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

General comments

We thank all three reviewers for providing their thoughtful and insightful review comments of our manuscript. We appreciate that the reviewers recognized the significance and impact of our work - “Very little imaging has been done on CAR synapses and to our knowledge this is the first live cell imaging study describing CAR microclustsers” (Reviewer 2); “This is an evolving field and little is known to date. Hence, this study could represent an insightful and important advance to the field” (Reviewer 3). A broad audience from both basic and clinical research sides will be interested in this work: “_This study will have a broad audience. Both scientists that study basic T cell signaling as well as clinicians that use CAR Ts will be interested in this study” (_Reviewer 2); “Audience is to both basic immunologist and cancer biologists” (Reviewer 3).

Meanwhile, we understand that the reviewers have raised a few major and minor issues, which we attempted to address. Most importantly, as suggested by both reviewer 1 and 3, we performed new experiments showing that LAT is not required for microcluster formation of the 1st generation of CAR (new Fig 4 and EV5). This finding suggests that the CAR-independent signaling is due to the intrinsic CAR architecture, and is not dependent on the co-signaling domains of CD28 and 4-1BB.

With the successful solutions to other issues, we believe the manuscript has been significantly improved and is ready for publication. Below we will provide point-to-point responses to each reviewer’s comments.

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

The authors compare the TCR alone to a CAR that contains signaling modules from three receptors- TCR, CD28 and 41BB. The data quality if good and the experiments done are. The difference is quite clear, and I would even like to see a little more of the evidence related to failure of the TCR system.

We appreciate the general positive comment of this reviewer.

More specifically:

Su and colleagues show that a third generation CAR with TCR zeta, CD28 and 41BB signal transduction pathways can activate a T cell for microcluster formation and Gads/SLP-76 recruitment, but not IL-2 production, without LAT. This is surprising because LAT is generally considered, as is up held here, as an essential adapter protein for T cell activation. However, this is not a "fair" experiment as the CAR has sequences from TCR, and two co-stimulatory receptor- CD28 and 41BB. It would be important and very straight-forward to test first and second generation CARs to determine if LAT independence is a function of the CAR architecture itself, or the additional costimulatory sequences. If it turns out that a first generation CAR with only TCR sequences can trigger LAT independent clustering and SLP-76 recruitment then the comparison would be fair and no additional experiment would be needed to make the point that the CAR architecture is intrinsically LAT independent. If the CD28 and/or 41BB sequences are needed for LAT independence then the fair comparison would be to co-crosslink TCR, CD28 and 41BB (an inducible costimulator such that anti-CD27 might be substituted to have a constitutively expressed receptor with this similar motifs) should be cross-linked with the TCR to make this a fair comparison between the two architectures.

We agree with the reviewer that it is critical to make a “fair” comparison between TCR and CAR by testing the 1st generation CAR, which only contains the TCR/CD3z domain. Our new data showed that LAT is not required for microcluster and synapse formation of the 1st generation of CAR, in both Jurkat and primary T cells (new Fig 4 and EV5). This result is similar to our previously reported result from the 3rd generation CAR, although the 1st generation CAR induced less IL-2 production and CD69 expression in LAT null cells than the 3rd generation CAR did (new Fig 6). This suggests that the LAT-independent signaling is intrinsic to the CAR architecture, as the reviewer suggested. The co-signaling domains from CD28 and 4-1BB contribute to, but are not required for bypassing LAT to transduce the CAR signaling.

The authors may want to cite work from Vignali and colleagues that even the TCR has two signaling modules- the classical ZAP-70/LAT module that is responsible to IL-2 and a Vav/Notch dependent module that controls proliferation. Its not clear to me that the issue raised about distinct signaling by CARs is completely parallel to this, but its interesting that Vignali also associated the classical TCR signaling pathway as responsible for IL-2 with an alterive pathways that uses the same ITAMs to control distinct functions. See Guy CS, Vignali KM, Temirov J, Bettini ML, Overacre AE, Smeltzer M, Zhang H, Huppa JB, Tsai YH, Lobry C, Xie J, Dempsey PJ, Crawford HC, Aifantis I, Davis MM, Vignali DA. Distinct TCR signaling pathways drive proliferation and cytokine production in T cells. Nat Immunol. 2013;14(3):262-70.

We appreciate the reviewer’s mentioning this paper from Vignali’s group. It provides insights into understanding LAT-independent signaling in CAR T cells. We cited this paper and added a discussion about the mechanism of LAT-independent signaling.

I would be very interested to see a movie of the LAT deficient T cells interacting with the anti-CD3 coated bilayers in Figure 2A. Since OKT3 has a high affinity for CD3 and is coated on the surface at a density that should engage anti-CD3 I'm surprised there is no clustering even simply based on mass action. The result looks almost like a dominant negative effect of LAT deficiency on a high affinity extracellular interaction. It would be interesting to see how this interface evolves or if there is anti-adhesive behavior that emerges.

We now presented a movie showing the detailed process of LAT deficient GFP-CAR T cells landing on the bilayers coated with OKT3 (new Movie EV5), in which the bright field images delineate the locations of the cells, the OKT3 signal marks TCR, and the GFP signal marks CAR proteins on the plasma membranes. No TCR clusters (as indicated by OKT3) were formed during the landing process. We think the binding of bilayer-presented OKT3 to TCR is not sufficient to trigger TCR microclusters. However, TCR microclusters could form in LAT-deficient cells if OKT3 is presented by glass surface. This point is raised by reviewer 2. We added a discussion on the difference between bilayer and glass-presented OKT3 in inducing microcluster formation.

Reviewer #1 (Significance (Required)):

While it interesting that the CAR is LAT independent, its obvious that the signalling networks are different as the CAR has two sets of motifs that are absent in the TCR, so the experiments as presented are not that insightful about the specific nature of the differences that lead to the different outcomes. At present its not a particularly well controlled experiment as the third gen CAR is changing too many things in relation to the TCR for the experiment to be interpreted. It would be easy to address this is a revised manuscript. To publish as is the discussion would need to acknowledge these limitations. The work is preliminary as science, but it might be useful to T cell engineering field to have this information as a preliminary report, which might be an argument for adding discussion of limitations, but going forward without more detailed analysis of mechanism.

This is an excellent point and we have addressed it. See our response above on the new data of the 1st generation CAR.

Reviewer #2 (Evidence, reproducibility and clarity):

Summary:

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

In this study, the authors have interrogated CAR signaling by imaging CD19-CAR microclusters as well as T cell signaling molecules recruited to CAR microclusters. They report differences spatial assembly between CAR and TCR microclusters that form on a lipid bilayer containing ligand. They also report that LAT is not required for CAR microcluster formation, recruitment of downstream signaling molecules or IL-2 production in Jurkat cells, while in primary T cells IL-2 production by CARs show more of a LAT dependence. From these observations, they conclude that CAR T cells have a rewired signaling pathway as compared to T cells that signal through the TCR.

Major comments:

• Are the key conclusions convincing?

The conclusions made by the authors about CAR microclusters are convincing. However, the conclusion that there is a "rewired signaling network" different from TCR microclusters needs to be more convincingly demonstrated in side-by-side comparisons of TCR and CAR microclusters and synapses.

1. One of the key conclusions in this study is that CAR microclusters form in the absence of LAT, but TCR microclusters require LAT (in JCam2.5 cells in Fig. 2 and primary T cells in Fig. 4B). The requirement of LAT for formation of TCR microclusters is surprising, given multiple reports (one of which the authors have cited) that TCRz and ZAP70 clusters form normally in the absence of LAT (pZAP microclusters form normally in JCam2.5 cells Barda-Saad Nature Immunology 2005 Figure 1; TCRz clusters form normally in LAT CRISPR KO Jurkat cells Yi et al., Nature Communications, 2019 Figure 5). The authors should carefully evaluate TCRz and ZAP70 clusters (that form upstream of LAT) in their assays.

We thank the reviewer for raising this excellent point. LAT-independent TCR clusters were reported in the two papers mentioned by the reviewer, which we think is convincing. However, there is a key difference in the experimental settings between these two papers and ours. We use supported lipid bilayer to present MOBILE TCR-activating antibody to activate T cells, whereas these two papers used IMMOBILE TCR-activating antibody attached to the cover glass. We reasoned that the mobile surface of supported lipid bilayer more closely mimics the antigen-presenting cell surface where antigens are mobile on the membrane. We added a new discussion about the difference between supported lipid bilayer and cover glass-based activation.

We agree with the reviewer on the careful evaluation of TCR and ZAP70 clusters. We had showed the data of TCR clusters as marked by TCR-interacting OKT3 (Fig 3A). We performed new experiments on ZAP70 clusters (new Fig EV3). Our data suggest that, similar to TCR clusters, ZAP70 clusters are not formed in LAT-deficient T cells, if activated by OKT3, but are formed if activated by CD19.

1. The authors make major conclusions about LAT dependence and independence of TCR and CAR microclusters respectively, by using JCam2.5 Jurkat cells and CRISPR/Cas9 edited primary cells. Of relevance to this conclusion, differences in the phosphorylation status of ZAP70 and SLP76 have been described between JCam2.5 cells lacking LAT (in which LAT was found to be deleted by gamma radiation) and J.LAT cells (in which LAT was specifically deleted by CRISPR/Cas9 in Lo et al Nature Immunology 2018). Of importance, pZAP and pSLP76 appeared fairly intact in J.LAT cells, but absent in JCam2.5 cells (Lo et al., Nat Immunol. 2018, Supp Fig 2). Therefore, the authors should evaluate TCRz, ZAP70, Gads and SLP76 in TCR and CAR microclusters in J.LAT cells. This may partly explain the discrepancy in LAT requirement for IL-2 production in JCam2.5 cells and primary cells with LAT CRISPRed out.

Jcam2.5 is a classical well-characterized LAT-deficient cell line that has been continuously used in the T cell signaling field (Barda-Saad Nature Immunology 2005, Rouquette-Jazdanian A, Mol. Cell, 2012; Balagopalan L, J Imm. 2013; Carpier J, J Exp Med, 2018; Zucchetti A, Nat. Comm. 2019). We agreed with the concern that the reviewer raised on the absence of pZAP70 and pSLP76 in JCam2.5 cells. As the reviewer suggested, we obtained J.LAT, which is LAT null but has intact pZAP70 and pSLP76. We introduced CAR into J.LAT and the wild-type control and performed the clustering assay as we did for Jcam2.5. Our results showed that, similar to Jcam2.5, CAR forms robust microclusters in J.LAT cells (new Fig EV2). More importantly, we presented data confirming the LAT-independent CAR clustering, SLP76 phosphorylation, and IL-2 production in human primary T cells (Fig 7). Therefore, the data from three independent cell sources support our conclusion on LAT-independent CAR signal transduction.

1. Since the authors are reporting differences between CAR synapses and TCR synapses, the authors should show side by side comparison of CAR and TCR synapses in Figure 1F.

We focused on characterizing CAR synapse in this manuscript and did not make any conclusion on the difference between TCR and CAR synapse. We are cautious about comparing CAR synapse to TCR synapse for technical reasons: it is critical to use antigen-specific TCRs (e.g. mouse OTI as a common model) to study the TCR synapse pattern so that the study will be physiologically relevant. However, we use human T cell line and human primary T cells for the CAR study. The technical barrier to introduce an antigen-specific TCR complex into these cells, and to activate these cells by purified peptide-MHC complex, is very high. And the result is interesting, but beyond the scope of the current work.

1. The authors should evaluate Gads microcluster formation in response to TCR stimulation via OKT3 (in Figure 4A). Given that it has been reported that TCRz, Grb2 and c-Cbl are recruited to microclusters in Jurkat cells lacking LAT by CRISPR deletion (Yi et al., Nature Communications, 2019), it is important to establish the differences between TCR microclusters and CAR microclusters in side by side comparisons in their assay system.

As the reviewer suggested, we evaluated Gads microcluster formation with TCR stimulation and found that Gads did not form microclusters in LAT-deficient cells (new Fig 5A). Because we only made conclusions on the Gads-SLP76 pathway, we think investigating Grb2 and c-Cbl microcluster, though interesting, is beyond the scope of this manuscript.

1. Similar to the comment about Gads above, the authors should evaluate pSLP76 microcluster formation in response to TCR stimulation via OKT3 in primary T cells lacking LAT in Figure 4C, i.e. side by side comparisons of pSLP76 in TCR and CAR synapses (with and without LAT) should be shown.

We totally agree and performed new experiment on pSLP76 in human primary T cells. Our data suggested that, similar to Jurkat, pSLP76 microclusters remain intact in LAT null primary cells (new Fig 7D and 7E).

• Should the authors qualify some of their claims as preliminary or speculative, or remove them altogether?

1. The data shown in Figure 3C shows a reduction in conjugate formation from 80% (WT) to 30% (LAT -). This is a severe reduction and does not support the authors' claim in the corresponding Figure legend that "LAT is dispensable for cell conjugate formation between Jurkat T cells expressing CAR and Raji B cells" and the Abstract that "LAT.....is not required for....immunological synapse formation". Statistical analysis for variance should be shown here.

We agree with the reviewer’s judgement. This cell conjugation analysis was performed using Jcam2.5 cells. As pointed by the reviewer, Jcam2.5 has additional defects in ZAP70 and SLP76 in addition to the lack of LAT. Therefore, we performed the same analysis again using J.LAT cells, which was recommended by the reviewer. Our new data showed that J.LAT cells form conjugates with Raji B cells in a similar rate as the wild-type cells do, as evaluated by statistical analysis (new Fig 6A). Therefore, we think these new data support the claim that LAT is dispensable for cell conjugate formation.

1. In a similar vein, based on data from Movie S5 (where in a single cell, CAR microclusters translocate from cell periphery to center), and Figure 3C where (as described above in point 1) conjugate formation appears to be severely reduced, the authors conclude in the Results and Abstract that "LAT....is not required for actin remodeling following CAR activation". This conclusion is not supported by the data and the authors should remove this claim. Alternatively, actin polymerization in CAR expressing cells (that are LAT sufficient and deficient) can be easily evaluated using phalloidin or F-Tractin.

As suggested by the reviewer, we evaluated actin polymerization in TCR or CAR stimulated cells using a filamentous actin reporter F-tractin. Our data showed that LAT is required for TCR-induced but not CAR-induced actin polymerization (new Fig 5C). Therefore, our results support the claim that LAT is not required for actin remodeling following CAR activation.

• Would additional experiments be essential to support the claims of the paper?<br> Request additional experiments only where necessary for the paper as it is, and do not ask authors to open new lines of experimentation.

Yes. Please see major comments above.

• 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.

Yes. It should take 3 months to complete these experiments, since reagents and experimental systems to do these experiments already exist.

• Are the data and the methods presented in such a way that they can be reproduced?<br> Yes. Methods are clearly explained.

We appreciate the reviewer’s recognition of the clarity of the methods part.

• Are the experiments adequately replicated and statistical analysis adequate?

There is no statistical analysis to evaluate differences between samples in Figures 3 and 4. These must be included.

We now added statistical analysis in Fig 5B and 6A (old figure 3 and 4).

Minor comments:

• Specific experimental issues that are easily addressable.

Please see Major Comments above. We believe that the recommended experiments are not difficult to execute since reagents exist and experimental systems are already set up.

• Are prior studies referenced appropriately?

Authors reference 13 and 14 for the following sentence in Results section 2: "Deletion or mutation of LAT impairs formation of T cell microclusters". However, in Reference 14 Barda-Saad et al., actually show that pZAP clusters are intact in JCam2.5 cells lacking LAT. Perhaps authors should clarify that LAT (and downstream signaling molecule) microclusters are impaired when LAT is deleted or mutated.

As the reviewer suggested, we now clarified that clustering of LAT downstream binding partners is impaired when citing reference (Barda-Saad et al).

• Are the text and figures clear and accurate?

Yes. But would be helpful if authors specify what "control" is in Fig. 3B and C. In Figure 3B it is lipid bilayers without CD19, while in 3C it is K562 cells that do not express CD19.

We now specified “control” in the figure.

