ChIP-seq and RNA-seq in wt and IRF9 ko macrophages (ref 10). Data in a follow-up study (ref. 11) and in this manuscript support our original conclusion by demonstrating the impact of the IRF9 ko on IFNγ responses. Importantly, we don’t claim this is true in all cell types, it may well depend on STAT/IRF9 expression levels and tonic IFN signaling.

Revision: n. a.

Revision plan: n. a.

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:

This manuscript by Geetha et colleagues addresses the differences and similarities in gene expression control between Type I IFN (IFNβ) and Type II IFN (IFNγ) signaling. The authors aim to determine the factors responsible for the partitioning of IFNβ and IFNγ-induced transcriptomes and their propagation of diverse biological responses. The authors mention the JAK-STAT paradigm of IFN signaling, which posits that ISGF3 dominates transcriptional responses to IFN-I, whereas GAF is critical for the generation of an IFNγ-specific transcriptome. However, recent investigations suggest that the ISGF3 complex may also play a role in IFNγ signaling. The authors investigate the contributions of IRF1, ISGF3, and noncanonical versions of the ISGF3 complex to the transcriptome divergence produced by IFN-I or IFNγ signaling. They used nascent transcript sequencing to determine how ISGs expression are temporally controlled by the ISGF3 complex or IRF1 and show that temporal control of ISG expression includes transcription factor recruitment, enhancer activation, changes of chromatin accessibility, and control of RNA pol II pausing. The authors also investigate cooperativity between STAT1 and IRF1 correlates with different nuclear interactomes. In its current form, the manuscript suffers from a lack of independent experiment replications and nuance in the interpretation of the results.

Major comments:

1. In Fig. 1d is difficult to interpret and misleading for many reasons. First, the cluster numbering is disconnected from the cluster order; why not numbering them based on the hierarchical clustering and writing the cluster number besides the cluster itself? Second, having a 2-color gradient is misleading; negative values shouldn't be in the same color tone than the positive values. Third, the authors did not provide adequate rationale behind using only the top 1,000 most expressed gene? Why not using all the differentially expressed genes in at least one of the condition to provide a comprehensive analysis? Could this potentially lead to bias in the data, and is there any information lost by not using the - lower - expressed genes fraction? Fourth, it is not clear what the color scale is representing and how the data was transformed. Was a mean centering of the expression values of the log2FC applied to the RNA-seq data to facilitate clustering? Mean centering and z-scoring is a common technique used to adjust expression data, but it can potentially exaggerate differences between samples. More information about the data and analysis should be provided, as it is difficult to determine whether this was a valid approach or not.
2. Fig 2c. The authors claim that RNA Pol II pausing is a major factor in controlling the dynamics of ISG transcription. However, they did not provide sufficient explanation of the results, and in all fairness there is not much variation between the clusters to sustain the claim that this is a major factor in ISG transcriptional control.
3. The large standard deviation bars in the claim that ChIP data confirmed the binding of ISGF3 components to the promoter of Mx2 cast doubt on the validity of the results and conclusions. The authors should consider additional experiments or complementary analyses to validate their findings. Or alternative, to adjust their claims accordingly.
4. On p.5, the authors mention "Representative browser tracks from the Gbp2 and Slfn1 genes further validate this observation" but they are simply referring to genome browser snapshot, i.e., specific genomic examples, extracting from the same single dataset. Without using an independent dataset, this can not "further validate" the initial findings.
5. IRF1 was successfully pulled down with STAT1 bait but not in the reciprocal experiment. The author should discuss this point as it is important for the conclusions. Could it potentially indicate issues with the technique used, and if this could introduce any bias into the results. The statement, "In contrast, interactors of the IRF1 bait did not include STAT1. This discrepancy could result from steric constraints of the tagged proteins due to the limitation of the 10nm distance reached by the biotin ligase," does not seem to be sufficient to explain this discrepancy.
6. The authors interpret their ATAC-seq and ChIP-seq results based on a 2kb window to the TSS of genes, not considering relatively close enhancers or longer range cis-regulatory interactions in their interpretation. For example, they mention on p.7 "Contrasting the strong binding of IRF9 and IRF1 to the Mx2 (cluster 2) and Gbp2 (cluster 9) promoters, respectively, we saw no evidence for direct binding to Lrp11 (cluster 3) and Ptgs2 (cluster 10)", but on Fig 3d they show only the proximal regions. No scale bars are shown either. Moreover, exploring the same published IRF1 ChIP-seq dataset, there is a clear IRF1 binding site at the promoter of Ptgs2, while the authors report none.
7. Lack of statistical analysis on chromatin accessibility claims: The authors claim that ATAC-seq data in BMDMs stimulated with IFNβ or IFNγ for a short (1.5 hours) or long (48 hours) period reveals a striking similarity between transcription and the general trends of chromatin accessibility at regions up to 1000 bp upstream of the TSS (Fig. 2a), suggesting continuous chromatin remodeling during the transcriptional response. However, I would like to know if this conclusion is well-supported by the correlation between the chromatin accessibility from ATAC-seq data from only one sample and the PRO-seq data. The need for additional experiments to verify claims such as the dependence of Ifi44 on IRF1 for gaining ATAC signal, as stated in the claim, "Expression required IRF1 for both, but accessibility of the Ifi44 regulatory region depended upon IRF1 whereas that of Gbp2 acquired an open structure independently of IRF1 (Fig. 5c)."
8. In the figure legends, there is missing information about the number of times experiments were replicated, suggesting that some were done a single time. Moreover, some graphs are missing statistical analysis, e.g., in Fig S3cS3e, S3f, the ChIP-qPCR experiments were done on biological triplicates, there is no mention of statistical test performed, it is not mentioned what the error bars represents (SD, SEM, etc.) and the variance is large, but the authors still interpret these results as significant enrichment of the transcription factors to the Mx2 promoter. Another example are the RNA Pol II pausing index ratios, which show minor variations and not are supported by statistics to support a possible significance. Proper description, replication and statistical analyses of the results are critical.
9. The authors used CRISPR-Cas9 genome editing to generate knockout cell lines. However, they did not verify the knockouts at the protein level. Further experiments could confirm that the targeted proteins are not expressed in the knockout cell lines.
10. On p.9, it is mentioned "IRF1 affects chromatin structure ...". Here chromatin structure is related to minor changes in chromatin accessibility, this can not be qualified as changes in chromatin structure.
11. The authors have not adequately addressed the methodological limitations in their discussion, which extends beyond the aforementioned comments. It is suggested they include a comprehensive discussion of the claims made pertaining to the necessity of IRF1 for accessibility and the potential biases in the interactomes, along with their associated consequences.
12. Although the TurboID experiments identify known STAT1 and IRF1 interactors, the proposed new interactors are numerous and none are validate through independent co-IP experiments. Moreover, the results are very noisy, with little differences between untreated BMDMs (where IRF1 is barely expressed) and IFN-treated conditions.
13. The work should be discussed in the context of the demonstrated physiopathological evidence of the IRF1 and IRF9 functions. IRF9 (Hernandez et al., JEM 2018) and more recently IRF1 (Rosain et al Cell, 2023) were identified as causing non overlapping phenotypes in human patients carrying loss-of-function mutations for these genes. The authors must interpret their results in this context.

Minor comments:

• In most graphs the expression values or log2FC are shown separately for IFNb and IFNg, however in the heatmaps (Fig 1d, S1d) the IFNb and IFNg results are intercalated keeping them side-by-side for each time point, which makes them more difficult to interpret. Suggestion to show the IFNb data first and followed by the IFNg results.
• Fig 1e. The color scales on the GO enrichment graphs are misleading since they use the same blue-to-red gradient for adj p-values ranging from 10-25 to 10-49 and 0.008 to 0.016, which could be considered non significant.
• The inconsistency in the title referring to IFNb as Type 1 but using IFNg instead of Type 2 nomenclature, perhaps consistency is best.
• The incomplete schema in Figure 1a, which only focuses on PRO-seq and does not include the ATAC-seq element.
• Figure 6d includes a color scale of -1 to +3, but it is unclear what these values represent and how they were calculated per interactor. The figure legend should be revised to clarify this information.
• The clearer labeling of Figure 5a and 5b.
• The statement that "IFN-I are the more important mediators of antiviral immunity" is not entirely accurate and may be an oversimplification, as there are certainly articles which suggest a larger role for type ll IFN elements than type l (ref: Yamane D et al., 2019 Nature microbiology). While yes, IFN-I plays a critical role in the innate immune response to viral infections, IFNγ also has antiviral activity and is involved in the adaptive immune response to viral infections, and in some instances to a larger extent than IFN l.
• The authors claim that a significant portion of ISG promoters is associated with ISGF3 upon IFNγ receptor engagement and that the transcriptomes of macrophages treated briefly with IFNβ or IFNγ exhibit remarkable similarity and sensitivity to Irf9 deletion. However, I am uncertain about the extent of consensus on this claim.
• Fig 1e, S1c. Graphs having circles of varying sizes in function of a value are named "bubble plots" and not "dot plots".
• Fig S1b, S3b. The PRO-seq was generated in triplicates, hence these graphs should include the Log2FC for the individual data points.
• Fig S3c legend. It is mentioned "Graph represents RT-qPCR of genomic Mx2". RT-qPCR usually stands for reverse transcription quantitative PCR, hence we suggest to change to "ChIP-qPCR" or qPCR. Confusingly, in the literature the term "RT-PCR" is used for real-time PCR and "qPCR" for quantitative PCR. Also, the authors should be specific about the "genomic" region targeted; the graphs mention "promoter", hence it would be appropriate to use the same designation in the legend.
• Fig S3e. The y-axis names are missing.
• In the genomic snapshot shown, only bars or fading triangles are shown in place of the gene body. The authors should provide an accurate gene structure; i.e., exons and introns.
• Raw cells are sometimes spelled as "Raw" and other times as "RAW". Please choose one for consistency.
• In p.10 l.20, the figure number is missing.