• Do you have suggestions that would help the authors improve the presentation of their data and conclusions?

Would be helpful if authors specify in every Figure or at least Figure legend the experimental bilayer system/ligand used, since they use both OKT3 and CD19 as ligands in the paper.

We now specified the ligand in the figure or legend.

Reviewer #2 (Significance):

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

If CAR microclusters and synapses are appropriately compared in a side by side comparison with TCR microclusters and synapses (as described in comments above), this study will be a conceptual advance in the field of CAR signaling. CAR microclusters have not been studied previously.

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

Very little imaging has been done on CAR synapses and to our knowledge this is the first live cell imaging study describing CAR microclusters.

We appreciate this reviewer’s comment on our work as a conceptual advance in understanding CAR signaling.

• State what audience might be interested in and influenced by the reported findings.<br> This study will have a broad audience. Both scientists that study basic T cell signaling as well as clinicians that use CAR Ts will be interested in this study.

We appreciate this reviewer’s recognition of the broad audience of this manuscript.

• 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.

T cell signaling and imaging of proximal T cell signaling responses.

Reviewer #3 (Evidence, reproducibility and clarity):

This manuscript by Dong and colleagues characterizes the molecular requirements and consequences of engaging a third-generation chimeric antigen receptor (CAR) directed to CD19. Utilizing a biological system of JCaM2.5, a Jurkat T cell mutant with dramatically low levels of LAT, expressing a CAR directed to CD19 fused to the cytoplasmic tails of CD28, 4-1BB and CD3z that is activated by CD19/ICAM1 reconstituted lipid bilayers, the authors demonstrate LAT is not required for microcluster formation, immunologic synapse formation or recruitment of GADS and pSLP76 to the plasma membrane. In contrast, LAT was required for anti-CD3 mediated microcluster formation and pSLP76 recruitment to the plasma membrane. However, LAT does appear to contribute to efficient synapse formation, PIP2 hydrolysis and IL-2 secretion when CAR+ JCaM2.5 or primary T cells are presented with Raji B cells, respectively. These data provide intriguing insights into the molecular requirements for third-generation CAR-T cell functions. The authors have developed quite a nice system to understand the molecular contributions for CAR-T function. A few suggestions are provided here to further enhance the accuracy and significance of the findings:

1. The authors can address whether the LAT-independent effects are due to the attributes of third generation CAR-Ts with inclusion of CD28 and 4-1BB cytoplasmic domains or whether these differences are intrinsic to all CAR-Ts (e.g., first and second generation CARs).

This is an excellent point. We have included new data showing LAT-independent cluster formation of the 1st generation CAR in both Jurkat and primary T cells (new Fig 4 and EV5). Therefore, we favor the second possibility as pointed by the reviewer that LAT-independent effects are intrinsic to CAR architecture.

1. Since a first-generation CAR-T forms non-conventional synapses (Davenport, et al., PNAS 2018), the authors should consider more detailed kinetic analysis to understand the formation and dissolution of the constituents of the synapse with their third generation CAR. This should include measurements of the duration of microcluster and synapse formation as well as further analysis of c- and p-SMAC constituents (e.g., LFA-1, TALIN, LCK and pSLP76) over time.

We agree with the reviewer on a more detailed characterization of the CAR synapse. We measured the duration of the unstable CAR synapse and time from cell landing to the start of retrograde flow (new Fig 2C). We also determined the localization of CD45, a marker for d-SMAC (new Fig 2D). We found that the formation of dSMAC is also not common in CAR T synapse, strengthening our conclusion that CAR forms non-typical immunological synapse.

1. The authors utilize two different activation platforms. While using CD19/ICAM1 reconstituted bilayers, CAR+ JCaM2.5 or CAR+ primary T cells demonstrate no differences compared to wildtype JCaM2.5 cells in the parameters studied. However, when using Raji B cells, the CAR+ JCaM2.5 cells or CAR+ primary T cells demonstrate a more intermediate phenotype with respect to cell conjugate formation (Figure 3C) and IL-2 production (Figure 4D). The authors should analyze whether the differences attributed to the different outcomes may be due to the stimulation mode. For example, is c-SMAC assembly and GADS or pSLP76 recruitment to the plasma membrane still LAT-independent when activated with Raji B cells?

As the reviewer suggested, we examined c-SMAC assembly in Raji B cells conjugated with CAR T cells. We found that the majority of CAR do not form cSMAC (new Fig EV4), which is consistent with the result from the bilayer activation system. Since both Gads and SLP76 are cytosolic proteins, they keep largely in the cytosolic pool which obscures their recruitment and clustering on the plasma membrane when imaged by confocal microscopy at the cross-section of cell-cell synapse.

1. The authors should consider whether CAR expression level affects their observations. For example, do lower levels of CAR expression make the system LAT-dependent? Further, what is the level of the CAR relative to endogenous TCR expression on their primary T cells.

We agree with the reviewer that it is informative to determine if LAT-independent signaling is dose dependent. We tried to measure the CAR concentration relative to the endogenous TCR/CD3z. By western blot using two different antibodies against CD3z, we detected TCR/CD3z expression, but found no bands corresponding to CAR. We believe this reflects a low expression of CAR in our system, which is confirmed by FACS. The general low expression of CAR makes it challenging to sort an even lower CAR-expressing population. Therefore, we sought alternative ways to determine the dose-dependence; we titrated the CD19 concentrations on the bilayer. As shown in the new Figure EV1, CAR formed microclusters similarly in the wild-type versus LAT-deficient cells in a wide range of CD19 concentration. Therefore, we conclude that the LAT-independent cluster formation is robust at low antigen density as well.

Minor comment:

1. Since JCaM2.5 has differences when compared to the parental Jurkat E6.1 T cell line, the authors should utilize JCaM2.5 reconstituted with wildtype LAT as a comparator.<br> Agreeing with this reviewer, we recognized that Jcam2.5 was generated by mutagenesis which may result in protein expression difference for genes besides Lat. As suggested by reviewer1, we used J.LAT, a genuine LAT knockout cell line that is generated by CRISPR-mediated gene targeting, to perform the clustering assay (new Fig EV2). Our results showed that, similar to Jcam2.5, CAR but not the TCR formed microclusters in J.LAT cells.

Reviewer #3 (Significance):

The mechanism(s) by which CAR-Ts function is of high significance from both scientific and clinical viewpoints. From a scientific viewpoint, it provides important basic mechanistic information of how T cells are being activated to kill tumor cells. By understanding the molecular requirements, additional generations of CARs can be designed to provide greater efficacy, overcome resistance and possibly less toxicity.

This is an evolving field and little is known to date. Hence, this study could represent an insightful and important advance to the field.

Audience is to both basic immunologist and cancer biologists.

We appreciate this reviewer’s comments on the high significance of our work to the field of both basic immunology and clinical application.

My expertise is in T cell signaling, T cell biology and immunotherapy.

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 Dong and colleagues characterizes the molecular requirements and consequences of engaging a third-generation chimeric antigen receptor (CAR) directed to CD19. Utilizing a biological system of JCaM2.5, a Jurkat T cell mutant with dramatically low levels of LAT, expressing a CAR directed to CD19 fused to the cytoplasmic tails of CD28, 4-1BB and CD3 that is activated by CD19/ICAM1 reconstituted lipid bilayers, the authors demonstrate LAT is not required for microcluster formation, immunologic synapse formation or recruitment of GADS and pSLP76 to the plasma membrane. In contrast, LAT was required for anti-CD3 mediated microcluster formation and pSLP76 recruitment to the plasma membrane. However, LAT does appear to contribute to efficient synapse formation, PIP2 hydrolysis and IL-2 secretion when CAR+ JCaM2.5 or primary T cells are presented with Raji B cells, respectively. These data provide intriguing insights into the molecular requirements for third-generation CAR-T cell functions.

The authors have developed quite a nice system to understand the molecular contributions for CAR-T function. A few suggestions are provided here to further enhance the accuracy and significance of the findings:

1. The authors can address whether the LAT-independent effects are due to the attributes of third generation CAR-Ts with inclusion of CD28 and 4-1BB cytoplasmic domains or whether these differences are intrinsic to all CAR-Ts (e.g., first and second generation CARs).
2. Since a first-generation CAR-T forms non-conventional synapses (Davenport, et al., PNAS 2018), the authors should consider more detailed kinetic analysis to understand the formation and dissolution of the constituents of the synapse with their third generation CAR. This should include measurements of the duration of microcluster and synapse formation as well as further analysis of c- and p-SMAC constituents (e.g., LFA-1, TALIN, LCK and pSLP76) over time.
3. The authors utilize two different activation platforms. While using CD19/ICAM1 reconstituted bilayers, CAR+ JCaM2.5 or CAR+ primary T cells demonstrate no differences compared to wildtype JCaM2.5 cells in the parameters studied. However, when using Raji B cells, the CAR+ JCaM2.5 cells or CAR+ primary T cells demonstrate a more intermediate phenotype with respect to cell conjugate formation (Figure 3C) and IL-2 production (Figure 4D). The authors should analyze whether the differences attributed to the different outcomes may be due to the stimulation mode. For example, is c-SMAC assembly and GADS or pSLP76 recruitment to the plasma membrane still LAT-independent when activated with Raji B cells?
4. The authors should consider whether CAR expression level affects their observations. For example, do lower levels of CAR expression make the system LAT-dependent? Further, what is the level of the CAR relative to endogenous TCR expression on their primary T cells.

Minor comment:

1. Since JCaM2.5 has differences when compared to the parental Jurkat E6.1 T cell line, the authors should utilize JCaM2.5 reconstituted with wildtype LAT as a comparator.

Significance (Required)

The mechanism(s) by which CAR-Ts function is of high significance from both scientific and clinical viewpoints. From a scientific viewpoint, it provides important basic mechanistic information of how T cells are being activated to kill tumor cells. By understanding the molecular requirements, additional generations of CARs can be designed to provide greater efficacy, overcome resistance and possibly less toxicity.

This is an evolving field and little is known to date. Hence, this study could represent an insightful and important advance to the field.

Audience is to both basic immunologist and cancer biologists.

My expertise is in T cell signaling, T cell biology and immunotherapy.

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).

In this study, the authors have interrogated CAR signaling by imaging CD19-CAR microclusters as well as T cell signaling molecules recruited to CAR microclusters. They report differences spatial assembly between CAR and TCR microclusters that form on a lipid bilayer containing ligand. They also report that LAT is not required for CAR microcluster formation, recruitment of downstream signaling molecules or IL-2 production in Jurkat cells, while in primary T cells IL-2 production by CARs show more of a LAT dependence. From these observations, they conclude that CAR T cells have a rewired signaling pathway as compared to T cells that signal through the TCR.

Major comments:

Are the key conclusions convincing?

The conclusions made by the authors about CAR microclusters are convincing. However, the conclusion that there is a "rewired signaling network" different from TCR microclusters needs to be more convincingly demonstrated in side-by-side comparisons of TCR and CAR microclusters and synapses.

1. One of the key conclusions in this study is that CAR microclusters form in the absence of LAT, but TCR microclusters require LAT (in JCam2.5 cells in Fig. 2 and primary T cells in Fig. 4B). The requirement of LAT for formation of TCR microclusters is surprising, given multiple reports (one of which the authors have cited) that TCR and ZAP70 clusters form normally in the absence of LAT (pZAP microclusters form normally in JCam2.5 cells Barda-Saad Nature Immunology 2005 Figure 1; TCR clusters form normally in LAT CRISPR KO Jurkat cells Yi et al., Nature Communications, 2019 Figure 5). The authors should carefully evaluate TCR and ZAP70 clusters (that form upstream of LAT) in their assays.
2. The authors make major conclusions about LAT dependence and independence of TCR and CAR microclusters respectively, by using JCam2.5 Jurkat cells and CRISPR/Cas9 edited primary cells. Of relevance to this conclusion, differences in the phosphorylation status of ZAP70 and SLP76 have been described between JCam2.5 cells lacking LAT (in which LAT was found to be deleted by gamma radiation) and J.LAT cells (in which LAT was specifically deleted by CRISPR/Cas9 in Lo et al Nature Immunology 2018). Of importance, pZAP and pSLP76 appeared fairly intact in J.LAT cells, but absent in JCam2.5 cells (Lo et al., Nat Immunol. 2018, Supp Fig 2). Therefore, the authors should evaluate TCR, ZAP70, Gads and SLP76 in TCR and CAR microclusters in J.LAT cells. This may partly explain the discrepancy in LAT requirement for IL-2 production in JCam2.5 cells and primary cells with LAT CRISPRed out.
3. Since the authors are reporting differences between CAR synapses and TCR synapses, the authors should show side by side comparison of CAR and TCR synapses in Figure 1F.
4. The authors should evaluate Gads microcluster formation in response to TCR stimulation via OKT3 (in Figure 4A). Given that it has been reported that TCR, Grb2 and c-Cbl are recruited to microclusters in Jurkat cells lacking LAT by CRISPR deletion (Yi et al., Nature Communications, 2019), it is important to establish the differences between TCR microclusters and CAR microclusters in side by side comparisons in their assay system.
5. Similar to the comment about Gads above, the authors should evaluate pSLP76 microcluster formation in response to TCR stimulation via OKT3 in primary T cells lacking LAT in Figure 4C, i.e. side by side comparisons of pSLP76 in TCR and CAR synapses (with and without LAT) should be shown.

Should the authors qualify some of their claims as preliminary or speculative, or remove them altogether?

1. The data shown in Figure 3C shows a reduction in conjugate formation from 80% (WT) to 30% (LAT -). This is a severe reduction and does not support the authors' claim in the corresponding Figure legend that "LAT is dispensable for cell conjugate formation between Jurkat T cells expressing CAR and Raji B cells" and the Abstract that "LAT.....is not required for....immunological synapse formation". Statistical analysis for variance should be shown here.
2. In a similar vein, based on data from Movie S5 (where in a single cell, CAR microclusters translocate from cell periphery to center), and Figure 3C where (as described above in point 1) conjugate formation appears to be severely reduced, the authors conclude in the Results and Abstract that "LAT....is not required for actin remodeling following CAR activation". This conclusion is not supported by the data and the authors should remove this claim. Alternatively, actin polymerization in CAR expressing cells (that are LAT sufficient and deficient) can be easily evaluated using phalloidin or F-Tractin.

Would additional experiments be essential to support the claims of the paper?<br> Request additional experiments only where necessary for the paper as it is, and do not ask authors to open new lines of experimentation.

Yes. Please see major comments above.

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.

Yes. It should take 3 months to complete these experiments, since reagents and experimental systems to do these experiments already exist.

Are the data and the methods presented in such a way that they can be reproduced?

Yes. Methods are clearly explained.

Are the experiments adequately replicated and statistical analysis adequate?

There is no statistical analysis to evaluate differences between samples in Figures 3 and 4. These must be included.

Minor comments:

Specific experimental issues that are easily addressable.

Please see Major Comments above. We believe that the recommended experiments are not difficult to execute since reagents exist and experimental systems are already set up.

Are prior studies referenced appropriately?

Authors reference 13 and 14 for the following sentence in Results section 2: "Deletion or mutation of LAT impairs formation of T cell microclusters". However, in Reference 14 Barda-Saad et al., actually show that pZAP clusters are intact in JCam2.5 cells lacking LAT. Perhaps authors should clarify that LAT (and downstream signaling molecule) microclusters are impaired when LAT is deleted or mutated.

Are the text and figures clear and accurate?

Yes. But would be helpful if authors specify what "control" is in Fig. 3B and C. In Figure 3B it is lipid bilayers without CD19, while in 3C it is K562 cells that do not express CD19.

Do you have suggestions that would help the authors improve the presentation of their data and conclusions?<br> Would be helpful if authors specify in every Figure or at least Figure legend the experimental bilayer system/ligand used, since they use both OKT3 and CD19 as ligands in the paper.

Significance (Required)

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

If CAR microclusters and synapses are appropriately compared in a side by side comparison with TCR microclusters and synapses (as described in comments above), this study will be a conceptual advance in the field of CAR signaling. CAR microclusters have not been studied previously.

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

Very little imaging has been done on CAR synapses and to our knowledge this is the first live cell imaging study describing CAR microclusters.

State what audience might be interested in and influenced by the reported findings.

This study will have a broad audience. Both scientists that study basic T cell signaling as well as clinicians that use CAR Ts will be interested in this study.

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.

T cell signaling and imaging of proximal T cell signaling responses.

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 compare the TCR alone to a CAR that contains signaling modules from three receptors- TCR, CD28 and 41BB. The data quality if good and the experiments done are. The difference is quite clear, and I would even like to see a little more of the evidence related to failure of the TCR system.

More specifically:

Su and colleagues show that a third generation CAR with TCR zeta, CD28 and 41BB signal transduction pathways can activate a T cell for microcluster formation and Gads/SLP-76 recruitment, but not IL-2 production, without LAT. This is surprising because LAT is generally considered, as is up held here, as an essential adapter protein for T cell activation. However, this is not a "fair" experiment as the CAR has sequences from TCR, and two co-stimulatory receptor- CD28 and 41BB. It would be important and very straight-forward to test first and second generation CARs to determine if LAT independence is a function of the CAR architecture itself, or the additional costimulatory sequences. If it turns out that a first generation CAR with only TCR sequences can trigger LAT independent clustering and SLP-76 recruitment then the comparison would be fair and no additional experiment would be needed to make the point that the CAR architecture is intrinsically LAT independent. If the CD28 and/or 41BB sequences are needed for LAT independence then the fair comparison would be to co-crosslink TCR, CD28 and 41BB (an inducible costimulator such that anti-CD27 might be substituted to have a constitutively expressed receptor with this similar motifs) should be cross-linked with the TCR to make this a fair comparison between the two architectures.

The authors may want to cite work from Vignali and colleagues that even the TCR has two signaling modules- the classical ZAP-70/LAT module that is responsible to IL-2 and a Vav/Notch dependent module that controls proliferation. Its not clear to me that the issue raised about distinct signaling by CARs is completely parallel to this, but its interesting that Vignali also associated the classical TCR signaling pathway as responsible for IL-2 with an alterive pathways that uses the same ITAMs to control distinct functions. See Guy CS, Vignali KM, Temirov J, Bettini ML, Overacre AE, Smeltzer M, Zhang H, Huppa JB, Tsai YH, Lobry C, Xie J, Dempsey PJ, Crawford HC, Aifantis I, Davis MM, Vignali DA. Distinct TCR signaling pathways drive proliferation and cytokine production in T cells. Nat Immunol. 2013;14(3):262-70.

I would be very interested to see a movie of the LAT deficient T cells interacting with the anti-CD3 coated bilayers in Figure 2A. Since OKT3 has a high affinity for CD3 and is coated on the suface at a density that should engage anti-CD3 I'm surprised there is no clustering even simply based on mass action. The result looks almost like a dominant negative effect of LAT deficiency on a high affinity extracellular interaction. It would be interesting to see how this interface evolves or if there is anti-adhesive behavior that emerges.

Significance

While it interesting that the CAR is LAT independent, its obvious that the signalling networks are different as the CAR has two sets of motifs that are absent in the TCR, so the experiments as presented are not that insightful about the specific nature of the differences that lead to the different outcomes. At present its not a particularly well controlled experiment as the third gen CAR is changing too many things in relation to the TCR for the experiment to be interpreted. It would be easy to address this is a revised manuscript. To publish as is the discussion would need to acknowledge these limitations. The work is preliminary as science, but it might be useful to T cell engineering field to have this information as a preliminary report, which might be an argument for adding discussion of limitations, but going forward without more detailed analysis of mechanism.

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

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

Rebuttal to reviewers ReviewCommons manuscript # RC-2020-00281

We would like to thank the reviewers and editors of Review Commons for evaluating our manuscript entitled “Transcriptional comparison of Testicular Adrenal Rest Tumors with fetal and adult tissues” and providing their valuable comments. We have listed the reviewers’ comments along with our response and amendments below.

Board Advice on initial submission:

This seems to be a study mainly relevant to the field of Testicular Adrenal Rest Tumors (TART). It presents the first RNAseq profiling of these tumors in multiple human samples at different stages. This has the potential to advance knowledge in this particular field. It would be less interesting to researchers interested in tissue spatial transcriptomics in general, since the experimental and computational tools are quite standard, but the findings may be important to the TART field.

Response: Indeed, this is the first study using transcriptomics to characterize Testicular Adrenal Rest Tumors, a frequent occurrence in patients with Congenital Adrenal Hyperplasia. It is also the first to find that the reported adrenal and testicular features of these tumors can be found in a single cell. We therefore believe this study is not only of interest to those working in the TART field, but also in development, endocrinology and andrology in general.

Comments Reviewer #1:

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

**Summary:**

The manuscript by Schroder M., et al describes the whole transcriptome of testicular adrenal rest tumors (TART) and shows that TART tissue is characteristically similar to adult adrenal and testicular rather than fetal adrenal and testicular tissues. The authors propose that their previous claim that TART is derived from an undifferentiated pluripotent progenitor is likely untrue and claim that TART likely originates from a mature cell type with both adrenal and testicular characteristics. The authors describe a unique cell type most similar to the adult adrenal, but with variable testis-specific gene expression patterns. The finding of overexpressed genes associated with ECM remodeling is interesting and may provide insight into the natural history of these tumors. A strength of the study is the number of tissue samples since surgery for these rare tumors is usually not performed.

**Major Comments:**

• The key conclusions are mostly based on RNA studies, thus their claims are preliminary.

Response: We agree that a major part of our conclusions is based on RNA studies. Although indeed primarily based on transcriptomics, this claim is, in our opinion, not preliminary as the identity of TART cells can definitively be deduced from their expression profile. Our second key conclusion, i.e. that TART cells comprise both adrenal and testicular features within the same, unique, TART specific cell, is based on immunohistochemistry of adrenal and testis-specific enzymes.

In Figure 1/Result p. 3: Authors claim that there were no exclusive HSD17B3 staining cells without CYP11B1, however Figure 1 looks like there are exclusively green (HSD17B) areas (especially TART3). The authors need to address this. It appears as if there are mature Leydig cells. This is important because the presence of Leydig cells would affect the interpretation of the findings.

Response: We do understand the concern of the reviewer. Aspecific background staining for HSD17B3 in TART samples complicated the differentiation between specific and background staining. This can be seen when comparing the staining in HSD17B3-positive (Leydig) cells with the background staining in non-Leydig cells in testis tissue and in a portion of TART cells. In TART, we found that cells with high intensity, specific HSD17B3 staining all also showed CYP11B1 staining, but not vice-versa. However, we do acknowledge that due to this -most likely background- staining, the occurrence of mature Leydig cells in TART cannot be completely excluded based on our results.

Therefore, we have tried to be more careful in our claims in the results section (page 3; TART cells express adrenal- and Leydig cell-specific steroidogenic enzymes paragraph) and we have addressed this in the discussion section (page 5/6):

High background staining for HSD17B3 complicated the differentiation between specific and background staining. For some cells this exclusive HSD17B3 staining might have been specific and therefore, despite that most HSD17B3-positive cells were positive for CYP11B1, the absence of mature Leydig cells in TART could not be guaranteed by these results.

Discussion: authors state that based on their previous observations that fetal Leydig cells have both adrenal and testis developmental potential. It was speculated that TART might have been derived from a totipotent progenitor cell type, but the current study shows that these tumors lack similarities with fetal tissues. Thus, the authors claim that these tumors are not derived from the transdifferentiation of pluripotent cells. However what is the origin of this mature distinct cell type? Is it not possible that this distinctive cell type is derived from a common progenitor since the testis and adrenal gland are derived from the same adrenogonadal primordium? Lack of similarities with fetal tissues at this late stage of development does not necessarily rule out a common progenitor origin.

Response: In this study, we compared the TART transcriptome with fetal tissues, as we hypothesized these might be similar considering the likely progenitor origin of TART cells. However, this was not the case, and we showed that the transcriptomic profile of TART resembles the transcriptomic profile of mature cell types, rather than their fetal counterparts. Therefore, we conclude that the hypothesis that TART arises from progenitor cells is not supported by our data. The reviewer is correct that we did not prove that it is not derived from pluripotent cells. We have therefore added the following text to the discussion:

Although we here find that the transcriptome of TART tissues are clearly distinct from fetal tissues, we did not prove that TART does not originate from fetal Leydig cells. TART being derived from a multipotent progenitor cell is still possible as we initially hypothesized, given the fact that TART is likely already present in utero and its resemblance to both testis and adrenal tissues which derive from a common primordium. Therefore, we were surprised to find TART to be more like adult adrenal and testis tissue, raising the possibility of TART being derived from a ‘mature’ progenitor cell type, i.e. adult stem Leydig cells or adrenal progenitor cells, that under influence of high ACTH levels and/or the localization in the testicular region might differentiate into a distinct cell type that expresses both adrenal- and testis-specific markers. However, this remains to be established.

**Minor Comment:**

In Methods: Was RNA isolated from FFPE sections or frozen tissue?

We agree that this was not clearly mentioned enough in our original manuscript, as both frozen (RNA isolation) and FFPE (IHC) material was used. We have now clarified in the methods section that the RNA was retrieved from frozen tissue samples (page 8; RNA isolation, library preparation, and sequencing paragraph).

Reviewer #1 (Significance (Required)):

This first study of transcriptome analysis of TART provides useful insight into the characteristics of these rare tumors that commonly develop in males with classic CAH. This study provides a foundation for further investigation of the biological pathways contributing to the development of TART, the most common cause of male infertility in CAH. This study is of interest to endocrinologists. Reviewed by a pediatric endocrinologist and molecular biologist - we are not completely aware of the sequencing analysis but are familiar with clustering and enrichment analysis.

Comments Reviewer #3:

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

The manuscript by Schröder et al describes the transcriptome sequencing of TARTs in CAH/CS in order to sort out the origin of TARTs. This is an interesting subject and the manuscript is well-written but I have a few comments that could be addressed.

• Some parts of the Results should be in the Methods and some in the Discussion. In the Results only the results should be given.

Response: We agree that we have incorporated some methodological sentences and some concluding remarks in the results sections to, in our opinion, improve the flow of the manuscript. As the manuscript guidelines differ between journals, we have for now decided not to change this. We will do so if this is wanted by the concerning journal.

Normally TARTs are not removed or biopsied, if not by mistake... Thus, most centers would not have tissue samples of TARTs at all. How come you have so many samples available?

Response: We thank the reviewer for highlighting this. As indeed TARTs are not routinely removed, the number of TART tissues included in our dataset is unique. Most of the TART samples were already obtained in 2004 because of reported pain and discomfort and in an attempt to improve semen quality in these patients. Removal of those particular TART samples have led to new insights that removal of longstanding TART did not improve semen parameters, nor parameters of pituitary-gonadal function (Claahsen-van der Grinten et al., 2007). Therefore, to date, the only indication for surgery for the removal of longstanding TART is the relief of pain or discomfort.

Ref 2 and 3 are rather old and similar. Could newer review references be used instead?

Response: We have changed those two references for a more recent review by Dr. Witchel on Congenital Adrenal Hyperplasia, who addresses both statements in a more recent review (Witchel, 2017).

Reviewer #3 (Significance (Required)):

New and significant study. Very interesting for people dealing with CAH patients.

References

Claahsen-van der Grinten, H. L., Otten, B. J., Takahashi, S., Meuleman, E. J. H., Hulsbergen-van de Kaa, C., Sweep, F. C. G. J., & Hermus, A. R. M. M. (2007). Testicular adrenal rest tumors in adult males with congenital adrenal hyperplasia: Evaluation of pituitary-gonadal function before and after successful testis-sparing surgery in eight patients. Journal of Clinical Endocrinology & Metabolism, 92(2), 612-615. doi:10.1210/jc.2006-1311

Witchel, S. F. (2017). Congenital Adrenal Hyperplasia. J Pediatr Adolesc Gynecol, 30(5), 520-534. doi:10.1016/j.jpag.2017.04.001

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

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

The manuscript by Schröder et al describes the transcriptome sequencing of TARTs in CAH/CS in order to sort out the origin of TARTs. This is an interesting subject and the manuscript is well-written but I have a few comments that could be addressed.

1. Some parts of the Results should be in the Methods and some in the Discussion. In the Results only the results should be given.
2. Normally TARTs are not removed or biopsied, if not by mistake... Thus, most centers would not have tissue samples of TARTs at all. How come you have so many samples available?
3. Ref 2 and 3 are rather old and similar. Could newer review references be used instead?

Significance

New and significant study. Very interesting for people dealing with CAH patients.

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 manuscript by Schroder M., et al describes the whole transcriptome of testicular adrenal rest tumors (TART) and shows that TART tissue is characteristically similar to adult adrenal and testicular rather than fetal adrenal and testicular tissues. The authors propose that their previous claim that TART is derived from an undifferentiated pluripotent progenitor is likely untrue and claim that TART likely originates from a mature cell type with both adrenal and testicular characteristics. The authors describe a unique cell type most similar to the adult adrenal, but with variable testis-specific gene expression patterns. The finding of overexpressed genes associated with ECM remodeling is interesting and may provide insight into the natural history of these tumors. A strength of the study is the number of tissue samples since surgery for these rare tumors is usually not performed.

Major Comments:

1. The key conclusions are mostly based on RNA studies, thus their claims are preliminary.

2. In Figure 1/Result p. 3: Authors claim that there were no exclusive HSD17B3 staining cells without CYP11B1, however Figure 1 looks like there are exclusively green (HSD17B) areas (especially TART3). The authors need to address this. It appears as if there are mature Leydig cells. This is important because the presence of Leydig cells would affect the interpretation of the finidings

3. Discussion: authors state that based on their previous observations that fetal Leydig cells have both adrenal and testis developmental potential. It was speculated that TART might have been derived from a totipotent progenitor cell type, but the current study shows that these tumors lack similarities with fetal tissues. Thus, the authors claim that these tumors are not derived from the transdifferentiation of pluripotent cells. However what is the origin of this mature distinct cell type? Is it not possible that this distinctive cell type is derived from a common progenitor since the testis and adrenal gland are derived from the same adrenogonadal primordium? Lack of similarities with fetal tissues at this late stage of development does not necessarily rule out a common progenitor origin.

Minor Comment:

In Methods: Was RNA isolated from FFPE sections or frozen tissue?

Significance

This first study of transcriptome analysis of TART provides useful insight into the characteristics of these rare tumors that commonly develop in males with classic CAH. This study provides a foundation for further investigation of the biological pathways contributing to the development of TART, the most common cause of male infertility in CAH. This study is of interest to endocrinologists. Reviewed by a pediatric endocrinologist and molecular biologist - we are not completely aware of the sequencing analysis but are familiar with clustering and enrichment analysis.

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

Learn more at Review Commons

---

Reply to the reviewers

---

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

The manuscript entitled "Vasohibin-1 mediated tubulin detyrosination selectively regulates secondary sprouting and lymphangiogenesis in the zebrafish trunk" by de Oliveira investigates the function of the carboxylpeptidase Vasohibin during the formation of the zebrafish trunk vasculature and reports a requirement of Vasohibin for secondary sprout formation and in particular the formation the lymphatic vasculature.

Having established the expression of Vasohibin in sorted ECs of 24 hpf embryos, the remaining study addresses the function of Vasohibin in this cell type. It is largely based on the use of a splice-site interfering morpholino. Particular commendable is the analysis, demonstrating that the KD of vash-1 indeed results in a significant reduction of detyrosination in endothelial tubulin. Findings in the vascular system then include: (i) the detection of increased division and hence supernumerous cells occurring selectively in 2nd sprouts from the PCV; (ii) an increased persistence of the initially formed 3 way connections with ISV and artery; (iii) reduced formation of parachordal lymphangioblasts and (iv) a reduced number of somites with a thoracic duct segment; (v) frequent formation of lumenized connections between PLs (where present) and ISV. To demonstrate specificity, the approach was repeated with a different morpholino and defects were partially rescued by MO-insensitive RNA.