Significance

Nature and significance of the advance:

The paper presents an investigation of the transcriptional response to IFNβ and IFNγ in mouse bone marrow-derived macrophages and identifies key factors controlling the dynamics of interferon-stimulated gene (ISG) expression. The study employs cutting-edge technologies such as PRO-seq and ATAC-seq to assess transcriptional and chromatin accessibility changes, respectively. The results can potentially provide new insights into the transcriptional regulation of ISGs and the factors controlling their expression, which have significant implications for understanding the immune response to viral infection and cancer. Overall, the work could represent a conceptual advance in the field of immunology and epigenetics surrounding the transcriptional regulation of IFN, but validations and further mechanistic results are required.

Contextualization of the work:

The study builds on previous research on the transcriptional response to interferons but provides a more detailed and comprehensive investigation of the underlying mechanisms. Some of the key references that the authors build on include studies on the role of IRF9 in interferon signaling, the regulation of chromatin accessibility during immune activation, and the characterization of interferon-stimulated gene expression. However, the current study goes beyond these previous studies by integrating multiple approaches to examine the transcriptional and epigenetic changes that occur during interferon signaling of two types, I and II.

Audience and potential impact:

The findings of the study are likely to be of interest to a wide range of researchers in the fields of immunology, molecular biology, and epigenetics, as well as those interested in the transcriptional regulation. The study may also be of interest to clinical researchers investigating the use of interferons as therapies for viral infections and cancer. The identification of factors controlling ISG expression may have implications for the development of new interferon-based therapies, as well as for understanding the mechanisms of resistance to interferon treatment in patients.

Field of expertise:

Overall, the study is contributing to our understanding of the differiential transcriptional response to interferons and the factors controlling ISG expression. Upon provide further mechanistic demonstrations and validations, the work could have significant implications for both basic and clinical research.

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 interesting study addresses underlying molecular mechanisms that distinguish transcriptomes induced by type I and II IFNs. It is widely accepted that they induce distinct and overlapping genes. The IFN field has shown that type I IFNs can induce both ISGF3 (STAT1/STAT1/IRF9) and GAS (STAT1/STAT1) activation, while type II IFNs only induce GAS elements. The current study adds to some of these observations, including the cooperation of ISGF3 and IRF1 at later time points. They show that ISGF3 and IRF1 can affect enhancers and modify chromatin accessibility. While some of these observations are incremental, this study would significantly interest the interferon community.
• Biological significance is limited as this study is largely descriptive and they do not test the hits obtained from BioID.
• The sequencing and BioID data are not submitted to public databases.

Significance

This interesting study addresses underlying molecular mechanisms that distinguish transcriptomes induced by type I and II IFNs. It is widely accepted that they induce distinct and overlapping genes. The IFN field has shown that type I IFNs can induce both ISGF3 (STAT1/STAT1/IRF9) and GAS (STAT1/STAT1) activation, while type II IFNs only induce GAS elements. The current study adds to some of these observations, including the cooperation of ISGF3 and IRF1 at later time points. They show that ISGF3 and IRF1 can affect enhancers and modify chromatin accessibility. While some of these observations are incremental, this study would significantly interest the interferon community.

While this reviewer does not suggest additional experimentation, this manuscript would be suitable as a resource paper.

My expertise is in innate immunity.

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 authors examine the differences between the genes induced by type I IFNs and IFNg by examining nascent transcripts over a prolonged period of time. The overall question being asked is very broad and the authors generate a massive amount of data that is hard to understand and interpret. The authors also fail to take into consideration the secondary genes induced by products of primary genes. For example, IFNb will induce other type I IFNs that will act in cis or trans to induce secondary ISGs. Also, it is not clear what effect cell death has on gene expression especially at later time points. The differential roles of ISFG3 (IRF9) and IRF1 are interesting but the biological meaning or outcomes of differences in gene expression at 24 and 48 hours is not entirely clear to this reviewer.

Major Issues:

• Figure 2. Difficult to interpret data as it is presented. Consider quantifying figure 2C in order to make "changes in Pol II pausing were more pronounced during IFNb signaling" statement more apparent.
• How are you distinguishing autocrine signaling in the BMDMs driven by IFN treatment from late transcripts (for example, at 48 hours are differential genes due to autocrine cytokine signaling or are they truly late transcripts)?
• Figure 3D. Authors choose Gbp2 (as positive control for IFNg driven gene), but don't show that Gbp2 is a IFNb independent gene. Consider using IRF1 KO BMDMs in this data as well.

Minor Issues:

• Simplify figure 4B- consider focusing on most differentially expressed genes between clusters
• Define known IFNg and IFNb driven genes when they are introduced in figure 2 rather than in discussion
• Clarify which cell types (IRF1 KO vs IRF9 KO) are used in figure 5 A/B.
• Unclear whether IRF1 expression in figure 3A is from whole cell lysate or nuclear fraction.
• Authors suggest IFNb treatment induces less IRF1 at later time points, however loading control also seems slightly lower than other considerations. Is it possible that IFNb treated cells are dying at later time points, given that type I IFN signaling can be pro-apoptotic.

Significance

I accepted the request to review the paper based on the abstract and the idea that this was mechanistic investigation of the role of ISGF3 and IRF1 in regulation of genes induced by type I IFN and IFNg. However, after thorough reading of the paper, I feel that I am not aptly qualified to evaluate all aspects of the manuscript. Many of the data are bioinformatic analysis and I do not have sufficient expertise to either understand the analysis or offer my interpretation of the conclusions drawn by the authors. I suggest that the best path forward is to find another suitable reviewer.. Hopefully, other reviewers have already offered you their suggestions to make an informed decision.

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 careful reading of our manuscript and constructive comments.

2. Description of the planned revisions

Reviewer 1) “Indeed, the manuscript describe the alteration of total brain O-GlcNAc levels, but understanding pathways or protein specific changes would allow to identify the mechanisms potentially at the basis of the development of intellectual disability.”

Response: While finding the pathways involved in phenotypes described here is beyond the scope of the present manuscript, we plan to include RNAi experiments elucidating cell types responsible for the sleep phenotype observed in sxc mutant flies.

Reviewer 3) “2) Lacing fly food with compounds can sometimes lead to phenotypes not actually caused by the drug. There are reports I have previously seen where the compound can make the food more aversive or attractive, both leading to results not due to the drug. Specifically, it has been previously reported that starved flies (if the compound leads to aversion from the food and causes starvation) will reduce the bouts of sleep in Drosophila ( Masek et al J Exp Biol 2014; Figure 4). Do the authors know if the TMG treated food eaten at the same level as normal food? Is there the potential for a starvation phenotype?”

Response: We appreciate this insight and we plan to perform this control experiment. Briefly, this will entail measuring male adult ingestion of Thiamet G laced food by adding Blue No. 1 dye and measuring absorbance of lysed flies, as previously described in Wong et al. 2009 (PMID: 19557170).

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

Reviewer 1) “In figure 1C the blot show a different MW range compared to blots 1A and 1B, author should correct. “ and “For figure 1 and 2 the dot graph are too small and difficult to read”

Response: The figures have been amended to address this.

Reviewer 3) “In the methods - Neuromuscular Junction Immunohistochemistry - which muscles and which types of boutons were imaged was not denoted in this section - it is described in results (lines 210-211) but should be in methods for ease to the reader.”

Response: The methods section has been amended.

“The statistics and data analyses are some of the best I have seen to date. One concern is the removal of a single outlier data point described at line 575. Was this necessary? Does it change the data? If not, I would recommend leaving it in. If it does, I would further recommend additionally biasing toward the alternative hypothesis by additionally removing the data point that lies furthest from the outlier. This would reduce bias.”

Response: Removal of an outlier does indeed change the results of the data. Following the suggestions of the reviewer, we re-analyzed our data removing the minimum for the group for which we previously removed an outlier (the maximum).

“1) line 391 mentions that feeding higher doses of TMG results in a non-rescue phenotype. Is there any data to support this statement (maybe supplementally) to give the reader the full picture of the availability of this compound? For example, how far above 250 uM does this happen?”

Response: This statement refers to adult Thiamet G feeding experiments, and the data to support this statement can be found in figures 2B and S2A. This statement has been amended for clarity and to include the caveat that even higher doses of TMG were not trialed.

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

Reviewer 1) “Authors employed RL2 antibody for O-GlcNac detection, however it recognized mainly high MW proteins and it would be nice to obtain the alteration profile of low MW proteins at the same conditions.”

Response: We agree that the use of a single method for detecting O-GlcNAcylation is limited, however, there is no reason to believe that immunoblotting using this antibody would bias the interpretation of the effects of mutations studied here on global O-GlcNAcylation. Specifically, there is no reason to believe low molecular weight proteins are recognized and modified by OGT differently to high molecular weight proteins. While gaining insight into substrate specific alterations in O-GlcNAcylation is of great interest to us, this is technically very challenging and beyond the scope of this study.

Reviewer 2) “… would it be possible for the authors to overexpress specifically in neurons wildtype OGT postnatally on a mutant background and quantify the effects on neuro-muscular synapse number and morphology? It would be interesting to compare these data with a similar experiment where they overexpress wildtype OGT in the corresponding muscle.”

Response: While temporal control of transgene expression is possible in Drosophila, it is not a technique that we routinely use and would require extensive optimization to include in the present manuscript.

_Reviewer 3) “In Figure 3D the authors show sxcWT compared with OgaKO with no significant difference at ~20 boutons in the count. Other work done by [47] in their reference list (ref 47: Figure 2D) shows an increase in OgaKO boutons vs WT and also shown in [50] (ref 50; Figure 4B) where # of boutons in 1B muscle 4 is increased in OgaKO significantly. There appears to be a difference in what was found with OgaKO vs controls in the authors' results vs these two manuscripts and it should be noted and explained to the reader.” _

Response: This is indeed an inconsistency we have observed, however, looking at reference Fenckova et al. 2022 (47 in our manuscript) we find that in figure legend 2 the following is stated: “None of the parameters is significantly affected in the OgaKO larvae (N = 30, in purple; OgaKO experiments were performed simultaneously and first published here [53] with significantly increased bouton counts (p <0.05) without multiple testing correction)” Reference [53] in the quote refers to Muha et al. 2020 (reference 50 in our manuscript). Therefore, it appears that this effect is too weak to withstand multiple correction testing, which we employ in our analysis.