Possible additional and relevant information could include data on a vash-1 promotor mutant to independently verify the MO-based functional analysis. Mutants would also allow analysis of further development, are the defects leading to the demise of the fish or is a later regeneration and normalization of the lymphatic vasculature observed?

We agree that a mutant would be desirable to validate the phenotypic analysis of the morpholinos used, and would also allow for further analysis. However, this is not achievable within a reasonnable time frame, especially in the context of current work restrictions.

In addtion to the two splice morpholinos currently used to knockdown vash-1 expression, we will use an ATG morpholino to further investigate our observations and hypothesis regarding the role of vash-1 in lymphatic vessels formation. We will also validate it by westernblot and attempt to rescue it with mRNA.

We have not investigated the phenotype past 4 dpf. We will add investigation of lymphatics and morphology at 5 dpf.

In addition, are other lymphatic vessel beds like the cranial lymphatics affected?

Using the Tg[fli1a:EGFP]y7 line, we have not been able to identify apparent differences in other vascular beds including the cranial lymphatics. However a detailed fine-grained investigation of the cranial vascular bed has not been performed. Given the focus of the present study on the trunk vasculature to understand the mechanisms of vash-1, we feel that a detailed analysis of cranial lymphatics would at this stage be somewhat out of scope.

PLs have been demonstrated to be at least partially guided in their movement by the CXCR4/SDF1 system and SVEP1. Has the expression of these factors been tested in vash-1 KDs?

We have not investigated the potential role of the CXCR4/SDF1 system and SVEP1 in vash-1 regulation of lymphangiogenesis. We will investigate the expression of cxcr4a, cxcl12a, cxcl12b and svep1 by in situ hibridization upon vash-1 knockdown.

With regards to the frequently observed connections of PLs and ISVs in vash-1 morphants, can the proposed lumen formation of these shunts be demonstrated e.g. by injection of Q-dots or microbeads into the circulation?

Although the lumenisation is very clear thanks to the membrane targeted expression of the label in this line, we will further analyse whether these abberant ISV to ISV connection can be perfused by Q-dots injections.

Concerning the mechanisms of these defects, is it possible to analyse the asymmetric cell division leading to 2nd sprouts in greater detail? Is the same number or are more cells sprouting form PCV and can the fli1ep:EGFP-DCX cell line in fixed samples be used to identify the spindle orientation in dividing cells?

We agree with the reviewer and plan to use the Tg[fli1ep:EGFP-DCX] fish line to investigate spindle asymmetry in uninjected embryos, as well as compare the spindle in control MO and vash-1 KD embryos. Vash-1 has been shown to regulate spindle formation in osteosarcoma cells (Liao et al., 2019). We will attempt to clarify whether this function is conserved in endothelial cells and contributes to the control of endothelial cell proliferation during initiation and formation of secondary sprouting.

We also agree that it is important to look at the PCV in the begining of secondary sprouting and will clarify whether the sprouting is initiated by an increased number of cells.

**Minor issues:** Page 5, Mat & Meth, please spell out PTU at its first mention.

This has been corrected accordingly (see page 4).

Page 6 Mat & Meth, Secondary sprout and 3-way connection parameters: The number of nuclei was assessed in each secondary sprouts (del s, singular) just prior...

This has been corrected accordingly (see page 5).

Page 16, 8th line from bottom: Recent work demonstrated that a secondary sprout either contributes (add s) to remodelling a pre-existing ISV into a vein, or forms (add s)a PLs (Geudens et al., 2019).

This has been corrected accordingly (see page 16).

Page 25, Legend to Fig. 2D-G: "...G,G' shows quantification of dTyr signal upon vash-1 KD..." Fig2 G,G' show immunostaining rather than quantification of the dTyr signal, which is shown Fig. 2H-J

This has been corrected accordingly (see page 26).

Fig. 1D / Fig. 2H-J please increase weight of the error intervals and / or change colour for improved visibility

This has been corrected accordingly (Fig. 1D and 2H-J), and we added n.s. to Fig. 1D.

Reviewer #1 (Significance (Required)):

Taken together the manuscript is comprehensively written and the study provides a conclusive analysis of the MO-mediated KD of Vasohibin in zebrafish embryonic development presenting significant novel findings. Known was a generally inhibitory function of Vasohibin on vessel formation and its enzymatic activity as a carboxylpeptidase responsible for tubulin detyrosination, affecting spindle function and mitosis. New is the detailed analysis of the Vasohibin KD on zebrafish trunk vessel formation and the description of a selective impairment of 2nd sprout formation. The manuscript is of interest for vascular biologists.

REFEREES CROSS-COMMENTING

I fully concur with the comments of reviewer #2, all three reviews find that this study is of significant interest to the vascular biology community as the relevance of tubulin detyrosination for developmental angiogenesis has not been investigated. Also all three reviews highlight the potential limitations of the use of splice morpholinos (suggested alternatives include ATG morpholinos and CRIPR mutants), the requirement to provide further evidence for a endothelial cell autonomous defect and the need to clarify some of the data representation.

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

**Summary:**

The manuscript by Bastos de Oliveira et al. describes an important investigation of the endothelial tubulin detyrosination during vascular development. Namely, they found detyronised microtubules in secondary sprouts, which is absent in MO-vash-1 treated embryos. The authors use the vash-1 morpholino approach to uncover the developmental consequences of suppressed detyrosination in angiogenesis and lymphangiogenesis in vivo in zebrafish. By a combination of transgenic lines, immunohistochemistry and time-lapse imaging, Bastos de Oliveira et al., have found that Vash-1 is a negative regulator of secondary sprouting in zebrafish. The authors showed that in the absence of Vash-1 more cells are present in the secondary sprouts due to increased cell proliferation; however lymphatic vascular network fails to form. The current manuscript requires additional experimental evidence to support the conclusions. Please see below the major technical concerns and minor comments.

**Major comments:**

-This study is based on analysis of the phenotypes observed in embryos injected with vash-1 morpholino. The authors use two different types of splice morpholinos, perform rescue experiments with RNA, and validate one MO-vash-1 with western blot. Morpholinos are not trivial to work with, and the results are variable hence additional controls need to be included, as following the recommendation put together by the zebrafish community (Stainier et, al., Plos Genetics, 2017). As the severity of the phenotypes comparing MO1 with MO2 is different and MO-vash-1 embryos appear developmentally delayed (Figure 2D-F and 5E-F overall size seem to be affected), additional MO is required, for example, ATG-MO or generation of CRISPR mutant would be favourable. All the morpholino used need to be validated using an antibody, RT-PCR and qPCR. It is essential to carry out the rescue experiments for all the MO used in this study and following the guidelines. Including the dose-response curve, data would be informative.

We agree with the reviewer and the recommendations of the zebrafish community. We will investigate the phenotypes with another KD strategy, such as the ATG-Morpholino suggested by the reviewer. We will also supply more validation of the MO2 including RNA rescue and westernblot (already included in Fig. 5 I).

We added dose-response curves (Supp. Figure 1 E,G) and a developmental morphology assessment for the morpholino 1 (Supp. Figure 1 A,B).

Given our extensive analysis of the effects of vash-1 KD, we believe the embryos in 2F are not developmentally delayed. However, the image in figure 2F does give that impression, and therefore may have triggered the reviewer’s concerns. We double checked and found that due to an oversight, we included a picture from a slightly different region of the trunk in comparision to Fig. 2D. We will add pictures of the same trunk region (Fig.2D-F) as we have done in all other figures. We nonetheless supply a supplementary figure 1 showing and quantifying the development of the analysed vash-1 morphants.

-In addition to EC, the levels of dTyr are lower in MO-vash-1 in neural tube and neurons spanning the trunk (Figrue 2 D-G'). These have been previously shown to be important for secondary sprouting. Is it possible that the observed phenotypes in the secondary sprouting are due to defects in these neurons?

We agree with the reviewer that a potential contribution of altered neuronal differentiation to the vascular phenotype should be clarified. We will assess the morphology of the neurons and their dendrites relevant for pathfinding (Lim et al., 2011) in vash-1 KD embryos, using a pan-neuronal zebrafish line, as well as via immunostaining against alpha-tubulin. Should we find evidence for changes in neuronal cells, we will attempt to clarify a cell autonomous role of vash-1 by transplantation experiments.

-Embryo number used in this study appears to be low especially in figure 3G, 5D, 5G, to conclude draw conclusions from these experiments, the number of embryos used should be higher than 20. Figure 4J please specify how many embryos were used.

We will increase the number of embryos per condition to a minimum of 20 embryos and update the averages in the text for 3G (control: 7 and vash-1 KD: 11 embryos).

In 5D and 5G each point is an embryo and more than 20 embryos per condition were used (in 5D 23-35 embryos per condition, in 5G 60-63 embryos/condition), we corrected the legend 5D and 5G (see page 27) and made it clear that each point in the graph corresponds to one embryo (5D- percentage of PLs associated with veins in each embryo; 5G- percentage of somites with toraxic duct in each embryo).

In 4J, 18 embryos were used for control (about 3 sprouts/embryo– 52 sprouts quantified) and 7 embryos for vash-1* KD condition (about 3 sprouts/embryo – 24 sprouts quantified). We corrected the number of control sprouts in the legend and added the number of embryos to increase clarity (see page 27).

-The authors hypothesise that VASH acts in the sprouting endothelial cells, based on the Q-PCR in Figure 1. However, in this experiment all EC have been sorted thus this remains ambiguous in which cell types vash-1 is expressed. Please provide the expression pattern for vash-1 across the developmental stages the phenotypes are observed.

We agree with the reviewer that it would be beneficial to understand the expression pattern of vash-1 in wild type embryos. We plan to perform in situ hybridization for vash-1 mRNA.

-Throughout the manuscript the authors refer the lymphatic identity, however, there is no evidence in the paper that the identity status has been assessed. To support these claims Prox1 immunohistochemistry or analysis of prox1 expression in the reporter line would be appropriate.

We agree with the reviewer and plan to perform a Prox1 immunostaining (Koltowska et al., 2015) in vash-1 KD embryos at 34-36 hpf (secondary sprouting) to investigate Prox1 levels upon vash-1 KD.

**Minor comments:**

-The authors refer to the literature where overexpression of VASH suppresses the angiogenesis. As the RNA injections were used in rescue experiments, the data of vash-1 RNA injections into the wild-type embryos would be beneficial.

We have injected vash-1 RNA into a control morpholino injected embryos (28 control embryos, 14 Vash-1 RNA injected embryos) and we observed a significant decrease in PLs at 52 hpf (average of -control: 87,5% somites with PLs to 67% somites with PLs in vash-1 RNA embryos). This could be due to a decrease of secondary sprouting, which would be in accordance with the current literature that vash-1 overexpression is anti-angiogenic. We will further investigate and add the results to figure 5. Figure 1. vash-1* mRNA injection leads to a decrease in somites with PLs (preliminary).

-In figures 2J, 3J, 3K, 3N, 4J, 5C, 5D and 5G the N number was set for examples as the number of sprouts, the number of somites with TD, number of ISV. To strengthen the observation in the manuscript quantification of the sprouts, PL, vISVs and lymphatic phenotypes with N set as the number of embryos would be more informative. Indicating the number of embryos used, in the graphs, would be helpful.

We agree with the reviewer and have added embryo numbers in all legends and graphs. In 2J, 3J, 3K, 4J each point is a sprout, a cell division or an ISV, corresponding to the N. We agree that the number of embryos could be more clearly stated, so we added the number of embryos analysed in the figure legend and will add them in the graphs.

In 5C, 5D and 5G each point corresponds to an embryo (clarified in the legend of Fig. 5- see page 27).

Fig. 5C refers to the percentage of somites with PLs in each embryo, 5D refers to percentage of the existing PLs in one embryo connected to a venous ISV, 5G corresponds to percentage of somites with a TD segment in each embryo.

-In Figure 5A, B and D the authors quantify what they refer to as a lumenised connection between the vISVs and PL. In the control image (second star), a somewhat lumenised structure is present, clarification of how the scores were set is missing.

In Fig. 5C we show a quantification of the percentage of somites with PLs per embryo, by counting the PLs identified with an asterisk in Fig. 5A-B. PLs are normally not lumenised, with few exceptions also ocurring in wild-type – see Fig. 4 in (S Isogai et al., 2001).

In Fig. 5D we quantified the proportion of PLs associated/connected with venous ISvs (see Methods section page 6), by 52 hpf in control and vash-1 morphants.

In 5B and 5F,F‘, the arrowheads identify lumenised PLs present in vash-1 KD embryos. We will add a quantification of kdr-l:ras-Cherry positive ISV-to-ISV connections, corresponding to the lumenised endothelial connections, since kdr-l:ras-Cherry signal labels endothelial (and not lymphatic) cells and is particularly strong at the luminal endothelial membrane of the vessel.

-In Figure 3 E and F the authors show the excessive sprouting phenotype between controls and Mo-vash-1. The images presented are taking from different parts of the embryos (middle of the trunk vs plexus region), hampering the comparison between the two groups. The quantification of the phenotypes in both experimental groups should be in the same region of the embryo, as the local difference can occur. It is key to provide representative images to support these observations.

The images presented are representative of the phenotype quantified, and the time-lapses were done in comparable regions of the zebrafish trunk (+- 1-2 somites in both groups due to drift during image aquisition), making the comparison possible.

-Figure 1D the vash-1 expression levels in EC seem very variable in this graph, therefore no conclusion can be drawn from this data, especially as the authors do not provide the p-values.

We added n.s. in the graph, to make it clear that the difference between developmental stages is not significant, potentially due to high biological variability between embryos, as seen in two primer pairs. We believe that presenting this biological variability is of importance to the readers.

We write on page 12 about this result: „During the sprouting phase (24hpf), vash-1 expression was 5-7 times higher in endothelial than in non-ECs, decreasing at 48 hpf (Fig. 1C-D). Although these results are not significant, they were independently confirmed with a second primer set.”. The only conclusion we made from this data is that Vash-1 is dynamically expressed in the zebrafish endothelium during development, as we now added in the discussion (page 14).

-In the introduction, the authors state: 'Although primary and secondary sprouts appear morphologically similar, with tip and stalk cells' - Please provide the reference that supports the claim that secondary sprouts have tip-stalk cells morphology/organisation.

Although many studies have investigated primary and secondary sprouting, identifying both shared as well as distinct molecular regulation, and show morphological details that are apparently similar, a formal claim that secondary sprouts show tip and stalk cell identities and behaviour is hard to find. Given that this is not relevant for the central findings of the work, we modified the sentence and added a reference “Although primary and secondary sprouts appear morphologically similar, with tip and stalk cells” (Sumio Isogai et al., 2003)…” See page 2.

We also updated the discussion for consistency: “Although the cellular mechanisms of primary and secondary sprouting in zebrafish appear very similar, with tip cell selection and guided migration and stalk cell proliferation, secondary sprouting utilises alternative signalling pathways and entails a unique specification step that establishes both venous ISVs and lymphatic structures.” (see page 15)

-The authors refer the increased cell division phenotypes observed in the movies, however, the movie files have not been available to the reviewers.

We will provide the movies.

Reviewer #2 (Significance (Required)):

This is an important study as uncovering the mechanistic details of angiogenic and lymphangiogenic negative regulators is of high value with the potential for therapeutic developments. To date, Vash-1 has been only studied in the context of tumour angiogenesis, vasculature in diabetic nephropathy and pulmonary arterial hypertension, and it remains unclear what is its role during development and how does it regulate vascular network formation. The tyrosination status of microtubule in endothelial cells is understudied. This study revealed, previously uncharacterised detyrosinated microtubules in endothelial cells in vivo. And further dissects how this process might be regulated, brings unique insights into the vascular biology field and beyond. Thus, delving into the cell biological mechanism such as microtubule dynamics and modification in vivo in embryo context is a significant step forward in setting new standards in the field.

I am developmental biologist who has experience in model organisms such as zebrafish and mouse. The main focus of my work is on developmental angiogenesis and lymphangiogenesis.

REFEREES CROSS-COMMENTING

After reading the other reviews comments, it seems that we all agree that this study is of high value to vascular biology field and beyond bringing novel findings.

Importantly the reviewers' comments are in line with each other and have identified several commonalities that should be addressed. Such as: Further validation of Morpholinos, or using alternative methods to replicate the findings. additional evidence that the observed phenotypes are primary due to vash-1 requirement within EC, and not due to the secondary effect in other cells such as CXCR4/SDF1 system and SVEP1, neurons or general delay of the embryos Further evidence of for VASH expression pattern the number of embryos used in the experiments, and how the data is represented.

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

Vasohibin-1 (Vash-1) is known to detyrosinate microtubules (MTs) and limit angiogenesis. Using in vivo live imaging and whole mount immunofluorescence staining of zebrafish trunk vasculature, Bastos de Oliveira et al. show that the MT detyrosination role of Vash-1 is conserved in zebrafish and that Vash-1 is essential for limiting venous sprouting and subsequent formation of lymphatics. Their findings suggest a role for MT detyrosination in lympho-venous cell specification.

**Major comments:**

1 . The authors claim that Vash-1 regulates secondary sprouting and lymphangiogenesis by detyrosinating MTs. However, no direct evidence of this link is provided in the manuscript. The authors only separately show that knockdown of vash-1 affects MT detyrosination and secondary sprouting and lymphangiogenesis. They have not shown a causative effect. The authors should therefore qualify the above stated claim as speculative. In other words, the authors should mention that their data only suggests that disruption of MT detyrosination is the underlying cause for aberrant secondary sprouting and lymphangiogenesis in vash-1 KD embryos.

We agree with the reviewer about the lack of evidence to state that the disruption of microtubule detyrosination leads to aberrant secondary sprouting. Although we believe this is the most parsimonius explanation for the secondary sprouts behavioural defects as cell division is disturbed and microtubule detyrosination is implicated in cell division (Barisic et al., 2015), we want to make clear that our data currently only suggest a specific role of microtubule detyrosination in secondary sprouting. Examples of this are page 14 of the discussion „These results suggest that Vash-1-driven microtubule detyrosination limits excessive venous EC sprouting and proliferation during lympho-venous development in zebrafish.” as well as the abstract.

We also corrected the sentence in the discussion (page 14): “In this study, we identified Vash-1-mediated microtubule detyrosination as a cellular mechanism as a novel regulator of EC sprouting from the PCV and the subsequent formation of lymphatic vessels in the zebrafish trunk.”

To avoid any overstatement, we also propose the following title change: Vasohibin-1 mediated tubulin detyrosination selectively regulates secondary sprouting and lymphangiogenesis in the zebrafish trunk.

As detailed in response to comment 2 below, we will however attempt to investigate the direct connection. Depending on the outcome, we will adapt conclusions and title accordingly.

2 . In order to provide more compelling evidence for a direct relationship between MT tyrosination and lymphangiogenesis, the authors could try mutating the carboxypeptidase domain of vash-1 or overexpressing a dominant negative transcript (that contains a mutated carboxypeptidase domain). If this gives the same phenotypes as the vash-1 morphants, it would indicate that the carboxypeptidase activity of Vash-1 (in detyrosinating MTs) is responsible for limiting secondary sprouting and promoting specification of lymphatics. This suggested experiment is fairly realistic in terms of both time and resources. For example, since the authors already have the human vash-1 cDNA cloned, making a dominant negative transcript from this would take around two weeks, imaging and analysis of embryos injected with this mRNA would take another four weeks. Therefore, in total, the suggested experiment would take around 6 weeks. Although the alternative experiment, that is, making a carboxypeptidase domain mutant of vash-1 would be a better choice in terms of reproducibility and long-term use of a stable line, it would admittedly take a relatively larger amount of time. Therefore, the ultimate choice would depend on the authors.

We will investigate this further by cloning and expressing a mutated vash-1 cDNA which translates a validated catalytically dead Vash-1 (Nieuwenhuis et al., 2017). However, this mutant has not been shown to function as dominant negative, so it is unclear whether it can be used as a dominant negative mutant.

3 . Both the data and methods are presented in a way that ensures reproducibility. The statistical analysis is very well done, in that the authors were very prudent in their choice of statistical tests. However, in many figures and subfigures (Fig. 2B, H-J; Fig. 3G, J, K, N; Fig. 4J; Fig. 5J), the number of replicates was not mentioned and instead only the sample size was stated. Whether this was just an oversight or if it should be taken to mean that the analysis was performed on just one replicate is unclear. The authors need to clarify this aspect of their analysis. Further, In Fig. 2H-J, Fig. 3G,J, K, N and Fig. 4J, the total number of data points in control MO vs vash-1 KD seem to be quite different. In other words, there seem to be a lot more data points in one experimental condition than the other. Does this difference fall within the acceptable range? If the authors were to compare a similar number of data points between the two experimental conditions, would the results of the statistical analysis still be the same?

We apreciate this comment and clarified the replicate numbers in the figure legends: Fig. 2B- 3 replicates (page 25), Fig. 2 H-J- quantification is 1 replicate (page 26), Fig. 2 D-G is representative of 3 replicates (page 25). Fig. 3 G,J,K,N – quantification is from 1 replicate (page 26), Fig. 3 B,C,E,F,H,I are representative of 2 experimental replicates (page 26). Fig. 4J – quantification is 1 replicate (page 27), Fig. 4 A-F is representative of 3 replicates (page 27). Fig. 5 J correspondes to 1 replicate (page 28).

We plan to increase replicates and numbers in quantifications shown in Fig. 3 G,J,K,N and Fig. 5 J as they are relevant for the conclusions of the manuscript, and adapt the text.

The quantifications of immunostaining signals are comparable between different samples of the same experiment but technically not easy accross different experiments, due to some variability of the immunostaining. However, the pattern we report in the quantifications and representative pictures is consistentely detected (reduced dTyr signal upon vash-1 KD in Fig 2 D-G; higher dTyr intensity in secondary rather than primary sprouts in Fig. 4 A-F). We added in the legend that the pictures of the embryos in these figures are representative of 3 biological replicates (see page 25 and 27).

We recognise the unequal sample size in control and vash-1 KD groups in Fig. 2H-J, Fig. 3G,J, K, N and Fig. 4J. Generally, the vash-1 KD group shows more variance than the control group (see Fig. 3 J-N, 4J for example), hence the reason why we analysed a higher sample size.

In the planned experiments (repeating quantifications of Fig. 3 J-N), we will analyse a similar number of embryos.

We corrected the figure legend of 2 H-J on the number of ISVs - 108 ISVs from 7 embryos for control and 150 ISVs for vash-1 KD, from 9 embryos (see page 26).

4 . The authors only provide KD data on the function of vash-1 using morpholinos. According to several recent guidelines concerning the use of morpholinos, this is not widely accepted in the zebrafish community as sufficient to provide robust insight into gene function. Please refer for example to the following publication: Guidelines for morpholino use in zebrafish, Stainier et al., PLOS Genetics, 2017. The generation of a vash-1 mutant is a necessary requirement for backing up morpholino KD data. Further, even though the authors state that embryos were selected on the pre-established criteria that they have normal morphology, beating heart, and flowing blood, certain morphological differences between control MO injected and vash-1 KD embryos could be observed in some figures. In Fig. 2D, F and Fig. 5A, B, E, F the vash-1 KD embryos seem smaller (extend of the dorso-ventral axis) than control MO injected embryos. The authors need to provide images showing the overall morphology of morpholino injected embryos and need to provide evidence that morpholino injections do not cause developmental delays.

We agree that a mutant would be desirable to validate the phenotypic analysis of the morpholinos used, and would also allow for further analysis. However, this is not achievable within a reasonnable time frame, especially in the context of current work restrictions. We have added a sentence about the need to confirm the loss of function phenotype with vash-1 mutants in the discussion (see page 14).

In addtion to the two morpholinos currently used to knockdown vash-1 expression, we will use an ATG morpholino to further investigate our observations and hypothesis regarding the role of vash-1 in lymphatic vessels formation. We will also validate it by westernblot and attempt to rescue it with mRNA.

We added a supplementary figure with pictures and quantifications of antero-posterior (Sup. Figure 1 C) and dorso-ventral length (Sup. Figure 1 D) of the analysed control and vash-1 morpholino injected embryos‘ development at 24, 34, 52 and 4dpf which shows no significant developmental delay and morphological defect. There is some occurrence of curvature of the tail at 34-52 hpf.

We added a sentence in the Methods section (pages 10) to clarify the morphant’s morphology and dosage-response curves.

We observe a 1-2 hour developmental delay of both the control and the vash-1 KD embryos compared to uninjected wild-type embryos, which led us to chose the 52 hpf time point to investigate the PLs. In uninjected embryos they are usually developed by 48hpf (Hogan et al., 2009).

Fig. 2 D shows a more anterior region of the zebrafish trunk than Fig. 2F (the tail has a smaller dorso-ventral length)- we will provide more comparable pictures from the same region.

Fig. 5B is slightly tilted – we will provide a picture with the same orientation.

Fig. 5 E and F have a similar length from dorsal aorta to the dorsal longitudinal anastomotic vessel. However, we appreciate a difference in the sub intestinal vascular plexus (SIVP), which is consistently underdeveloped in the vash-1 KD embryos.

Figure 2- vash-1 deficient embryos show underdeveloped intestinal vascular system at 4 dpf.

**Minor comments:**

a. The authors should back their qPCR data for vash-1 expression (Figure 1) by standard mRNA in situ hybridization, given the large degree of variability in vash-1 expression. Do they observe a dynamic expression in the vasculature using this technique?

We agree with the reviewer that an in situ hybridization would be beneficial to understand the expression pattern of vash-1 in wild type embryos. Accordingly, we will look at vash-1 expression by in situ hybridization in WT embryos.

The number of nuclei per sprout in Fig. 3J does not correspond with the number of divisions per sprout presented in Fig. 3K. The authors observe one or two cell divisions per sprout in ctr MO injected embryos (Fig. 3K), however, Fig. 3J shows that the majority of ctr. sprouts contains only one cell. This is even more dramatic for vash-1 MO injected embryos, which can have up to four divisions, therefore should contain six cells. However, the maximum number of cells the authors report is three to four cells. How do these observations go together?

We believe these quantifications are not contradicting. The number of endothelial nuclei was assessed just prior to the connection to the ISV and the cell division quantification was done in a time-lapse from the time of secondary sprout emergence until the resolution of the 3-way connection. It is expected that there are more cell divisions during a longer time frame, as cells migrate dorsally or ventrally out of the sprout.

Fig. 5I and J have the same data points for control MO and vash-1 MO1. Does this mean that both graphs are from the same experiment? If so, the authors could combine the two graphs into one. If the two graphs are not from the same experiment, both would need to have independent controls.

Fig 5 I and J are indeed from the same experiment. They are now combined into one graph (see Fig. 5 J).

d. The percentage of somites with PLs in vash-1 MO1 injected embryos in Fig. 5I is half the value shown in Fig. 5C. Although this kind of variability might be expected in biological samples, perhaps the authors could briefly discuss the issue and its implications on reproducibility in the manuscript so as to have the readers be aware of it, especially since the rescue of the vash-1 morpholino phenotype back to 50% from 25% is the same value the authors observed in the vash-1 KD alone in Fig. 5C. Here the value is 50% for the morpholino injection.

We added a sentence discussing the phenotypic variability in the discussion (see page 16), and we added a dosage response curve for the PLs (Sup. Figure 1 F), showing that embryos injected with the same amount of morpholino show variability in the percentage of somites with PLs at 52hpf. We added a more representative picture of PLs for vash-1 morphant in Fig. 5I ( Y-axis of Fig. 2H and 4J correspond to ratios, which have no units. Nontheless, we added AU/AU to these graphs to make it clearer. We added the bars in Fig. 5D.

It would help to have an inference or conclusion at the end of each results section.

We added one conclusion sentence per results section (see pages 11-14).

Reviewer #3 (Significance (Required)):

Conceptual: As per my knowledge, this is the first study that looks at microtubule modifications in the context of a vertebrate organism past the gastrulation stage, as opposed to similar studies that have been done in cell culture or invertebrates (S. cerevisiae, C. elegans and D. melanogaster). Moreover, this study is one of few that address a novel link between the cytoskeleton and the process of cell fate specification.

Previous studies have separately shown that Vash-1 limits angiogenesis and detyrosinates MTs. The current study combines the two observations in the context of endothelial cells, and hypothesizes that perhaps the function of Vash-1 in limiting angiogenesis and at the same time promoting lymphatic development could be due to its role in MT modification at the molecular level and the consequent effect of this on cell division and/or fate specification at the cellular level. In short, this study aims to connect the long-standing gap in knowledge between cytoskeletal modifications and cell dynamics (in particular, division and specification) in a vertebrate organism. I therefore believe that the current study would be an exciting finding for research communities that study cytoskeletal influence on cellular dynamics and also those in the broad area of vascular biology.

My field of expertise relates to vascular biology, specifically developmental angiogenesis and the behavior of endothelial cells in zebrafish.

References

Barisic, M., Silva E Sousa, R., Tripathy, S. K., Magiera, M. M., Zaytsev, A. V., Pereira, A. L., Janke, C., Grishchuk, E. L., & Maiato, H. (2015). Microtubule detyrosination guides chromosomes during mitosis. Science, 348(6236), 799–803. https://doi.org/10.1126/science.aaa5175

Hogan, B. M., Bos, F. L., Bussmann, J., Witte, M., Chi, N. C., Duckers, H. J., & Schulte-Merker, S. (2009). Ccbe1 is required for embryonic lymphangiogenesis and venous sprouting. Nature Genetics, 41(4), 396–398. https://doi.org/10.1038/ng.321

Isogai, S, Horiguchi, M., & Weinstein, B. M. (2001). The vascular anatomy of the developing zebrafish: an atlas of embryonic and early larval development. Developmental Biology, 230(2), 278–301. https://doi.org/10.1006/dbio.2000.9995

Isogai, Sumio, Lawson, N. D., Torrealday, S., Horiguchi, M., & Weinstein, B. M. (2003). Angiogenic network formation in the developing vertebrate trunk. Development, 130(21), 5281–5290. https://doi.org/10.1242/dev.00733

Kimura, H., Miyashita, H., Suzuki, Y., Kobayashi, M., Watanabe, K., Sonoda, H., Ohta, H., Fujiwara, T., Shimosegawa, T., & Sato, Y. (2009). Distinctive localization and opposed roles of vasohibin-1 and vasohibin-2 in the regulation of angiogenesis. Blood, 113(19), 4810–4818. https://doi.org/10.1182/blood-2008-07-170316

Koltowska, K., Lagendijk, A. K., Pichol-Thievend, C., Fischer, J. C., Francois, M., Ober, E. A., Yap, A. S., & Hogan, B. M. (2015). Vegfc Regulates Bipotential Precursor Division and Prox1 Expression to Promote Lymphatic Identity in Zebrafish. Cell Reports, 13(9), 1828–1841. https://doi.org/10.1016/j.celrep.2015.10.055

Liao, S., Rajendraprasad, G., Wang, N., Eibes, S., Gao, J., Yu, H., Wu, G., Tu, X., Huang, H., Barisic, M., & Xu, C. (2019). Molecular basis of vasohibins-mediated detyrosination and its impact on spindle function and mitosis. Cell Research, June. https://doi.org/10.1038/s41422-019-0187-y

Lim, A. H., Suli, A., Yaniv, K., Weinstein, B., Li, D. Y., & Chien, C. Bin. (2011). Motoneurons are essential for vascular pathfinding. Development, 138(21), 4813. https://doi.org/10.1242/dev.075044

Nicenboim, J., Malkinson, G., Lupo, T., Asaf, L., Sela, Y., Mayseless, O., Gibbs-Bar, L., Senderovich, N., Hashimshony, T., Shin, M., Jerafi-Vider, A., Avraham-Davidi, I., Krupalnik, V., Hofi, R., Almog, G., Astin, J. W., Golani, O., Ben-Dor, S., Crosier, P. S., … Yaniv, K. (2015). Lymphatic vessels arise from specialized angioblasts within a venous niche. Nature, 522(7554), 56–61. https://doi.org/10.1038/nature14425

Nieuwenhuis, J., Adamopoulos, A., Bleijerveld, O. B., Mazouzi, A., Stickel, E., Celie, P., Altelaar, M., Knipscheer, P., Perrakis, A., Blomen, V. A., & Brummelkamp, T. R. (2017). Vasohibins encode tubulin detyrosinating activity. Science, 358(6369), 1453–1456. https://doi.org/10.1126/science.aao5676

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

Vasohibin-1 (Vash-1) is known to detyrosinate microtubules (MTs) and limit angiogenesis. Using in vivo live imaging and whole mount immunofluorescence staining of zebrafish trunk vasculature, Bastos de Oliveira et al. show that the MT detyrosination role of Vash-1 is conserved in zebrafish and that Vash-1 is essential for limiting venous sprouting and subsequent formation of lymphatics. Their findings suggest a role for MT detyrosination in lympho-venous cell specification.