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 written by Czajewski et al. "Rescuable sleep and synaptogenesis phenotypes in a Drosophila model of O-GlcNAc transferase intellectual disability" is a novel approach to examining genetic missense mutations representing a patient derived OGT mutation in quantifiable phenotypes coupled with genetic and pharmacological manipulation. The authors find novel contradictory results in synaptic bouton parameters than previous work leading to increased interest in these results. The authors also use pharmacological intervention to reverse the phenotypes derived from the OGT mutations creating an interesting path forward for these types of studies. The manuscript was well written, the experiments are sounds, and the analyses are extremely well done. The manuscript would benefit from addressing a few concerns:

Minor:

In the methods - Neuromuscular Junction Immunohistochemistry - which muscles and which types of boutons were imaged was not denoted in this section - it is described in results (lines 210-211) but should be in methods for ease to the reader.

The statistics and data analyses are some of the best I have seen to date. One concern is the removal of a single outlier data point described at line 575. Was this necessary? Does it change the data? If not, I would recommend leaving it in. If it does, I would further recommend additionally biasing toward the alternative hypothesis by additionally removing the data point that lies furthest from the outlier. This would reduce bias.

Major:

In Figure 3D the authors show sxcWT compared with OgaKO with no significant difference at ~20 boutons in the count. Other work done by [47] in their reference list (ref 47: Figure 2D) shows an increase in OgaKO boutons vs WT and also shown in [50] (ref 50; Figure 4B) where # of boutons in 1B muscle 4 is increased in OgaKO significantly. There appears to be a difference in what was found with OgaKO vs controls in the authors' results vs these two manuscripts and it should be noted and explained to the reader.

The results working with Thiamet G (TMG) is very interesting and needs a bit more clarification. I tried to find other research where TMG is fed to Drosophila, and could not find this, and I suspect this is novel and very interesting, especially as a tool. However, I do have concerns about the details for this feeding and would like to further understand a few things that came up in the manuscript that need to be addressed:

1. line 391 mentions that feeding higher doses of TMG results in a non-rescue phenotype. Is there any data to support this statement (maybe supplementally) to give the reader the full picture of the availability of this compound? For example, how far above 250 uM does this happen?
2. Lacing fly food with compounds can sometimes lead to phenotypes not actually caused by the drug. There are reports I have previously seen where the compound can make the food more aversive or attractive, both leading to results not due to the drug. Specifically, it has been previously reported that starved flies (if the compound leads to aversion from the food and causes starvation) will reduce the bouts of sleep in Drosophila ( Masek et al J Exp Biol 2014; Figure 4). Do the authors know if the TMG treated food eaten at the same level as normal food? Is there the potential for a starvation phenotype?

Significance

There is a novel technique in this manuscript that could enhance OGT research in Drosophila, which is significant.

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 authors of this manuscript describe the effect on neuronal development and function in drosophila of OGT mutations derived from patients with intellectual disability. OGT is an enzyme that adds the posttranslational modification O-GlcNAc to proteins. Once added, O-GlcNAc can be removed by the enzyme OGA. O-GlcNAc cycling on and off proteins has been associated not only with intellectual disability but a range of other brain-dependent disorders. However, molecular mechanisms by which O-GlcNAc cycling may affect brain development and function are largely unclear. It is equally unclear whether and how disorders associated with OGT mutations may be treated. This manuscript presents evidence that it is possible to rescue neurological phenotypes dependent on patient-derived OGT mutations using genetic or pharmacological manipulations of OGA. These results strongly suggest that at least some aspects of the phenotype of patient-derived OGT mutations depend on O-GlcNAc cycling rather than other mechanisms. Excitingly, they also suggest that it may be possible to treat patients suffering from OGT mutations with drugs that target OGA after the baby has been born. This last point is critical not only for the field of OGT-associated disorders but for the whole field of intellectual disability.

Minor comment:

While the manuscript delivers its message clearly with a simple and concise language, the manuscript would become even stronger if the observation that some aspects of intellectual disability can be treated postnatally is substantiated with additional methods that are more specific. The current data are also somewhat difficult to interpret because the pharmacological and genetic manipulations of OGA used so far may not be a direct rescue of the OGT mutations, which the authors also point out. For example, would it be possible for the authors to overexpress specifically in neurons wildtype OGT postnatally on a mutant background and quantify the effects on neuro-muscular synapse number and morphology? It would be interesting to compare these data with a similar experiment where they overexpress wildtype OGT in the corresponding muscle. These experiments would both strengthen their finding that it is possible to rescue neurodevelopmental conditions postnatally and give further evidence to the molecular mechanism by which OGT affects neurodevelopment.

Significance

In summary, while it is a short manuscript and on a topic studied previously, its data are novel, clearly presented and would appeal to researchers within and outside the field of O-GlcNAc. It is ready for publication as it has been submitted but including more experiments along the lines suggested above would help it reach one level higher.

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 from Czajewski and colleagues demonstrates that patients-derived OGT mutation can lead to reduced O-GlcNAc levels, which can be rescued by genetic OGA ablation or pharmacological OGA inhibition. Several studies in the last decade demonstrated that O-GlcNac homeostasis is crucial for brain development and function and that its alteration is deeply involved in neurodegeneration and cognitive decline. Therefore rescuing protein O-GlcNAcylation by targeting OGT/OGA cycling could represent a valuable therapeutic approach for intellectual disability. Results obtained by the authors are very promising and support the notion of a mutual interplay between OGT and OGA in regulating brain O-GlcNAc levels. Furthermore, the partial rescue of synaptogenesis and sleep stability support the efficacy of OGA reduction in rescuing O-GlcNAc levels.<br /> I do not find Major flaws in the manuscript structure or in the experimental approach, however I believe that the manuscript would benefit of the analysis of the molecular target that lead to brain defects under OGT mutation and that are rescued after OGA inhibition. Indeed, the manuscript describe the alteration of total brain O-GlcNAc levels, but understanding pathways or protein specific changes would allow to identify the mechanisms potentially at the basis of the development of intellectual disability . Furthermore, it would be also interesting to understand if the mutation of OGT has direct or indirect effects on Ser/Thr phosphorylation levels.

Minor comments:

Authors employed RL2 antibody for O-GlcNac detection, however it recognized mainly high MW proteins and it would be nice to obtain the alteration profile of low MW proteins at the same conditions.

In figure 1C the blot show a different MW range compared to blots 1A and 1B, author should correct.

For figure 1 and 2 the dot graph are too small and difficult to read

Significance

General assessment: I believe that the study is well executed and interesting since it nicely demonstrate the influence of OGT mutation on O-GlcNAc levels and the efficacy of OGA reduction in rescuing the process and in improving synaptogenesis and sleep stability. However, I also believe that a better understanding of the molecular mechanisms involved could substantially improve the study.

Advance: the present study provide further knowledge about the physio/pathological role of OGT/OGA cycling in the brain.

Audience: basic researchers

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]

Here we describe the experiments that will be performed as specifically requested by the reviewers to strengthen the main conclusions of the manuscript, and we also justify the utilization of a specific shRNA for KIS knockdown.

2. Description of the planned revisions

Reviewer 1

1. • Building on the experiments they perform in a KIS knock-down context (e.g. Fig. 3B, or previously described spine phenotype), the authors should investigate whether inhibiting PTBP2 in this context (through shRNA or expression of a phospho-mimetic construct) might suppress the phenotypes observed when inactivating KIS.*

We will use shRNAs against PTBP2 to test the functional interactions with KIS on CamK2b splicing as in Fig. 3B and we will also assess possible effects on spine morphogenesis. Since it does not show dominant negative effects, the phosphomimetic mutant of PTBP2 will not be included in these experiments.

• Based on Figures 1E and 3A, it seems that KIS downregulation affects both exon inclusion and exon skipping, and that its function in exon usage is only partly explained by modulation of PTBP2-dependent exons. Have the authors analyzed the populations of PTBP2-dependent exons that are regulated by KIS in an opposite manner? This may point to specific classes of transcripts (in terms of expression pattern, function, molecular signature) important in the context of endogenous neuronal differentiation.*

We will extend our analysis to exons that are regulated by KIS and PTBP2 in the same direction. We fully agree with the reviewer that these data may uncover gene sets with specific functional implications different from those in which KIS and PTBP2 counteract each other.

Reviewer 2

1. • FigEV4 (also introductory text on p3): RRMs 3 and 4 of PTBP1/2/3 fold as a single back to back packed didomain - with the so-called linker contributing to the didomain fold (e.g. PMID: 24688880, PMID: 16179478) and also extending the RNA binding surface by creating a positive patch (e.g. PMID:20160105 PMID: 24957602). AlphaFold successfully predicts the didomain in full length PTBP2 (https://alphafold.ebi.ac.uk/entry/A0A7I2RVZ4). The authors should therefore use AlphaFold2 to predict the RRM3-4 di-domain structure of wt and phosphomimetic mutant PTBP2s. Phosphorylation of S434 or S434D, which is on the C-terminal end of RRM3 may have no predicted effect on RRM3 alone (FigEV4), but it could conceivably disrupt didomain packing, which could itself have important knock-on consequences for RNA binding. In addition, the introduction of negative charges at S434 might affect the ability of R438, K440 & K441 to interact with RNA. An image of the didomain charge density of WT and mutant PTBP2 would be useful to address this.*

As suggested by the reviewer, we have considered the di-domain structure of RRM3 and RRM4, and AlphaFold2 predicted no effects by the phosphomimetic residues. We will add these data to the revised version of the manuscript.

• Figure 4 could also easily go further in experimentally testing the effects of individual phosphomimetic mutations upon protein-protein interactions (Alphafold predicts that S178D, but not S308 or S434D, should affect Y244 mediated interactions, such as MATR3). The co-IP approach in Fig 4A could readily be used with FLAG-PTBP2 mutants. Likewise, consequences of individual mutations upon RNA binding (Fig 4D) could be tested. The use of a Y244N mutant here would test whether the loss of RNA binding is a consequence of the loss of protein-protein interactions. Such experiments are not essential, but they are readily carried out and have the potential to unravel the consequences of the individual phosphorylation events (more correctly of phosphomimetic mutants).*

Extending the analysis to this residue will be a very interesting contribution to the article. After building the Y244N mutant we will test PTBP2 interactions with protein partners and the splicing reporter RNA as in Fig. 4A-D.

Reviewer 3

• Part of the reported splicing changes might reflect an indirect consequence of an altered differentiation contributing to the correlation observed in figure 1F. It would be interesting to confirm splicing changes using shorter incubation times with the shRNA compared to the 11 days used in this study.*

The levels of splicing regulators such as PTBP1 and PTBP2 change quite markedly during the initial phases of neuronal differentiation (Zheng et al 2012). However, we observed no change in their levels when comparing KIS knockdown to control conditions, suggesting no major upstream effects on the differentiation program per se. In any event, we will analyse expression levels of transcription factors key to neuronal differentiation.

• Previous papers of the group described a function of KIS in translation (Cambray et al 2009, Pedraza et al 2014). This is not discussed here. For example, the possibility that RBPs are regulated by KIS at the translation level is not excluded by the analysis in Fig EV2a.*

In our experiments coexpressing KIS with splicing factors in HEK293 cells (Fig. 4A) we did not observe any reduction in their levels. We will include the corresponding immunoblots from input samples in the revised version of the paper. We will also measure PTBP2, Matrin3 and hnRNPM levels by immunoblots in KIS-knockdown cortical neurons to test this possibility further.

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

(None at this stage)

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

Reviewer 3

1. • To minimize possible off target problems, the RNAseq analysis would be more convincing if replicated with a second shRNA to knockdown KIS.*

The efficiency of the selected shRNA had been validated both by the supplier (Sigma) and in our previous work, which also included a complementation assay (see Fig. 4 in Pedraza et al 2014).

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 manuscript, the authors explored the function of the protein kinase KIS in splicing regulation associated with neuronal differentiation in vitro. KIS is a serine threonine kinase known to phoshorylate splicing factors such as SF1 and SUGP1, and to be preferentially expressed in adult brain in mammals. Using an shRNA based approach, the authors characterize cassette exon usage upon partial KIS depletion in cultured mouse cortical neurons. In parallel using mass spectrometry of proteins in KIS overexpressing HEK293 cells, they identify potential KIS substrates including the splicing regulator PTBP2. They confirm that recombinant KIS can phosphorylates PTBP2 in vitro. They show a correlation between KIS-activated and PTBP2-inhibited exons using published data for this factor. They report opposite effects of KIS and PTBP2 on CamKIIB splicing and Finally, coimmunoprecipitation and FRET experiments suggest that KIS inhibits the interactions of PTBP2 with known protein binders, hnRNPM and Matrin3 as well as with RNA. Altogether these data suggest that KIS downregulates PTBP2 during neuronal differentiation.

Major comments:

Overall the manuscript is well written and the data are interesting. However several points could have been more extensivelly studied or discussed to achieve a stronger demonstration of the role of KIS in PTBP2 phosphorylation and neuronal differentiation.

1. To minimize possible off target problems, the RNAseq analysis would be more convincing if replicated with a second shRNA to knockdown KIS.
2. Part of the reported splicing changes might reflect an indirect consequence of an altered differentiation contributing to the correlation observed in figure 1F. It would be interesting to confirm splicing changes using shorter incubation times with the shRNA compared to the 11 days used in this study.
3. Standard deviation is more relevant to describe data dispersion in all figures.
4. Previous papers of the group described a function of KIS in translation (Cambray et al 2009, Pedraza et al 2014). This is not discussed here. For example, the possibility that RBPs are regulated by KIS at the translation level is not excluded by the analysis in Fig EV2a.

Minor comments:

Figure 1:

The authors state that : "KIS...accumulates in nuclear sub-structures adjacent to those formed by splicing factors". As the figure presents in fact GFP-KIS, it should be mentioned, and how this localisation is relevant for endogenous KIS should be adressed.

Fig EV1: SI range in pannel D is very different from that in pannel C and Fig1E.

On page 4 "KIS expression reached maximal levels in hippocampal cultures (Fig 1B)." However the figure legend indicate that this analysis was performed with cortical neurons. The use of cortical or hippocampal neurons along the manuscript should be clarified.

page 4 " KISK54A, a point mutant without kinase activity" The authors should indicate the reference.

Figure EV2C: It is not clear whether the Coomassie staining and autoradiography do correspond to the same gel.

Figure 3C The authors use a dual fluorescence reporter to analyse PSD95 exon 18 splicing. However the well to well variability in such experiments might be elevated. Not only the cells number in a single well but also the number of replicates should be indicated and well to well variability reported.

Figure 3D. The precise timing for the transfection and culture of cells before staining is unclear

Figure 4A. The input should be loaded to evaluate the coIP efficiencies and ascertain that KIS does not downregulate Matrin3 and hnRNPM levels.

Figure EV4A. No difference of Matrin3 binding is to be seen on the gel. In addition, the authors should confirm that PTBP2 or binders are phosphorylated by recombinant KIS. The preparation of GST-KIS is not described. Page 6: "We found that PTBP2-inhibited exons are significantly (FDR=0.001) enriched in KIS knockdown neurons, supporting the notion that KIS acts on AS, at least in part, by inhibiting PTBP2 activity." This should be rephrased as in fact PTBP2-inhibited exons are enriched among KIS activated exons. Page 10: "SUGP1 is one of the most enriched proteins in our KIS phosphoproteome (see Fig 2A)". Phosphorylation and interaction with KIS was already reported by Arfelli and coll. 2023 supplementary figure 2.

" It forms part of the spliceosome complex, interacts with the general splicing factor U2AF2 and has been reported to play an important role in branch recognition by its association with SF3B1." A reference is needed there.

The authors previously reported a differentiation defect in cultured neurons 'Cambray et al, 2008' that was not observed by another group (Manceau et al., PLOS One 2012). This should be discussed in view of these more recent results. Is there any differentiation defect in the experiments reported there?

Statistical values are difficult to read in the figures. Please use larger fonts.

Significance

This manuscript brings new elements supporting the function of the protein kinase KIS in splicing regulation in neurons. In particular it identifies for the first time the splicing regulator PTBP2 as a substrate for KIS.

It will be of interest to a specialized audience of researchers interested in splicing regulators in neuronal differentiation.

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

The manuscript by Moerno-Aguilera et al. shows that the brain enriched protein kinase KIS targets the well known neuronal splicing regulator PTBP2 and several of its interaction partners. As a consequence, PTBP2 activity is down-regulated. Using cultured primary immature neurons they show that KIS expression increases during differentiation and that shRNA knockdown of KIS alters the splicing of many alternative exons. Phosphoproteomic anlaysis of HEK293 cells transfected with KIS or a kinase dead mutant (K545A) show that it phosphoryates both PTBP2 as well as a cluster of proteins that are known to interact with PTBP2 or its paralog PTBP1. By comparing the new data on KIS-dependent splicing with previous data-sets on PTBP2-dependent splicing targets they show that KIS appears to act antagoniostically with PTBP2 when it acts as a repressive regulator, but not when it is an activator. Using combinations of wt and kinase-dead KIS with PTBP2 mutants in the 3 main phsophorylation sites (3SA - non-phosphorylatable, S3D - phosphomimetic) to look at the effects on a known PTBP2 functional target, PSD95, they show that the likely effect of KIS is to antagonise PTBP2 function by phosphorylation at one or more of three residues (S178, S308, S434). Finally, they show that transfected KIS (but not K54A) reduces known protein-protein interactions of PTBP2 and that the triple phosphomimetic PTBP2 mutant shows reduced binding to RNA. Alphafold2 predictions show that the S178 phosphomimetic mutant might alter the conformation of the RRM2 domain, in particular altering the environment of Y244, which has been shown in PTBP1 to be critical for interaction with MATR3 and other coregulators.

Major points

In general, the conclusions drawn are consistent with the data. I have a few suggestions where the authors could either extend their findings with a few straightforward additional experiments, or clarify some of the existing data.

FigEV4 (also introductory text on p3): RRMs 3 and 4 of PTBP1/2/3 fold as a single back to back packed didomain - with the so-called linker contributing to the didomain fold (e.g. PMID: 24688880, PMID: 16179478) and also extending the RNA binding surface by creating a positive patch (e.g. PMID:20160105 PMID: 24957602). AlphaFold successfully predicts the didomain in full length PTBP2 (https://alphafold.ebi.ac.uk/entry/A0A7I2RVZ4). The authors should therefore use AlphaFold2 to predict the RRM3-4 di-domain structure of wt and phosphomimetic mutant PTBP2s. Phosphorylation of S434 or S434D, which is on the C-terminal end of RRM3 may have no predicted effect on RRM3 alone (FigEV4), but it could conceivably disrupt didomain packing, which could itself have important knock-on consequences for RNA binding. In addition, the inrtoduction of negative charges at S434 might affect the ability of R438, K440 & K441 to interact with RNA. An image of the didomain charge density of WT and mutant PTBP2 would be useful to address this.

Figure 4 could also easily go further in experimentally testing the effects of individual phosphomimetic mutations upon protein-protein interactions (Alphafold predicts that S178D, but not S308 or S434D, should affect Y244 mediated interactions, such as MATR3). The co-IP approach in Fig 4A could readily be used with FLAG-PTBP2 mutants. Likewise, consequences of individual mutations upon RNA binding (Fig 4D) could be tested. The use of a Y244N mutant here would test whether the loss of RNA binding is a consequence of the loss of protein-protein interactions. Such experiments are not essential, but they are readily carried out and have the potential to unravel the consequences of the individual phorphoryation events (more correctly of phosphomimetic mutants).