Major comments:

1 . The authors claim that Vash-1 regulates secondary sprouting and lymphangiogenesis by detyrosinating MTs. However, no direct evidence of this link is provided in the manuscript. The authors only separately show that knockdown of vash-1 affects MT detyrosination and secondary sprouting and lymphangiogenesis. They have not shown a causative effect. The authors should therefore qualify the above stated claim as speculative. In other words, the authors should mention that their data only suggests that disruption of MT detyrosination is the underlying cause for aberrant secondary sprouting and lymphangiogenesis in vash-1 KD embryos.

2 . In order to provide more compelling evidence for a direct relationship between MT tyrosination and lymphangiogenesis, the authors could try mutating the carboxypeptidase domain of vash-1 or overexpressing a dominant negative transcript (that contains a mutated carboxypeptidase domain). If this gives the same phenotypes as the vash-1 morphants, it would indicate that the carboxypeptidase activity of Vash-1 (in detyrosinating MTs) is responsible for limiting secondary sprouting and promoting specification of lymphatics. This suggested experiment is fairly realistic in terms of both time and resources. For example, since the authors already have the human vash-1 cDNA cloned, making a dominant negative transcript from this would take around two weeks, imaging and analysis of embryos injected with this mRNA would take another four weeks. Therefore, in total, the suggested experiment would take around 6 weeks. Although the alternative experiment, that is, making a carboxypeptidase domain mutant of vash-1 would be a better choice in terms of reproducibility and long-term use of a stable line, it would admittedly take a relatively larger amount of time. Therefore, the ultimate choice would depend on the authors.

3 . Both the data and methods are presented in a way that ensures reproducibility. The statistical analysis is very well done, in that the authors were very prudent in their choice of statistical tests. However, in many figures and subfigures (Fig. 2B, H-J; Fig. 3G, J, K, N; Fig. 4J; Fig. 5J), the number of replicates was not mentioned and instead only the sample size was stated. Whether this was just an oversight or if it should be taken to mean that the analysis was performed on just one replicate is unclear. The authors need to clarify this aspect of their analysis. Further, In Fig. 2H-J, Fig. 3G,J, K, N and Fig. 4J, the total number of data points in control MO vs vash-1 KD seem to be quite different. In other words, there seem to be a lot more data points in one experimental condition than the other. Does this difference fall within the acceptable range? If the authors were to compare a similar number of data points between the two experimental conditions, would the results of the statistical analysis still be the same?

4 . The authors only provide KD data on the function of vash-1 using morpholinos. According to several recent guidelines concerning the use of morpholinos, this is not widely accepted in the zebrafish community as sufficient to provide robust insight into gene function. Please refer for example to the following publication: Guidelines for morpholino use in zebrafish, Stainier et al., PLOS Genetics, 2017. The generation of a vash-1 mutant is a necessary requirement for backing up morpholino KD data. Further, even though the authors state that embryos were selected on the pre-established criteria that they have normal morphology, beating heart, and flowing blood, certain morphological differences between control MO injected and vash-1 KD embryos could be observed in some figures. In Fig. 2D, F and Fig. 5A, B, E, F the vash-1 KD embryos seem smaller (extend of the dorso-ventral axis) than control MO injected embryos. The authors need to provide images showing the overall morphology of morpholino injected embryos and need to provide evidence that morpholino injections do not cause developmental delays.

Minor comments:

a. The authors should back their qPCR data for vash-1 expression (Figure 1) by standard mRNA in situ hybridization, given the large degree of variability in vash-1 expression. Do they observe a dynamic expression in the vasculature using this technique?

b. The number of nuclei per sprout in Fig. 3J does not correspond with the number of divisions per sprout presented in Fig. 3K. The authors observe one or two cell divisions per sprout in ctr MO injected embryos (Fig. 3K), however, Fig. 3J shows that the majority of ctr. sprouts contains only one cell. This is even more dramatic for vash-1 MO injected embryos, which can have up to four divisions, therefore should contain six cells. However, the maximum number of cells the authors report is three to four cells. How do these observations go together?

c. Fig. 5I and J have the same data points for control MO and vash-1 MO1. Does this mean that both graphs are from the same experiment? If so, the authors could combine the two graphs into one. If the two graphs are not from the same experiment, both would need to have independent controls.

d. The percentage of somites with PLs in vash-1 MO1 injected embryos in Fig. 5I is half the value shown in Fig. 5C. Although this kind of variability might be expected in biological samples, perhaps the authors could briefly discuss the issue and its implications on reproducibility in the manuscript so as to have the readers be aware of it, especially since the rescue of the vash-1 morpholino phenotype back to 50% from 25% is the same value the authors observed in the vash-1 KD alone in Fig. 5C. Here the value is 50% for the morpholino injection.

e. The Y-axis label is missing in Fig. 2H and Fig. 4J. Figure 5D lacks bars showing median and standard deviation.

f. It would help to have an inference or conclusion at the end of each results section.

Significance

Conceptual: As per my knowledge, this is the first study that looks at microtubule modifications in the context of a vertebrate organism past the gastrulation stage, as opposed to similar studies that have been done in cell culture or invertebrates (S. cerevisiae, C. elegans and D. melanogaster). Moreover, this study is one of few that address a novel link between the cytoskeleton and the process of cell fate specification.

Previous studies have separately shown that Vash-1 limits angiogenesis and detyrosinates MTs. The current study combines the two observations in the context of endothelial cells, and hypothesizes that perhaps the function of Vash-1 in limiting angiogenesis and at the same time promoting lymphatic development could be due to its role in MT modification at the molecular level and the consequent effect of this on cell division and/or fate specification at the cellular level. In short, this study aims to connect the long-standing gap in knowledge between cytoskeletal modifications and cell dynamics (in particular, division and specification) in a vertebrate organism. I therefore believe that the current study would be an exciting finding for research communities that study cytoskeletal influence on cellular dynamics and also those in the broad area of vascular biology.

My field of expertise relates to vascular biology, specifically developmental angiogenesis and the behavior of endothelial cells in zebrafish.

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 Bastos de Oliveira et al. describes an important investigation of the endothelial tubulin detyrosination during vascular development. Namely, they found detyronised microtubules in secondary sprouts, which is absent in MO-vash-1 treated embryos. The authors use the vash-1 morpholino approach to uncover the developmental consequences of suppressed detyrosination in angiogenesis and lymphangiogenesis in vivo in zebrafish. By a combination of transgenic lines, immunohistochemistry and time-lapse imaging, Bastos de Oliveira et al., have found that Vash-1 is a negative regulator of secondary sprouting in zebrafish. The authors showed that in the absence of Vash-1 more cells are present in the secondary sprouts due to increased cell proliferation; however lymphatic vascular network fails to form. The current manuscript requires additional experimental evidence to support the conclusions. Please see below the major technical concerns and minor comments.

Major comments:

-This study is based on analysis of the phenotypes observed in embryos injected with vash-1 morpholino. The authors use two different types of splice morpholinos, perform rescue experiments with RNA, and validate one MO-vash-1 with western blot. Morpholinos are not trivial to work with, and the results are variable hence additional controls need to be included, as following the recommendation put together by the zebrafish community (Stainier et, al., Plos Genetics, 2017). As the severity of the phenotypes comparing MO1 with MO2 is different and MO-vash-1 embryos appear developmentally delayed (Figure 2D-F and 5E-F overall size seem to be affected), additional MO is required, for example, ATG-MO or generation of CRISPR mutant would be favourable. All the morpholino used need to be validated using an antibody, RT-PCR and qPCR. It is essential to carry out the rescue experiments for all the MO used in this study and following the guidelines. Including the dose-response curve, data would be informative.

-In addition to EC, the levels of dTyr are lower in MO-vash-1 in neural tube and neurons spanning the trunk (Figrue 2 D-G'). These have been previously shown to be important for secondary sprouting. Is it possible that the observed phenotypes in the secondary sprouting are due to defects in these neurons?

-Embryo number used in this study appears to be low especially in figure 3G, 5D, 5G, to conclude draw conclusions from these experiments, the number of embryos used should be higher than 20. Figure 4J please specify how many embryos were used.

-The authors hypothesise that VASH acts in the sprouting endothelial cells, based on the Q-PCR in Figure 1. However, in this experiment all EC have been sorted thus this remains ambiguous in which cell types vash-1 is expressed. Please provide the expression pattern for vash-1 across the developmental stages the phenotypes are observed.

-Throughout the manuscript the authors refer the lymphatic identity, however, there is no evidence in the paper that the identity status has been assessed. To support these claims Prox1 immunohistochemistry or analysis of prox1 expression in the reporter line would be appropriate.

Minor comments:

-The authors refer to the literature where overexpression of VASH suppresses the angiogenesis. As the RNA injections were used in rescue experiments, the data of vash-1 RNA injections into the wild-type embryos would be beneficial.

-In figures 2J, 3J, 3K, 3N, 4J, 5C, 5D and 5G the N number was set for examples as the number of sprouts, the number of somites with TD, number of ISV. To strengthen the observation in the manuscript quantification of the sprouts, PL, vISVs and lymphatic phenotypes with N set as the number of embryos would be more informative. Indicating the number of embryos used, in the graphs, would be helpful.

-In Figure 5A, B and D the authors quantify what they refer to as a lumenised connection between the vISVs and PL. In the control image (second star), a somewhat lumenised structure is present, clarification of how the scores were set is missing.

-In Figure 3 E and F the authors show the excessive sprouting phenotype between controls and Mo-vash-1. The images presented are taking from different parts of the embryos (middle of the trunk vs plexus region), hampering the comparison between the two groups. The quantification of the phenotypes in both experimental groups should be in the same region of the embryo, as the local difference can occur. It is key to provide representative images to support these observations.

-Figure 1D the vash-1 expression levels in EC seem very variable in this graph, therefore no conclusion can be drawn from this data, especially as the authors do not provide the p-values.

-In the introduction, the authors state: 'Although primary and secondary sprouts appear morphologically similar, with tip and stalk cells' - Please provide the reference that supports the claim that secondary sprouts have tip-stalk cells morphology/organisation.

-The authors refer the increased cell division phenotypes observed in the movies, however, the movie files have not been available to the reviewers.

Significance

This is an important study as uncovering the mechanistic details of angiogenic and lymphangiogenic negative regulators is of high value with the potential for therapeutic developments. To date, Vash-1 has been only studied in the context of tumour angiogenesis, vasculature in diabetic nephropathy and pulmonary arterial hypertension, and it remains unclear what is its role during development and how does it regulate vascular network formation. The tyrosination status of microtubule in endothelial cells is understudied. This study revealed, previously uncharacterised detyrosinated microtubules in endothelial cells in vivo. And further dissects how this process might be regulated, brings unique insights into the vascular biology field and beyond. Thus, delving into the cell biological mechanism such as microtubule dynamics and modification in vivo in embryo context is a significant step forward in setting new standards in the field.

I am developmental biologist who has experience in model organisms such as zebrafish and mouse. The main focus of my work is on developmental angiogenesis and lymphangiogenesis.

REFEREES CROSS-COMMENTING

After reading the other reviews comments, it seems that we all agree that this study is of high value to vascular biology field and beyond bringing novel findings.

Importantly the reviewers' comments are in line with each other and have identified several commonalities that should be addressed. Such as: Further validation of Morpholinos, or using alternative methods to replicate the findings. additional evidence that the observed phenotypes are primary due to vash-1 requirement within EC, and not due to the secondary effect in other cells such as CXCR4/SDF1 system and SVEP1, neurons or general delay of the embryos Further evidence of for VASH expression pattern the number of embryos used in the experiments, and how the data is represented.

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 entitled "Vasohibin-1 mediated tubulin detyrosination selectively regulates secondary sprouting and lymphangiogenesis in the zebrafish trunk" by de Oliveira investigates the function of the carboxylpeptidase Vasohibin during the formation of the zebrafish trunk vasculature and reports a requirement of Vasohibin for secondary sprout formation and in particular the formation the lymphatic vasculature.

Having established the expression of Vasohibin in sorted ECs of 24 hpf embryos, the remaining study addresses the function of Vasohibin in this cell type. It is largely based on the use of a splice-site interfering morpholino. Particular commendable is the analysis, demonstrating that the KD of vash-1 indeed results in a significant reduction of detyrosination in endothelial tubulin. Findings in the vascular system then include: (i) the detection of increased division and hence supernumerous cells occurring selectively in 2nd sprouts from the PCV; (ii) an increased persistence of the initially formed 3 way connections with ISV and artery; (iii) reduced formation of parachordal lymphangioblasts and (iv) a reduced number of somites with a thoracic duct segment; (v) frequent formation of lumenized connections between PLs (where present) and ISV. To demonstrate specificity, the approach was repeated with a different morpholino and defects were partially rescued by MO-insensitive RNA.

Possible additional and relevant information could include data on a vash-1 promotor mutant to independently verify the MO-based functional analysis. Mutants would also allow analysis of further development, are the defects leading to the demise of the fish or is a later regeneration and normalization of the lymphatic vasculature observed? In addition, are other lymphatic vessel beds like the cranial lymphatics affected? PLs have been demonstrated to be at least partially guided in their movement by the CXCR4/SDF1 system and SVEP1. Has the expression of these factors been tested in vash-1 KDs? With regards to the frequently observed connections of PLs and ISVs in vash-1 morphants, can the proposed lumen formation of these shunts be demonstrated e.g. by injection of Q-dots or microbeads into the circulation? Concerning the mechanisms of these defects, is it possible to analyse the asymmetric cell division leading to 2nd sprouts in greater detail? Is the same number or are more cells sprouting form PCV and can the fli1ep:EGFP-DCX cell line in fixed samples be used to identify the spindle orientation in dividing cells?

Minor issues: Page 5, Mat & Meth, please spell out PTU at its first mention.

Page 6 Mat & Meth, Secondary sprout and 3-way connection parameters: The number of nuclei was assessed in each secondary sprouts (del s, singular) just prior...

Page 16, 8th line from bottom: Recent work demonstrated that a secondary sprout either contributes (add s) to remodelling a pre-existing ISV into a vein, or forms (add s)a PLs (Geudens et al., 2019).

Page 25, Legend to Fig. 2D-G: "...G,G' shows quantification of dTyr signal upon vash-1 KD..." Fig2 G,G' show immunostaining rather than quantification of the dTyr signal, which is shown Fig. 2H-J

Fig. 1D / Fig. 2H-J please increase weight of the error intervals and / or change colour for improved visibility

Significance

Taken together the manuscript is comprehensively written and the study provides a conclusive analysis of the MO-mediated KD of Vasohibin in zebrafish embryonic development presenting significant novel findings. Known was a generally inhibitory function of Vasohibin on vessel formation and its enzymatic activity as a carboxylpeptidase responsible for tubulin detyrosination, affecting spindle function and mitosis. New is the detailed analysis of the Vasohibin KD on zebrafish trunk vessel formation and the description of a selective impairment of 2nd sprout formation. The manuscript is of interest for vascular biologists.

REFEREES CROSS-COMMENTING

I fully concur with the comments of reviewer #2, all three reviews find that this study is of significant interest to the vascular biology community as the relevance of tubulin detyrosination for developmental angiogenesis has not been investigated. Also all three reviews highlight the potential limitations of the use of splice morpholinos (suggested alternatives include ATG morpholinos and CRIPR mutants), the requirement to provide further evidence for a endothelial cell autonomous defect and the need to clarify some of the data representation.

Reviewer #3:

Kim et al. studied the roles of GATA1 binding and H3K27ac on chromatin interactions associated with binding of the architectural factor CTCF. The authors focused on a TAD and sub-TAD surrounding the human beta-globin locus. The sub-TAD that is flanked by tissue-invariant CTCF sites that form contacts with each other only in erythroid cells. Loss of GATA1 binding at the beta-globin enhancer (LCR) and perturbation of H3K27ac affects some CTCF associated contacts, both involving the sub-TAD as well as the (tissue-invariant) TAD boundaries.

The mechanisms by which tissue specific contacts are established are of great interest in the field. However, while this report provides some correlative data to support a link between H3K27ac and CTCF contacts, the study in its current form is too preliminary and failed to address alternative explanations.

1) A major concern for this study: the MEL/chr11 cell line might not be a suitable model system. The CTCF/Rad21 binding profiles in the MEL/chr11 seem quite different from normal human erythroid cells. Specifically, the signals of CTCF binding at C4, C5 and C6 are almost at the baseline (Fig2A/B and Fig3D/E) in MEL/chr11. However, those sites should have comparable CTCF binding strength with C3 and C7, (e.g. Fig2E). The diminished CTCF binding at the C4-C6 sites might affect the sub-TAD structures in MEL/chr11 cells.

2) Loss of GATA1 binding at HS2 or HS3 affects GATA1 binding at other sites as well. It is also possible that loss of GATA1 binding impairs binding by other nuclear factors that might be involved in sub-TAD formation. The same holds for histone modifications other than H3K27ac. This should be discussed.

3) Fig.2: the correlation between H3K27ac and CTCF contacts does not hold at all examined sites. For example, H3K27ac is reduced also at C7 and C3 that maintain contacts. Same in Fig.3 where increases in H3K27ac is also increased near CTCF sites whose contacts don't increase.

4) Fig.2: total H3 is reduced at C5 to the same extent as asH3. How was H3K27ac normalized?

5) Why were p300 inhibition experiments done in K562 cells? In Fig2E and Fig4D it seems that the H3K27ac signals are not concordant between MEL/chr11 and K562 cells. For example, in Fig2E, the H3K27ac at C3 site is very strong in the MEL/chr11; However, in Fig4D, the H3k27ac signal at C3 site is near baseline. The H3K27ac signals also suggest that the MEL/chr11 cell line is different from normal human erythroid cells and also the leukemia cell line K562. The different H3K27ac profiles at CTCF binding sites confound the interpretation.

6) Fig1E, the interactions between HS5-3'HS1 (C4-C5) were not significantly changed in the HS3/HS2 mutated MEL/chr11 cells. Actually this well known interaction between HS5-3'HS1 was not even detected by 3C in the WT control MEL/chr11 cells, contrary to previous studies. Again, this suggests that the MEL/chr11 cell line is not an ideal system for this study. The reduced interactions between beta-globin and C5/HS5 could be the result of loss of GATA1 binding but not CTCF. For clarity it would be better to plot the 3C data scale to the actual genomic distance.

7) The authors should make clearer distinctions throughout whether they are considering supposedly tissue-invariant CTCF contacts near the TAD boundaries or the tissue-specific sub-TAD. It appears that both can change upon their perturbations.

8) Data mining: "acetylation level was decreased remarkably in the boundaries of longer interactions (over than 100 Kb)" why is this remarkable? How does contact distance inform the role of H3K27ac? What is the correlation genome wide between H3K27ac and CTCF ChIA PET? This could be done using data sets from a variety of cell types.

9) Fig.6A: GATA1 knock down seems to affect H3K27 broadly and not any more extensively (or even less) at CTCF sites.

Reviewer #2:

The authors analyzed long and short range chromatin interactions in erythroid cell lines after perturbation of globin associated enhancers, histone acetylation, GATA1 expression, or CBP/p300 expression. The data show that CTCF binding sites are associated with high levels of H3K27 acetylation in the context of short range interactions. Deletion of GATA1 binding sites in the human LCR in hypersensitive sites 2 or 3 caused a reduction in H3K27ac at nearby CTCF sites and impaired the interaction of those sites, without interfering with the binding of CTCF or the cohesin subunit Rad21. Treatment of cells with TSA or inhibition of CBP/p300 or GATA1 expression had a similar effect.

The manuscript provides interesting and novel observations with respect to GATA1 binding and H3K27ac at nearby CTCF sites. However, the mechanism(s) by which GATA1 modulates H3K27ac levels at CTCF sites remain(s) unknown.

Specific comments.

1) One possible interpretation of the results, which has not been addressed by the authors, is that the reduction in histone acetylation levels may modulate the stiffness of chromatin, which may particularly affect short-range interactions.

2) Figure 3: TSA treatment also affects other histone acetylation events. This should be mentioned. Furthermore, the data in Figure 3D are not consistent with data in Figure 1E with respect to the control cells. In Figure 1E it is shown that the anchor C5 interacts with beta and C6, Figure 3C shows no interactions between these elements.

3) Figure 4: There is no description of how the CBP/p300 depleted cells were generated. Was this a single shRNA? If so, it would be important to repeat the experiment using a second shRNA targeting a different region of the RNA to avoid off-target effects. Furthermore, it would be important to show the CBP/p300 binding pattern across the globin locus and the CTCF sites. Does H3K27ac at the CTCF sites correlate with CBP/p300 binding? These data should be available from K562 cells.

4) Figure 6: there is no description of how the GATA1 depleted cells were generated. Again, was this a single shRNA?

Reviewer #1:

In this study, Kim et al described the GATA1-dependent histone H3K27ac on a subset of CTCF binding sites and CTCF-mediated chromatin interactions in erythroid cells. The authors generated a GATA1 binding site deletion in MEL cells and observed impaired CTCF-mediated interactions around the beta-globin gene cluster. They further modulated H3K27ac by TSA treatment of CBP/p300 knockdown, and noted that altered H3K27ac affected interactions between a subset of CTCF sites. The authors further performed global correlation analysis between public ChIA-PET and H3K27ac in K562 cells with or without GATA1 depletion, and presented evidence supporting a modest effect of H3K27ac at some CTCF sites upon depletion of GATA1 in K562 cells. Based on these findings, the authors concluded that GATA1-dependent H3K27ac mediates erythroid-specific chromatin interactions between CTCF sites.

Overall, this manuscript provides molecular details of the role of GATA1-mediated transcriptional programs in regulating chromatin interactions between CTCF sites. The experiments related to GATA1 binding site mutations were well designed and executed. The authors also make use of existing public datasets to extend the analysis to global scales. However, the current study falls short in providing strong evidence to support several conclusions due to a combination of insufficient data and suboptimal data quality. The overall conclusions that GATA1 regulates chromatin interactions through H3K27ac-mediated effects do not significantly extend our current understanding based on previous studies in the authors' group (Kang et al., 2017 BBA 1860:416) and others (Hsu et al., 2017 Mol Cell 66:102). The lack of replicate experiments for several ChIP-seq studies limited the robustness of relevant conclusions.

In summary, although the current study provides additional insights into the role of GATA1-dependent transcriptional programs in regulating CTCF-mediated chromatin interactions, the overall findings were not developed in sufficient depth in light of previous studies and some conclusions were not sufficiently supported by the evidence.

Major points:

1) The authors previously showed that interactions between CTCF sites associated with sub-TADs are dependent on the binding of GATA1 to LCR enhancers (Kang et al., 2017 BBA 1860:416). In this study, the authors provided further details of the relationship between GATA1 binding, H3K27ac, and CTCF-mediated interactions. These data provide additional insights into the molecular mechanisms of GATA1-mediated effect on chromatin interactions, although they do not add significantly to previous studies from these authors and others (e.g. Hsu et al., 2017 Mol Cell 66:102).

2) To determine the effect of GATA1 binding site KO on CTCF-mediated interactions, the authors performed 3C-qPCR analysis (Fig. 1E). Given the higher resolution and ability to quantitatively contextualize the data obtained from next-gen sequencing based 3C methods, I recommend these be done in the future. Similar for the studies in Figs. 3B-E and 4E.

3) Page 5, line 115-117. It is stated that 'when it is compared to the results from knocking down GATA-1...'; however, no results from GATA-1 knockdown were included for the comparison. These studies would be helpful to determine whether the effect is directly caused by loss of GATA1 binding at the LCR enhancers.

4) It is unclear whether GATA1 site KO also affected H3K27ac signals at other CTCF-bound chromatin regions besides the b-globin gene cluster (C3 to C7). For example, in Fig. 2E, it would be helpful to include C2 and C8 regions in the ChIP-seq track as controls. Additional analyses of global changes of H3K27ac and CTCF binding also will be helpful to determine whether the effect is limited to the b-globin gene cluster.

5) In Fig. 3, to determine the effect of H3K27ac on chromatin interactions, the authors treated the cells with the HDAC inhibitor TSA. A major caveat of these studies is related to the pleiotropic effects of TSA on global H3K27ac and gene expression, thus the indirect effects on chromatin interactions due to altered gene regulation and/or erythroid maturation. This limitation should be discussed.

6) In Fig. 5, it would be important to include analyses by separating CTCF binding sites depending on whether or not they are associated with TAD boundaries. Is there correlation between the range of CTCF-mediated interactions and TAD boundary? These studies will provide a global assessment of the findings based on the beta-globin gene cluster in Figs. 1-4.

7) The effects of GATA1 knockdown on H3K27ac signals at CTCF binding sites at the global or selected regions are marginal (Fig. 6A, C,D, noting the scales of y-axis not started with 0 in Fig. 6C,D). It is unclear whether this is due to suboptimal data quality (see below) or limited effects at a global scale. These results raise questions about the extent to which the GATA1-mediated effect on H3K27ac at CTCF sites and CTCF-mediated interactions in erythroid cells.

8) There are concerns about the quality of several genomic datasets including H3K27ac ChIP-seq that may limit the robustness of relevant analyses. Specifically, the quality of H3K27ac ChIP-seq in K562 cells is suboptimal and no apparent H3K27ac enrichment is noted at the CTCF-binding sites (C3, C4, C6 and C7). Comparing Fig. 2E with Figs. 4D and 6A, the signal-to-noise ratio is highly variable in different experiments. No replicate experiment or quality control analysis was provided.

Preprint Review

This preprint was reviewed using eLife’s Preprint Review service, which provides public peer reviews of manuscripts posted on bioRxiv for the benefit of the authors, readers, potential readers, and others interested in our assessment of the work. This review applies only to version 1 of the manuscript.

Summary

This study investigates the relationship between the binding of the transcription factor GATA-1 and histone acetylation on CTCF sites in the same genomic region. Although the reviewers find this work interesting, they raise concerns about the strength of the novel conclusions that can be drawn at this stage. Specifically, the reviewers are concerned about the strength of the data supporting a specific role of GATA1 binding in regulating CTCF-associated sub-TAD interactions (Reviewer #1 and #3), the lack of technical details on several key experiments (Reviewer #1, #2 and #3), the lack of consideration of alternative interpretations on GATA1- and/or H3K27ac-mediated chromatin regulation (Reviewer #2 and #3), and the use of MEL/chr11 cell line as the experimental model system (Reviewer #1 and #3).

Reviewer #3:

The manuscript by Sahm et al. describes the transcriptomic comparison of breeders and non-breeders in two species of mole rats from genus Fukomys. The remarkable aspect of Fukomys mole rats is that the breeders live significantly longer than workers. The authors produced new breeder couples by pairing animals from different family groups and then compared their transcriptomes to non-breeders of the same age from the original colonies.

There were very few differences identified between breeders and non-breeders. Most transcriptomic changes were confined to gonads and endocrine glands. This is somewhat unsurprising and these organs become active for breeding. The pathways that became activated were related to ribosome biogenesis, protein translation, and MYC signaling, which all reflect physiological activation of these tissues in breeders.

Interestingly, some activation of these signaling pathways was also observed in non-gonadal tissues, which contradicts the common dogma that downregulation of these pathways is associated with lifespan extension. This is a remarkable result that suggests that short-lived model organisms may not correctly reflect signaling effects required for longevity in long-lived species. The authors also point out to higher glucocorticoid levels in workers and speculate that that may be showing Cushing's like syndrome.

Overall, this is an important study of high interest.

Items to address:

1) It is not clear how long were the animals maintained in a breeder status prior to analysis.

2) Figure 3A, the rationale for the comparison of changes in young breeder versus non breeder animals to changes that occur with age over time is unclear.

3) Figure 4, were the changes driven by gonads mainly? How many were non-gonadal? For example, in Figure 2, muscle showed 0 DEGs by status, but in pathway analysis we see upregulation of pathways. Please explain.

4) Table 2: TOR is downregulated in breeders while ribosome processes are upregulated, please explain.

5)The finding of Cushing's syndrome needs more support. Please consider comparing Cushing's related transcriptome changes as a whole to the changes observed in mole rats, otherwise this conclusion may need to be toned down.

Reviewer #2:

Sahm et al. have analyzed gene expression of 22 samples across 15 tissues of Fukomys mole rat, and with these resources, they try to explain why breeders and non-breeders have different lifespans in those species. Below I highlight key aspects of the approach (and related conclusions, or lack thereof) that I believe represent serious issues. One positive note, again, the questions and data of this paper that I believe are highly interesting and important if the author can pursue it in a more focused and solid way.

1) First, in my opinion, the discussion of the paper is not well synthesized, and the content does not help the reader for the aim of study and of what is discussed in the title (e.g., HPA) and abstract. While the entire discussion fails to build logics between HPA stress and aging in mole rats, it tells the story of intervention and hypothesis testing instead, with the last of the results immediately jumping into comparisons of positive selection genes and DEG (It is unclear why it is relevant even one gene is found under positive selection).

2) Given hundreds of samples have been sequenced in the paper, there is no extensive examination of batch effect, which could ultimately, in its present form, put all the results (e.g., DEG analysis, pathways enrichment, multifactors analysis) and conclusions at high risk. Particularly, the PCA analysis has indicated the species, tissue, and the combination of both variables accounted for 98.4 % of the total variance in the data set--it becomes more important to know how the authors have organized the sequencing strategy.

3) Unsophisticated use of enrichment analysis on pathways likely leads to misleading conclusions - a large portion of the analyses presented use an unreasonable approach to identify genes with expression shift across tissues and species to make conclusions that I believe are largely misleading about genes important for longevity. The authors identify pathways that have overall (gene-wide) changes in gene expression as p values looks like a potential indication of expression shift. While this approach MAY be lucky enough to catch a few of these 'true positives", the VAST majority of what it will identify will be pathways with global accumulated changes (which is what the point would be), but rather small pathways or pathways fully of with wired p values or undetectable expression (and thus perhaps some of the least important genes). Then, the extension of these approaches to pathways further muddies the waters of any discussions. The approaches for detecting affected global pathways have been the subject of a very large body of literature, and there are well developed hierarchical/empirical models for testing these hypotheses that the authors have not referred to. And, of course, it is strongly encouraged that the authors formulate a new model/index to reevaluate pathways analysis as the data and experimental design is unique compared to other studies.

4) Last but not the least, small KEGG set, i.e., KEGG, that with very few genes could also be confusing such whole categories analysis. The author should check this potential bias using random sampling 'pseudo-KEGG set' with the same gene number and/or identify a threshold to filter each true KEGG category considered.