Minor

Do KIS regulated exons show enrichment of motifs associated with PTBP2, consistent with the proposed model - particularly CU-rich motifs upstream of exons that are more repressed upon KIS shRNA treatment.

For the splicing analysis pipeline, how were exon-exon junction reads treated? If "only exons with more than 5 reads in all samples" were considered, will this not exclude highly regulated exons that are completely skipped under one condition?

The Introduction mentions U2AF homology (UHM) domains, but neglects to discuss their known binding partners - ULMs (UHM ligand motifs), which contain an essential tryptophan. It would be useful for the discussion to highlight whether any direct KIS interactors possess ULMs and how this relates to the phospho-targets identified here. The authors may wish to draw the parallel with the structurally analagous way that PTBP1 (and presumably PTBP2) interact with their short peptide ligand motifs.

Figure EV2C. The S3A and S308A mutations clearly reduce phosphorylation. However, the effects of S178A and S434A are far less clear. Presumably the quantitation shown in the lower panel of EV2C relies on normalization to PTBP2 protein input, which appears quite variable in the Coomassie gel. It might be better to repeat the experiment with uniform protein inputs. Minimally, details of the quantitation approach should be added to Materials and Methods.

Fig 3D shows PTBP2 overexpression, but the main text (p7) states KIS overexpression.

Fig 4B should have a scale bar for the FRET signal

Fig 4E should indicate the location of S178

Significance

This interesting, clear and concise manuscript provides important new insights into the way that a neuron specific kinase can regulate neuronal splicing networks by phosphorylating and thereby downregulating the known neuronal splicing regulator PTBP2. Alternative splicing is known to play a particularly important role in neurons, so this demonstration of an additional layer of regulation by post-translational modification should make the manuscript of wide interest to investigators of splicing regulation, neuronal differentiation and maturation.

Issues that are not addressed in the manuscript include; i) how does KIS specifically target PTBP2 and related proteins? The UHM domain can mediate interaction with ULM containing splicing factors (such as U2AF2, SF3B1), but none of the identified targets have known ULMs. ii) the consequences of individual phoshomimetic mutants upon protein-protein interactions and RNA binding could readily be explored further using computational and experimental methods already used in the manuscript.

For context, this reviewer has a direct interest in the mechanisms of regulation of alternative splicing, but not in the context of neurons (though I am familiar with a lot of the relevant literature), and I do not have expertise in neuronal cell biology.

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 manuscript, the authors characterized the molecular function of the brain-enriched kinase KIS by combining transcriptome-wide approaches with molecular and functional studies. They uncover that KIS regulates isoform selection of genes involved in neuronal differentiation and inhibits through phosphorylation the capacity of the splicing regulator PTB2 to interact with both target RNAs and protein partners.

Major comments

• This is a very clear and well-written manuscript presenting high-quality and carefully controlled experimental results. The authors used an impressive range of approaches (transcriptome-wide exon usage, phospho-proteomic, imaging, biochemical assays..) to profile exon usage alterations upon KIS knock down and provide a mechanistic understanding of how KIS regulate the splicing activity of PTBP2. Specifically, they convincingly demonstrate that the phosphorylation of PTBP2 by KIS leads to both dismantling of PTBP2 protein complexes and impaired RNA binding. My only main concerns relate to the understanding of the biological context in which the mechanism studied may be at play. That KIS can counteract PTB2 activity through direct phosphorylation has been very clearly shown by the authors using overexpression of KIS and /or PTB constructs in different contexts (HEK293T cells, N2A cell line, hippocampal neurons). Whether this occurs endogenously in the context of neuronal differentiation, and how much this contributes to the overall phenotypes induced by KIS inactivation, is less clear. While fully investigating the interplay between KIS and PTB2 in the context of neuronal differentiation is beyond the scope of this study, the three following points could be addressed to provide some evidence in this direction.

• Building on the experiments they perform in a KIS knock-down context (e.g. Fig. 3B, or previously described spine phenotype), the authors should investigate whether inhibiting PTBP2 in this context (through shRNA or expression of a phospho-mimetic construct) might suppress the phenotypes observed when inactivating KIS.

• Based on Figures 1E and 3A, it seems that KIS downregulation affects both exon inclusion and exon skipping, and that its function in exon usage is only partly explained by modulation of PTBP2-dependent exons. Have the authors analyzed the populations of PTBP2-dependent exons that are regulated by KIS in an opposite manner? This may point to specific classes of transcripts (in terms of expression pattern, function, molecular signature) important in the context of endogenous neuronal differentiation.
• The authors should better discuss when and where they think PTBP2 phosphorylation by KIS might be relevant. Is there evidence that this process (or PTBP2 complex assembly) might be regulated upon differentiation or plasticity?

Minor comments

1. Figures and associated legends are overall very clear and well-organized. Addressing the following points would however help improving the clarity of some Figures:
   • In Figure 2EV2C legend, the characteristics of the 3SA constructs are not described
   • the difference between Figure EV1A and Figure 1H classifications is unclear, nor the interpretation regarding the different GO classes identified
2. Whether PTBP2 is endogenously the major target of KIS explaining transcriptome-wide changes in exon selection is a possibility that remains to be demonstrated. Thus, the authors should correct and tune down the following sentences: "KIS phosphorylation counteracts PTBP2 activity and thus alters isoform expression patterns ..." (end of introduction) "PTBP2 being one of the most relevant phosphotargets" (results, end of the second section)

Significance

• The splicing regulator PTBP2 is a known master regulator of neuronal fate whose tightly controlled expression drives the progenitor-to-neuron transition as well as the establishment of neuronal differentiation programs. How this protein is regulated at the post-translational level has so far remains poorly investigated. In this manuscript, the authors provide a thorough mechanistic understanding of how KIS-mediated phosphorylation of PTB2 impacts on its regulation of exon usage. They also provide a transcriptome-wide view on the function of the brain-enriched KIS kinase in exon usage, uncovering its broad functions in alternative splicing. If the physiological context in which KIS-mediated phosphorylation of PTB2 is induced remains to be precisely defined, this work opens interesting new perspectives on regulatory mechanisms at play during neuronal differentiation. Providing extra lines of evidence indicating that KIS acts on neuronal functions through PTBP2 phosphorylation will help further strengthen this aspect.
• This manuscript will be of interest to different large communities interested on one hand on the regulation of gene expression programs underlying neuronal differentiation and on the other hand on the molecular regulation of major complexes involved in alternative splicing and isoform selection. It opens new perspectives related to the spatiotemporal regulation of neuronal isoform selection.

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)): *

* Srinivasan et al. present a comprehensive study on systematizing the structure-dynamics-function relation of lipid transfer proteins (LTPs), combining extensive molecular simulations and complementary experiments. Indeed, the current state-of-the-art in the field is quite chaotic and fractional, and such systematic studies are necessary to advance our general and conceptual understanding of the mechanisms of action of LTPs. The selected techniques and research strategies are all suitable, their description is sufficient and enables reproducibility; the obtained results are carefully presented and discussed; the conclusions are adequately supported by the data.

Given my primarily computational background, I evaluated mainly the simulation part of the manuscript. Considering experiments, I do not see any significant flows or deficiencies that could diminish the value of the data and following conclusions given in the manuscript. I would even suggest improving the abstract by more explicitly saying that this work includes experimental measurements because it currently reads like purely computational work was performed. *

We thank Reviewer #1 for the positive evaluation of our work. The abstract has now been updated to include that our work allows us to interpret existing data but also to design and perform new experimental measurements.

* Major comments: *

1) Although I like the central message of the paper and have no objections, I am curious whether the conclusion "a more "dynamic" or/and "mobile" part of the protein interacts with the membrane or any other (macro)(bio)molecule" makes sense globally and is not limited to LTPs. For example, it is a reasonable assumption that a more flexible part of the protein, i.e., capable of adopting necessary binding configurations, would be a more likely interacting spot. Locking in a less flexible and more specific configuration upon binding with a target molecule is also anticipated and quite typical, e.g., when ligands interact with target proteins, thereby blocking their function. The authors themselves recognize this paradigm as referring to the enzymes' dynamics. It would be great if authors could comment more on dynamics-function relation, referring to the existing literature, where such observations were/were not observed for different protein families. Performing simulations on proteins that do not exhibit such a feature and do not belong to LTPs, but, e.g., structurally similar to some of the studied LTPs, would be an excellent addition too, highlighting this signature characteristic of LTPs.

We have now added a discussion comparing the mechanism we observe with those described for other proteins such as membrane transporters and receptors. Since those proteins are very different and have been already thoroughly characterized (including with molecular simulations) we don’t think that additional simulations are required. Also, concerning protein binding dynamics, we refer to the excellent review of Wade and coworkers: "Acc. Chem. Res. 2016, 49, 5, 809–815"

"____Notably, the conformational plasticity we observe for LTPs is reminiscent of other, previously described, functional protein mechanisms, including enzyme dynamics during catalysis (____DOI: 10.1126/science.1066176____), the alternating-access model of membrane transporters (____https://doi.org/10.1038/nsmb.3179____) or GPCR dynamics (____https://doi.org/10.1021/acs.chemrev.6b00177____). In all these cases, protein dynamics is strongly coupled to ligand binding (____https://doi.org/10.1021/acs.accounts.5b00516____) and protein function, be it for signaling, transport or enzymatic activity. Unlike for these fields, however, the contribution of structural and spectroscopic studies to uncover LTP dynamics remains quite limited, and our simulations provide an important contribution to fill this gap. We hope that our results will motivate researchers to increase efforts to experimentally quantify LTPs conformational plasticity, e.g. by structural determination of LTPs in different states (or bound to different lipids) or by single-molecule spectroscopy studies."

*Minor comments: *

*

1) Fig 1d. What is so special in Lysine compared to Arginine? Is there any disbalance in their presence in studied proteins? Any correlations between the binding affinity of certain amino acids and their overall presence on the protein surface? *

Indeed, there is disbalance in the presence of lysine and arginine residues in our proteins. The relation between the number of these residues in our dataset is Lys:Arg = 1.6:1. On top of that, and as described in (Tubiana T et al PLoS Comput Biol. 2022 ;18(12):e1010346) lysine is preferred over arginine in peripheral membrane proteins, likely because it induces fewer perturbations in the lipid bilayer. Our data also agree with Tubiana et al, concerning the correlation between abundance of specific residues on the protein surface and membrane binding.

* 2) Fig S1. GM2A and TTPA seem to be irreversibly adsorbed to the membrane on the microsecond timescale in most replicas. Is anything special in these proteins? Did this affect the sampling of a claimed membrane-binding interface?*

Our interpretation of the different adsorption profile of GM2A and TTPA is that these two proteins appear to have higher membrane affinity in our computational assay in comparison with the other proteins in our dataset. However, this has no effect on the membrane-binding interface as the proteins are still able to undergo significant tumbling before binding to the lipid bilayer, as demonstrated by the angle between the two main protein axes and the bilayer normal before membrane binding (Fig. S8 in Supplementary Information).

* 3) A related follow-up question. Multiple replicas were performed to identify the membrane-binding interface. However, if I understand well, the initial orientation of the protein with respect to the membrane was always the same. I found it a pity since performing multiple replicas starting from different initial geometries (e.g., rotating the protein in a somewhat systematic way) would likely result in a more efficient exploration of the conformation space. Can the authors comment on whether this predefined initial configuration could negatively affect the results? Performing a few additional simulations for the most problematic proteins I mentioned earlier (GM2A and TTPA) could be a nice opportunity to apply this strategy. *

In our protocol, all proteins start from the same initial orientation but undergo significant tumbling in solution before interacting with the lipid bilayer, including for the two most extreme cases, GM2A and TTPA (Fig. R1). Hence, we think that there is no bias for what pertains to the final membrane interacting region. We have added the Fig. R1 in Supplementary Information (Fig. S8) and added the following text in the Methods Section:

"____Despite starting from a single orientation, all proteins undergo extensive tumbling before binding to the bilayer, as illustrated by the angle between the two principal protein axes and the membrane normal for the two proteins that display the highest binding propensity, GM2A and TTPA (Fig. S8)."

* 4) How was the volume of the cavity affected by mutations in STARD11 and Mdm12? Do these data somehow correlate with the experimentally observed reduced efficiency of the lipid transfer? *

Our data on the volume of the cavity in STARD11 and Mdm12 are inconclusive. However, we caution from such a simplistic interpretation, since it completely neglects the lipid-bound conformation that normally has a much larger cavity than the apo form (Fig. 3).

*5) I would appreciate it if the authors considered playing with the templates of the main Figures at later stages because in the current version, and when printed on A4 paper, the readability of certain graphs and pictures is uncomfortable and sometimes even impossible. Obviously, the final schematics would depend on the journal and its formatting. *

We will modify the templates of the main Figures to improve readability according to journal formatting.

* **Referees cross-commenting** *

* I would like to acknowledge the thoughtful and detailed reviews provided by other reviewers. I do like their reports, and I believe that by addressing the reviewers' comments and incorporating their revisions, the article will significantly improve in terms of scientific rigor and contribution to the field. *

*Reviewer #1 (Significance (Required)):

This manuscript is a solid scientific work addressing gaps in our knowledge about Lipid Transfer Proteins by employing state-of-the-art methods. It advances the field on conceptual and fundamental levels. This study is of interest to both computational biophysicists and physical chemists (to whom I belong myself) as well as experimentalists, who seek a rational explanation of the experimental observations. *

We thank the reviewer again for the positive evaluation of the significance of our work.

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

* Summary:

In a combined computational and experimental study, the authors provide insights into general features of lipid transfer proteins (LTPs), which play key roles in lipid trafficking: Through molecular dynamics simulations of a diverse set of 12 shuttle-like LTPs, they demonstrate that LTPs consistently exist in an equilibrium between two or more conformations, whose populations are modulated by a bound lipid, and that residues significantly involved in these collective conformational changes typically interact with a membrane. Their simulations indicate that conformational plasticity is a general feature of LTPs, leading them to suggest that the ability to change conformations is essential for LTP function. They test the generality of this hypothesis through in cellulo assays of two LTPs (STARD11 and Mdm12) that were not originally simulated. While experiments of STARD11 support their hypothesis, those presented for Mdm12 provide ambiguous results. *

*

Major comments: *

* Throughout the manuscript, it's stated that common 'dynamical features' correlate with LTP function. The accuracy of this statement is unclear since 'dynamical features' are never precisely defined and, while equilibrium conformational ensembles are characterized, dynamics (ie kinetics or time-dependent observables) are not. Please clarify.*

We plan to improve the scholarly presentation of our article to clarify this issue. In short, two distinct properties modulate protein function: 1. Conformational plasticity, i.e. the (thermodynamic) ability of the protein to adopt different conformations (and with different populations depending on the bound substrate). 2. Conformational “dynamics”, i.e. the propensity to exchange between these different thermodynamic states. This ability depends on the free energy barriers between different states and it is intrinsically a kinetic (rather than thermodynamic) property.

*More importantly, further evidence is needed to determine a correlation with *function*. LTPs are suggested to have faster transfer rates (a measure of function) if the apo form adopts a substantial population of holo-like conformations, akin to enzyme preorganization. This is further tested by rationally mutating STARD11 and Mdm12. However, the support for this conclusion and if these mutations alter the LTPs conformational ensembles as desired is unclear: *

In our opinion, the interpretation suggested by Reviewer #2 that there is a “correlation” between transfer rates and the overlap of apo-like and holo-like conformations, though fascinating, cannot be derived from the available data at this stage, and we did not mean to imply as such. Rather, lipid transport is a complex phenomenon that involves several steps (membrane binding/unbinding, lipid uptake/release,…). Our simulations indicate that protein conformational plasticity, including potentially the overlap between apo-like and holo-like conformations, also influences lipid transfer rates. We will clarify this aspect in the text.

* Is there a quantitative correlation between the overlap of apo and holo conformational distributions (as could be quantified by KL divergence or Wasserstein distance, for example) and difference in transfer rates as suggested by Fig S6?*

We plan to compute quantitative correlation between apo and holo conformational distribution for Fig.S6 and for mutant simulations (see answer below) but, as discussed above, we are skeptical that we will observe a clear correlation.

* The conclusion and the generality of the findings would be greatly strengthened if a correlation can be shown for other LTPs through additional simulations of mutants whose transfer rates have been previously characterized experimentally in the literature. (For example: Ryan 2007 PMID 17344474, Grabon 2017 PMID 28718450, Iaea 2015 PMID 26168008, among many others)*

We are currently running simulations of several mutants to address this point and provide additional data/context.

* While differences in the apo conformational ensembles of the WT and mutants are observed in Fig S7b and d, if these mutations reduce overlap with holo-like conformations is not determined. Simulations of the WT holo forms are needed to properly test this hypothesis. *

We are currently performing these simulations.

• For Mdm12, mutations are specifically made to "lock the protein in the apo-like state;" however, the mutant adopts conformations distinct from the apo form as show in Fig S7d. How do the authors interpret the results of the cellular assays considering this and could it help explain why the mutant has similar kinetics to WT? What may explain the puzzling results of similar transfer kinetics but differing mitochondrial morphology? *

As discussed above, interpretation of lipid transport rates based exclusively on apo and holo conformational population is premature, as this is a complex mechanism that depends on many variables. For what concerns the experimental results, we think three explanations are possible: 1. Mitochondrial morphology could be more sensitive to small variations in lipid composition than our METALIC assay. 2. Our assay only quantifies transport of unsaturated PC and PE species, and we can’t quantify variations in transport of other lipid species that are likely to also be transported by ERMES, such as PS and PA. 3. According to a recent structural model (Wozny et al, Nature 618, 88–192, 2023), Mdm12 might be part of a tunnel-like LTP complex in which it doesn't establish direct interactions with nearby organellar membranes. As such, its mechanism might be different from the one described here for other shuttle-like lipid transport domains. We will discuss these possibilities in the main text.

• Confounding factors potentially complicate the interpretation of the in cellulo experiments. Simpler in vitro experiments may be better suited to determine if altering LTP's biophysical properties, namely rationally altering the population of apo- vs holo-like configurations, quantitatively affects transport rates as suggested.*

We agree with Reviewer #2 that this information could be useful. However, this is beyond our technical abilities, and it would require lengthy and expensive experiments that are unlikely to be completed within a reasonable time framework for a revision (3 months). We have rather opted to better discuss our model in the context of published in vitro lipid transport experiments.

• The abstract, intro, and title highlight that the manuscript's findings are indicative of and correlated with *function* but on p. 12 it's foreseen "that future studies will focus on the functional consequence of such observation." Please reconcile these conflicting statements and ensure connections to function are accurately described. The current title is rather bold. *

We will rewrite and clarify the extent of our hypotheses and validations.

* All mentions of "correlation" throughout the manuscript need to be quantitatively evaluated or properly qualified. In addition to that mentioned above regarding Fig S6, what is the correlation coefficient between residues' contribution to PC1 and membrane interaction frequency (Fig 2)? *

To address this point, we will quantify the correlation between residues' contribution to PC1 and membrane interaction frequency. However, we expect a low correlation between residues' contribution to PC1 and membrane interaction frequency for at least two main reasons. __ First, not all residues contributing to PC1 interact with membranes, but only a subset, as discussed above. Second, our methodology to compute membrane binding, based on the geometric distance between residues and bilayer, is intrinsically quite noisy (since residues in proximity of bona fide membrane binding regions will also appear as involved in membrane binding), thus making quantification of correlations somewhat inaccurate. Rather, we will try to explain in the text that our observations are not of "correlation" but rather of dependence/association, and we will use quantitative measures to quantify these properties (such as rank correlation coefficients or multivariate analyses).__

* Residue's contributions to collective conformational changes are found to be indicative of membrane binding. Yet, membrane interacting residues are identified from CG simulations that cannot capture such collective conformational changes due to the use of an elastic network. Given that the CG simulations agree with previous experimental findings, this suggests that collective conformational changes are not important for membrane binding. *

We disagree with this interpretation by Reviewer #2 of our data: we do not claim that residue's contributions to collective conformational changes is indicative of membrane binding. Rather, membrane binding happens at protein regions displaying high contribution to collective conformational changes. This distinction is subtle but important: protein motion does not determine membrane binding regions. Rather, it appears that, for LTPs, membrane binding regions are also characterized by collective motions (suggesting function). We will clarify this in the main text.

*Are similar conclusions drawn from residues' RMSFs? In other words, are local conformational fluctuations just as indicative of membrane binding? *

We will compute protein residues’ RMSFs and compare it with the membrane binding data. However, given that RMSF is representative of thermal fluctuations, we again expect a bad correlation between RMSF and membrane binding. On the other hand, we indeed observe that most membrane binding regions are protein loops, but this is not unexpected (e.g. Tubiana et al, PLoS Comput Biol. 2022 Dec; 18(12): e1010346.). However, such observation does not provide any information on lipid transport, but only on the mechanism of membrane binding. Rather, the observation of a relationship between membrane binding and global motion is more interesting, since the latter is often indicative of protein function.

*The stated correlation may in fact be spurious and instead arise because residues at the entrance to LTP's hydrophobic cavities need to be positioned at the membrane surface for productive lipid uptake and these same residues must undergo significant conformational changes to allow lipid entry. *

This is exactly what we think it is happening and what our data suggest. However, one must remember that our simulations allow us to predict the membrane binding interface, that is often difficult to determine experimentally (and often via indirect evidence). Hence our data provide novel evidence in this direction.

*Is proximity to cavity entrance more or less correlated with membrane binding than 'dynamics'? *

If we consider that, as discussed before, dynamics does not correlate with membrane binding (there are many dynamical regions that are not at the membrane interface), it is safe to assume that proximity to cavity entrance would correlate more with membrane binding. However, we have to consider that often we do not know where the cavity entrance in LTPs is located simply based on structure alone, and hence our approach provides important clues into this process.

p.12 speculatively suggests "the high degree of protein dynamics we observed in membrane proximal regions could potentially facilitate the energetically unfavorable reaction that involves the extraction of a lipid from a membrane." Yet, the logic behind this idea does not make sense since a free energy barrier, an equilibrium thermodynamic quantity, cannot be lowered by changes in dynamics. Please explain.*

Our current understanding of the mechanism of lipid extraction is quite poor. However, both using chemical intuition and following a recent MD study on one LTP (Rogers et al, 2023, Plos Comp Biol), it is safe to assume that the hydrophobic environment around the lipid is important for its stabilization in the lipid bilayer. Hence, reducing the number of hydrophobic contacts between the lipid and its environment could facilitate transport. A highly dynamic protein, by cycling between different conformations, could “stir” the bilayer, and hence decrease the number of contacts between the lipid and its environment favoring transport. We will clarify this point in the text.

*Examining how the LTPs impact membrane properties would offer insight into the functional relevance of such residues for lipid extraction. *

Indeed, our point above is connected to this one. We are performing simulations to compute hydrophobic contacts in bilayer as proposed in (Rogers et al, 2023, Plos Comp Biol).

The authors highlight that a bound lipid alters LTPs' conformational ensembles akin to "conformational selection" or "induced fit." How sensitive are these findings to the bound lipid species? Do LTPs with multiple known substrates exhibit an increasing diversity of holo conformations and are different conformations stabilized by different substrates? Would similar observations (Fig 3) be made with a lipid that is not known to be transferred by a given LTP? An interesting future direction would be to examine if lipid substrate specificity could be assessed by comparing conformational ensembles to that of a known substrate and/or by overlap with the apo ensemble.

We deem that the role of lipid specificity on LTP conformational plasticity is beyond the scope of the current work. While this topic is certainly worth future investigations, we must point out that (i) not all proteins bind/transport multiple lipids (at least according to current knowledge) and (ii) only few LTPs have been structurally characterized bound to different lipids (Osh4, Osh6, …). This limitation prevents a wide generalization, and we prefer not to speculate on this topic. So far, we have tested our approach for Osh4 bound to cholesterol or PI(4)P and found that indeed the protein exhibits different holo conformations (in agreement with the experimental data) when bound to different substrates. We have added a short comment on this topic in the Discussion section.

"____We foresee that future studies will focus on the functional consequence of such observation, and most notably to the characterization of the extent to which such conformational changes affect multiple steps of protein function, including membrane binding or lipid extraction and release, and whether these are further modulated when different lipids are being transported."

For LTPs to transfer lipids between membranes, transitions between apo and holo forms ought to occur when LTPs are membrane bound. How does membrane binding influence the conformational ensembles observed in solution? Does it promote conformational changes between apo- and holo-like structures, as suggested to regulate lipid uptake and release by previous studies of Osh/ORP, Ups/PRELI, and START family members? (For example: Miliara 2019 PMID 30850607, Watanabe 2015 PMID 26235513, Grabon 2017 PMID 28718450, Iaea 2015 PMID 26168008, Kudo 2008 PMID 18184806, Dong 2019 PMID 30783101) While answering these questions would require further computational effort, doing so will allow more accurate assessment of the role of conformational changes in LTP function.

We can’t unfortunately currently quantify how membrane binding influences the conformational ensembles observed in solution, as the slowdown in diffusion at the water-membrane interface makes this task computationally challenging (and certainly not feasible within the time framework of a review). We have so far tested two different proteins and have not succeeded in converging their conformational distribution when membrane-bound despite long MD simulations that lasted several months (even though the non-converged data indicate sampling of both “open” and “closed” conformations). Interestingly, our observations are in qualitative agreement with a recent study on CPTP (Rogers et al, PLOS Comp Biol, 2023), where membrane-bound CPTP is able to sample different conformations (“open” and “closed”) but not to transition between the two states in 300 ns-long MD simulations.

* The authors motivate the study with the *assumption* that a common molecular mechanism of LTP function exists. Yet LTPs have evolved diverse sequences, structures, and substrate preferences; thus there seems to be no a priori requirement (or even necessarily a benefit) for a single molecular mechanism. What evidence then supports this premise? While previous studies are limited to individual LTPs, when viewed altogether retrospectively, they suggest features that could be shared among LTPs. Synthesizing previous studies and more thoroughly referencing them (only 5 are cited in the intro on p. 3) would strengthen both the premise and findings of the manuscript. *

Indeed, despite having different structures, substrates and the ability to target distinct organelles, previous evidence on LTPs seem to suggest a potential role for protein conformational plasticity for function, e.g. for Osh/ORP (Jun Im et al, Nature 2005; Canagarajah et al, JMB 2008; Moser von Filseck et al, Nat Comm, 2015; Lipp et al, Nat Comm. 2019,...), StART (Arakane et al, PNAS, 1996; Feng et al, Biochemistry, 2000; Grabon et al, JBC, 2017; Khelashvili et al, eLife, 2019;...) and PITP domains (Tremblay et al, Archives of Biochemistry and Biophysics, 2005; Ryan et al, MBOC, 2007; …). Our simulations provide additional evidence in this direction and allow for generalizing these observations, allowing to draw parallelisms with “enzyme-like” or transporter-like” features that could be exploited for further design of testable hypotheses. We will rewrite our text to better contextualize/acknowledge previous findings and to clarify these points.

*The LTPs investigated are known to target distinct membranes. Should they then be expected to share structural or sequence-based features predictive of membrane binding interfaces, as motivates the analysis in Fig 1d, 1e, and S3? Or is it beneficial for LTPs to recognize membranes in different ways? *

Since membrane binding is membrane/organelle-specific, it is possible that residue’s diversity in membrane binding interfaces could indeed be beneficial for this diversity. We will add this comment as a potential explanation of our finding of a lack of conserved sequence-based features for membrane binding interfaces.

*

Minor comments:*

* 2 "making lipid transfer across the cytoplasm a potentially energetically favorable process": Is it meant that it is less energetically costly than transfer without a LTP? Why it would be energetically favorable is unclear (and would indicate that the LTP sequesters lipids away from membranes instead of transferring them between membranes). *

Yes, this is what we meant. We will rewrite this appropriately.

* 3 "The excellent agreement between the membrane interface determined from the simulations and the experimentally-proposed one available for... Osh6" is missing a citation. *

We have now added the relevant citation.

* The plots in Fig 1d and S3 are difficult to interpret. Bar plots, for example, would allow easier comparison and evaluation. Currently, it seems that most proteins individually exhibit some of the same trends observed among the whole set, counter to the conclusion on p 5. *

We will improve the presentation of our Figures.

* Negatively charged residues engage in a number of membrane interactions (Fig 1d and S3). What is a potential explanation for this unconventional observation? *

One possible interpretation is that negatively charged residues could interact with positively charged moieties (ethanolamine, choline) of PC and PE lipids.

* How much variance is captured by PC1, and how many PCs are needed to capture most of the variance in the conformations? *

PC1 explains 38 % of the total variance, by average, whereas PC2 accounts for 17 % of it. Therefore, PC1 and PC2 capture most of the variance in almost all cases.

We have also added this to the text:

"____We specifically focused on PC1 as it explains most of the variance in the dynamics (38% on average for all the proteins in our dataset, see Supplementary Table 2).____ "

We have computed this variance and we have added this analysis in Supplementary Information.

* Plots in Fig 3, especially panels c and d are difficult to see. Please make the panels larger (perhaps a 3 x 4 layout instead of 2 x 6 would work better). *

We will improve the presentation of our Figures.

* 8 "these conformational changes are localized in protein regions that interact with the lipid bilayer" is contradicted by the results in Fig 2b showing that all residues with large contributions to PC1 do not interact with the membrane and discussed on p 5. *

As discussed above, we don’t observe “correlation” between membrane binding and conformational plasticity, but we rather observe that membrane binding regions display high conformational plasticity (the opposite is not true). We will further clarify in the text.

*

8 "in the absence of bound lipids, it is able to sample multiple conformations" is not supported by the orange distributions in Fig 3d that appear unimodal. Is it instead meant that the apo form exhibits larger variance in cavity volume? *

Yes, this is what we meant. We’ll clarify.

*

Please clarify if the elastic network was constructed to maintain the holo or apo structures of each protein and if a bound lipid was used in the CG simulations. *

For membrane binding CG simulations, we used the apo structure and no bound lipid was used in the simulations. However, analogous simulations in the holo form (not shown) have essentially identical membrane binding interfaces.

*

Was *CHARMM* TIP3P used? *

Yes.

* Please clarify how membrane interacting residues were defined and how interaction frequency was calculated from the longest duration of interaction. *

We will add this explanation in the Methods. The method is identical to (Srinivasan et al, Faraday Discussion, 2021).

* Refs 16 and 45 refer to the same paper. *

Thanks, it is now corrected!

* Reviewer #2 (Significance (Required)): *

* General assessment: *

* The work aims to tackle a grand question regarding membrane homeostasis mechanisms-what are universal principles underlying LTP function-and offers initial insights; however, further evidence is needed to support the conclusions as written, and some key results require further investigation and explanation. *

*Advance and audience: *

*

By concurrently investigating the largest number of lipid transfer proteins to-date, the authors provide data invaluable for uncovering general mechanisms of non-vesicular lipid transport and advancing our understanding of membrane homeostasis mechanisms. By illuminating the wide-spread importance of conformational plasticity among lipid transfer proteins, the work presents a conceptual advance in our understanding of lipid transfer mechanisms and unifies previous studies. Because the manuscript emphasizes common biophysical principles and draws connections to enzyme biophysics, it ought to be of interest not only to membrane biologists but biochemists and molecular biologists more broadly.*

We thank Reviewer #2 for the very positive evaluation of the significance of our work and for the in-depth analysis provided that will certainly help improve the quality of our work.

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

*The article "Conformational dynamics of lipid transfer domains provide a general framework to decode their functional mechanism." by Sriraksha Srinivasan, Andrea DiLuca, Arun Peter, Charlotte Gehin, Museer Lone, Thorsten Hornemann, Giovanni D'Angelo and Stefano Vanni study the interaction of Lipid transport Domains with membranes. This is done mainly by molecular modelling but also with selected experimental validations. *

* Major comments: *

* - The key conclusions are generally well supported by the analysis. - The authors could however analyze in more details some aspects in which specific cases appear. For example, p3 "multiple binding and unbinding events, as shown by the minimum distance curves" does not give an entire description of the variability seen in Fig S1, e.g. LCN1 versus GM2A.*

We now discuss in more detail the variability seen in Fig. S1 and attribute it to different membrane binding affinities of the proteins in our dataset. We also discuss how this variability could reflect the diversity of organellar membranes to which these proteins bind in vivo.

"____Notably, the proteins in our dataset display distinct binding affinities, with some proteins showing very transient binding while others remain membrane-bound for most of the simulation trajectory (Fig. S1). This behavior could be, in part, attributed to the wide diversity of organellar membranes to which the LTDs in our dataset bind to in vivo, and to the comparative simplicity of our in silico model DOPC lipid bilayers."

• Later the "excellent agreement" for the data in Fig S2 is not quantified which does not allow the reader to know whether it better than would have been with other methods (SASA, OPM, DREAM). *

We have explicitly quantified this agreement by providing a direct comparison between the experimental results and our in silico assay, and we further compared it against two alternative methods: OPM and DREAMM. In detail, we have identified 12 experimentally-characterized spots suggested to be involved in membrane binding in our protein dataset (see shaded blue regions in Fig. S2). Of those 12, our method identifies all of them (100%), while DREAMM identifies 7 of them (58 %) and OPM 4 out of 8 (50 %), since of the 12 proteins we tested, only 7 are available in the OPM database. Overall, even if our approach is much noisier than the others, and thus suggesting multiple binding regions that are not currently supported by experimental observations, using physics-based methodologies appears to remain a preferable strategy to characterize the binding of peripheral proteins to lipid bilayers. Given the limited size of our dataset, we prefer not to make a direct comparison between our assay and OPM/DREAMM in the main text as this won't be representative of the various methodologies.

*p5 commenting on Fig2b the case of Osh6 that appears to disagree should probably be mentioned. *

We now discuss this case, and attribute to this disagreement to insufficient sampling for the peculiar case of Osh6:

"____One interesting exception in our database appears to be Osh6, where the experimentally determined membrane-binding region at the N-terminus (https://doi.org/10.1038/s41467-019-11780-y) is only marginally binding to the lipid bilayer in silico and it also appears to have limited contribution to PC1. However, our simulations are unable to sample the large conformational changes that the N-terminal lid of Osh6 has been proposed to undergo from its lipid-bound to its apo state, indicating that insufficient sampling could be the reason for this apparent discrepancy."

*

-The data and the methods are generally well presented allowing to be reproduced.

• The experiments adequately replicated with adequate statistical analysis. *

* Minor comments: *

* - When presenting the dataset the authors could probably detail a bit more the protocol undertaken to chose the cases. In particular it is unclear whether the chosen proteins have any membrane selectivity, which in principle could be affected by the choice of lipid used here.*

We have now added in Table 1 a column with a list of potential organelles the different LTPs have been shown to localize to (source: UniProt). As model membrane bilayer, we opted to use a pure DOPC bilayer, for both simplicity and to compare membrane binding in a uniform setting. We foresee that future studies investigating the membrane specificity of the various proteins will shed further light into the molecular mechanism of LTPs. Finally, we also indicate that our choice of proteins was mainly driven by the availability of lipid-bound structures in the protein data bank. We have added the following sentences in the main text:

"____Specifically, we selected all LTPs for which a crystallographic structure in complex with a lipid was available at the start of our project, plus two additional proteins (GM2A and LCN1) to increase the structural diversity of our dataset (Fig. 1a)"

and

"____Notably, the proteins in our dataset display distinct binding affinities, with some proteins showing very transient binding while other remain membrane-bound for most of the simulation trajectory (Fig. S1). This behavior could be, in part, attributed to the wide diversity of organellar membranes to which the LTDs in our dataset bind to in vivo, and to the comparative simplicity of our in silico model DOPC lipid bilayers."

*- The authors could probably give some indication of how much of the variance is explained by PC1 and comment briefly on the choice to ignore other PCs. *

PC1 explains 38 % of the total variance, on average. This means that PC1 has a large contribution to the variance, especially in comparison to the other PCs. For instance, PC2 only accounts for 17 % of the total variance. This is the reason we limited our discussion to PC1. We have added a table in supplementary Information quantifying the variance explained by PC1 and PC 2 and added the following sentence in the main text:

"____We specifically focused on PC1 as it explains most of the variance in the dynamics (38% on average for all the proteins in our dataset)____. "

* - When analyzing the residues involved in the interaction with the membrane the results could probably be compared with that of the systematic analysis performed recently: Tubiana, T., Sillitoe, I., Orengo, C., & Reuter, N. (2022). Dissecting peripheral protein-membrane interfaces. PLOS Computational Biology, 18(12), e1010346. *

We have added in the text a reference to the work by Tubiana et al and we have further stressed that our results agree with previous observations (including theirs). This includes the preference for Lys over Arg and the importance of protruding hydrophobes:

"____Concomitant analysis of all LTDs (Fig. 1d) indicates that the membrane binding interface of LTDs is enriched in the positively charged amino acid Lysine, as this amino acid is less membrane-disruptive than Arginine22, and aromatic/hydrophobic ones (Phe, Leu, Val, Ile). This confirms previous observations, as (i) binding of negatively charged lipids via positively charged residues and (ii) hydrophobic insertions are two of the main mechanisms involved in membrane binding by peripheral proteins22-27."

* - In the discussion on allostery/conformational selection might not be centered so much on enzymes. *

We thank the reviewer for this important observation. We have now included in the Discussion the following paragraph that provides additional references and discussion of membrane transporters and receptors.

"____Notably, the conformational plasticity we observe for LTPs is reminiscent of other, previously described, functional protein mechanisms, including enzyme dynamics during catalysis (____DOI: 10.1126/science.1066176____), the alternating-access model of membrane transporters (____https://doi.org/10.1038/nsmb.3179____) or GPCR dynamics (____https://doi.org/10.1021/acs.chemrev.6b00177____). In all these cases, protein dynamics is strongly coupled to ligand binding and protein function, be it for signaling, transport or enzymatic activity. Unlike for these fields, however, the contribution of structural and spectroscopic studies to uncover LTP dynamics remains quite limited, and our simulations provide an important contribution to fill this gap. We hope that our results will motivate researchers to increase efforts to experimentally quantify LTPs conformational plasticity, e.g. by structural determination of LTPs in different states (or bound to different lipids) or by single-molecule spectroscopy studies."

* Reviewer #3 (Significance (Required)): *

*

The article shows convincing results on the debated issue of the mechanism of lipid transport by lipid transfer proteins. *

First the study employs molecular modelling to allow a rather large test on 12 cases. The molecular dynamics experiments allow the authors to draw clear hypotheses on role of protein dynamics on the interaction with membranes and the effect on bound lipids on the modification of this dynamics.

*Then the authors use this knowledge to design experiments that largely confirm those hypotheses. The results should therefore be interesting for a large audience of biochemists and cell biologists interested in lipid transport in the cell. *

We thank Reviewer #3 for its very positive evaluation and contextualization of our work.