Reviewer #1:

This is a primarily descriptive study reporting gene expression differences between breeding and non-breeding individuals of two long-lived Fukomys species. These mole rat species are interesting for this type of analysis because it has been previously shown that breeders live ~2-fold longer than non-breeders. Thus, it may be possible to learn about mechanisms that determine longevity by studying differences between breeders and non-breeders. Although the manuscript is primarily descriptive, there is value in observational data sets such as this, which can be hypothesis generating and can spur future, more mechanistic studies. A lot of analyses are presented - with many different comparisons of differentially expressed genes, GO terms, etc. - but they have not yet produced a lot in terms of true biological insight. As it was largely limited to gene expression, it is also a bit one-dimensional. The data will likely be of interest and value to scientists who study the comparative biology of aging and systems biology of aging, but without additional biological insights derived from the data, the manuscript will probably not capture the interest of a broad scientific audience or even the broader gerontology community.

My largest concerns with the content of this manuscript are related to interpretation of the data. While this type of comparative gene expression analysis can generate hypotheses, it cannot strongly support or refute causal relationships without additional experimentation. The authors repeatedly appear to interpret correlative observations with causation and make claims that overreach the data. The title itself is a good example of this where the authors claim "the HPA stress axis shapes aging rates". There is no direct evidence to support this claim or others similar in nature throughout the manuscript.

Another area of overinterpretation is with respect to the relevance of these findings as a test of the validity of prior aging theories or mechanisms. It seems very likely that this is a somewhat unique evolutionary case where upon the switch from non-breeder to breeder there is a dramatic rewiring of physiology at many levels (transcriptional, translational, post-translational, metabolome, epigenome, etc.). Simply because this switch does not match patterns seen in longevity interventions such as caloric restriction, reduced GH/IGF-1 signaling, or mTOR inhibition does not refute or call into question literature describing potential mechanisms for how those interventions act. It likely simply reflects that this is a different path to achieve longevity.

Specific comments:

• The title is problematic as described above.
• The first sentence in the abstract is problematic as it implies causality between sexual activity/reproduction and longevity. Sexual activity/reproduction are associated with life expectancy. It is interesting to consider whether these two things could perhaps be uncoupled in this animal, as has been shown for reduced fecundity and longevity in invertebrate models.
• The phrase "oppose crucial findings" in the abstract is also problematic. First, the word "crucial" does not seem to make sense in this context. What makes them crucial? Second, it is intuitive and expected, indeed perhaps required, that reproduction would be associated with anabolic processes, and this study does not show causality between the observed changes in these processes and longevity, so it does not actually "oppose" the prior findings in genetic models with reduced IGF-1/GH signaling where lifespan is extended.
• I have a problem with the premise that it is possible to "confirm or falsify" results from cross-species or intra-species studies through the type of approach taken here as implied in line 47. Confirming or falsifying results implies something about the quality of the prior data itself. I assume what the authors mean is confirming or falsifying the underlying hypotheses or assumptions. Even that is questionable, however, since the mechanisms by which longevity are determined could simply be different within species versus across species versus cases like this where you have dramatically different life expectancies in breeders versus non-breeders.
• I thought the discussion of the GH/IGF-1 results was fairly balanced, but I would encourage the authors to consider more deeply the within species versus across species observations in the literature. The evidence for reduced GH/IGF-1 increasing lifespan comes from within species studies and appears to hold true from worms to dogs and likely in humans as well, although the correlation between body size and lifespan is a bit more complicated in humans for obvious reasons. Within species, smaller individuals who have reduced GH/IGF-1 tend to live longer - that's a correlation. In worms, flies, and mice a reduction in growth signaling has been shown to be sufficient - and very likely causal - for enhanced longevity through genetic and pharmacological studies. The comment that these interventions are all performed during development is not exactly true - in mice rapamycin at least works in adulthood and even when only given transiently or intermittently in adulthood. Across species, larger species tend to live longer. Perhaps the mechanisms going from non-breeders to breeders more resemble the evolutionary longevity strategies that have been taken at the species level rather than the mechanism that appear to determine longevity within species. Personally, I don't see this as a contradiction or a controversy within the field.
• I find the "short-lived" versus "long-lived" species argument to be overly speculative and arbitrary. Interventions such as CR, mTOR inhibition, reduced IGF-1, etc. extend lifespan at least from worms to mice, which is a >50-fold difference in lifespan. Mice to people is ~30-fold. Compared to mice, worms are shorter-lived (by a fold difference metric) than mice are compared to humans.
• The authors state that breeders and non-breeders have "massively diverging aging rates", but I think it is important to keep in mind that differences in lifespan do not necessarily imply differences in aging rate. Especially going from the long-lived state to the short-lived state. Perhaps there is good evidence that functional and molecular declines and diseases/pathologies of aging are accelerated in the non-breeders and delayed in the breeders, which would support this assertion, but this is unclear from the manuscript.
• The phrase "unilaterally described as harmful" in the conclusion is simply not true. GH/IGF-1 signaling limits lifespan in worms, flies, and mice but none of the papers cited unilaterally claim that it is harmful. In fact, some of those same papers note that high GH/IGF-1 signaling often confers a selective advantage in terms of faster maturation and reproduction. So, at the species level, this is beneficial. Even at the individual level, high GH/IGF-1 may be associated with better outcomes in the wild where predation is a factor. High GF/IGF-1 in these species is detrimental (only?) in the context of aging/longevity in a relatively safe laboratory environment at the individual level.

Preprint Review

This preprint was reviewed using eLife’s Preprint Review service, which provides public peer reviews of manuscripts posted on bioRxiv for the benefit of the authors, readers, potential readers, and others interested in our assessment of the work. This review applies only to version 2 of the manuscript.

Summary

Mole-rats live in social colonies that contain two breeders and many non-breeding workers. This study looks at gene expression between breeding and non-breeding individuals of two long-lived mole-rat species in the genus Fukomys. The large lifespan difference between breeders and non-breeders has made these animals an important model system for the study of aging, and the data collection reported in this manuscript is very impressive. The paper is descriptive data, with limited data analyses and experimental support for the conclusions. However, given the scale and rarity of the dataset, it will be a valuable resource for the aging research community once the quality and physiological relevance of the data are demonstrated by additional bioinformatic analyses and experimental validations.

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

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

Response to Reviewers

We are grateful to the Reviewers for their thoughtful and helpful assessment of our work. Below we include a point-by-point response to the Reviewers' critiques concerning the interpretation of our results and the power of our system to elucidate key dynamics of fission yeast homology-directed repair (HDR). We appreciate that the Reviewers judged our assay to be a valuable new tool for studying DSB repair in S. pombe. In general, the Reviewers also felt that our data provides new insights into homology search during HDR in fission yeast, including 1) that multiple DSB-donor encounters often precede repair and 2) that the activity of the helicase Rqh1, which dissolves strand invasion structures, alters the kinetics and efficiency of HDR in S. pombe. The Reviewers also raised several concerns with regards to 1) some technical aspects of the experimental approach, 2) the display of the data, and 3) the interpretation of the data. The Reviewers requested additional experiments to address the efficacy of our 5 minute observational time window and the rate of spontaneous damage in the Rqh1 null background, which we are able to provide in a resubmission. We will also clarify experimental details that the Reviewers found confusing in the original text. Lastly, the Reviewers highlighted minor needed figure adjustments that we will incorporate.

Point-by-point Response:

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

Vines et al adapted a system that has been used in S. cerevisiae to study the homology search and homologous recombination repair events by live cell imaging. The authors utilized a system they set up in a fission yeast strain that has a fluorescently tagged endonuclease induced DSB site and monitored RAD52 focus formation in both haploid and diploid cells. The main findings presented are that multiple strand invasion events occur during DSB repair and the role of Rqh1 in promoting these multiple events. For example, cells with Rqh1 loss either have a single strand invasion event that quickly leads to repair or a very long extensive repair time. Overall the results are intriguing with new insight into DSB repair being presented.*

We appreciate the Reviewer’s recognition that our work provides new insights into homology-directed repair (HDR) in fission yeast.

The manuscript would benefit from having another system to help to support or validate the key findings and/or the use of some mutants to help uncouple the different roles of Rad51 and/or Rqh1.

While we agree with the Reviewer that using orthogonal approaches is always desirable, it is not clear what other experimental platform can address the dynamic events with single cell resolution that underlie our observations here; indeed, this was the motivation behind designing this new approach. However, we will provide additional, detailed context to support our findings in the revised manuscript that highlights how orthogonal experimental strategies (e.g. DSB repair outcome assays) already in the literature (e.g. Hope et al., PNAS, 2006) are consistent with our findings. Importantly, however, there is no other population-based system we are aware of that could demonstrate, for example, that Rqh1 shows two different behaviors in individual cells (repair failure and more rapid repair). See more in response to comment 7, below.

\*Major comment:**

1) In Figure 1C, and also Figure 2D, the RAD52 focus observed does not appear in the same location as the LacO cassette. I assume this is because of the way the images are cropped. It would be nice if the authors are saying that the RAD52 focus co-localizes with the inducible DSB location for this to be more readily apparent in the representative images. *

Co-localization events, indicated with the yellow circles, are assessed within raw 3D data that is then flattened for representation in 2D in the figures. For Figure 1C, the two events in the example cell indeed overlap in 3D space. However, in Figure 2D (cells lacking Rad51) we do not observe any colocalization events in the example (and there are no time points annotated with yellow circles).

2) In Figure 3A, the authors claim that the mean time to repair an endonuclease induced DSB is 50 min +/- 20 min. It is unclear whether or not this experiment is done in a diploid strain.

We apologize if we were not clear. All experiments presented in the manuscript are carried out in diploid cells. What varies is whether there is a lac operator integrated at one copy of Chr II (all experiments except Fig. 2A) or on both copies (only Fig. 2A). This will be clarified in the revised text.

3) In Figure 3, whether or not this experiment represents asynchronous cells can greatly influence the timing of DSB repair, as the cell cycle is a huge contributor to HDR repair.

We agree with the Reviewer - the cell cycle has a critical influence on DSB repair mechanism. The diploid fission yeast in which we induced and observed DSBs are indeed asynchronous. However, in fission yeast, which spend over 80% of their cell in G2, we can assess cell cycle by morphology; cytokinesis coincides with the beginning of G2, which then persists until mitotic entry (which is also very obvious from the nuclear shape as visualized by Rad52-mCherry). Moreover, we previously found that HO endonuclease only induces DSBs during S phase (Leland et al., eLife, 2018). Given this, for individual cells we observe site-specific DSBs beginning in late S and early G2 phases and all of our analysis is done at this phase of the cell cycle. These observations are further validated by the observation that an HO-induced DSB undergoes very high rates of gene conversion in fission yeast (Prudden et al, EMBO J., 2003).

4) In Figure 3D, since a major finding of the paper is that there are multiple invasion events, it would be nice to show some representative images of a few cells where multiple pairings occur.

In Supplementary Figure 2A, we provided an example of a cell with multiple encounters between the DSB and donor. This will be more clearly highlighted in the revised text.

5) It is known from Eric Greene's work that RAD51 mediated homology search can do multiple samplings of 8-9 nucleotide segments. Have the authors considered the area around the DSB site and how many potential pairing sites there might be in this region? Is it possible that having a LAC array with repeated segments might be influencing this the pairing since there would be multiple templates?

We acknowledge that the homology of the region surrounding the DSB is important for faithful recognition of a homologous donor and that there could be many pairing sites surrounding our induced DSB after end resection. Such local sampling, however, would not be discernible due to the resolution of the light microscope (>0.2µm). We will address this noteworthy point during our discussion in the revision. Importantly, we placed the lacO array over 3 kb away from the locus where the HO recognition site is integrated on the homologous chromosome to attempt to avoid exactly the Reviewer’s concern.

6) It would aid the reader if there were some picture schematics of what the authors think is occurring throughout the paper in the Figures. Since this is a results/discussion, this approach would be appropriate in lieu of a model figure at the end (which would also be very nice).

We agree that diagrams would aid in communication of our hypotheses and interpretations, and these will be included in the revision.

7) Since the multiple strand invasion events is a major finding of the paper, it is important to test the hypothesis that multiple strand invasion events are occurring a different way. A few ideas would be to examine Lorraine Symington's work on BIR where she observes multiple template switching events (Smith, CE, Llorente, B, Symington, LS (2007) Nature, 447(7140): 102-105) or something analogous to Wolf Heyer's recent study in Cell on template switching that the authors already cited. Another idea is to try a RAD51 mutant. For example, Doug Bishop's group has created a RAD51 mutant that uncouples the homology search from strand exchange, Rad51-II3A mutant (Cloud, V et al (2012) Science, 337(6099): 1222). Perhaps a mutant like this might be able to further support the key finding here.

While our findings share parallels with the works raised by the Reviewer, we would argue that there is a fundamental difference between BIR-type assays and the one we present here, namely that we are visualizing multiple strand invasion events at the homologous chromosome in a normal, high fidelity repair event rather than multiple strand invasion events during BIR, which frequently result in translocations. Moreover, as the two chromosomes are perfectly homologous in our assay, we cannot leverage sequencing to reveal past strand invasion events that took place during HDR. We also cannot, unfortunately, access multiple simultaneous strand invasion events due to the diffraction limit of the light microscope. We concede that it would be informative to further dissect strand invasion using tools such as the Rad51-II3A mutant described in budding yeast in work referenced above by Reviewer #1 and developed in fission yeast by Sarah Lambert’s group (Ait Saada et al., Mol. Cell, 2017). However, with the present limitations on our laboratory access and the timeline necessary to carry out this experiment, we feel this is currently beyond the scope of this work.

8) It is surprising that Rqh1 doesn't have a role in DNA end resection since this is a conserved function from budding yeast to man. Would similar results to what is observed in Figure 4 be observed in a Dna2 or Exo1 mutant?

We acknowledge that Rqh1 orthologs in other organisms (BLM/Sgs1/etc.) have been shown to contribute to DSB end resection. However, previous work from our group indicates that Rqh1 is entirely dispensable for long-range resection in fission yeast (Leland et al., eLife, 2018). Interestingly, in this work we also demonstrated that it is only upon loss of either the 53BP1/Rad9 orthologue Crb2 or Rev7 that Rqh1 is able to compensate for loss of Exo1. It remains unclear whether this is a peculiarity of fission yeast (perhaps because they rely heavily on HR due to extensive time in G2) or if it is a direct consequence of the long G2 itself. Regardless, we demonstrated that cells lacking Exo1 cannot generate sufficient ssDNA tracts to load visualizable Rad52-mCherry (Leland et al., eLife, 2018). Given this, we cannot address this genetic background in this assay. The essential role for Dna2 in replication has also precluded its analysis.

\*Minor comment:**

1) As mentioned in the first line of the abstract, HDR is generally considered error-free as opposed to a pathway that "can be" error-free. *

We acknowledge that HDR (and more specifically HR) is often error-free, but there are notable exceptions such as when a non-homologous donor is utilized for repair or when the polymerases engaged during repair incorporate errors (work from Haber and colleagues). We will expand and clarify this sentence in the revision.

2) In Figure 2D, it is unclear whether this experiment is done in diploid cells. The rest of the figure is in diploid cells but two LacO cassette are not present past the first frame. Please clarify in the legend and/or figure panel. As mentioned above, this is also confusing in Figure 3.

As above, we monitored repair events in diploid cells only – this will be clarified in the revised text.

*Reviewer #1 (Significance (Required)):

The most important advancement in this paper is that multiple strand invasion events occur during homologous recombination and the role of the Rqh1 in this process. Rqh1 is important protein whose mutation is implicated in human disease such as Bloom syndrome and cancer. In addition, misregulation of double-strand break repair and particularly of Rad51 is associate with cancer. Therefore, understanding the basic mechanisms of how Rad51 mediates double-strand break repair and the role of Rqh1 in this process is critical for understanding fundamental aspects of cancer development. * We appreciate the Reviewer’s assessment of the impact of this work.

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

In this study, Vines et al developed a microscopy-based assay to determine the kinetics of a site-specific interhomolog repair event, in living fission yeast cells. They detect efficient homology search and homology-directed repair in the system. They also observe that repair is likely to involve multiple site-specific and Rad51-dependent co-localization events between the DSB and donor sequence, suggesting that efficient inter-homologue repair involves multiple strand invasion events. Loss of the RecQ helicase Rqh1 leads to repair through a single strand invasion event. However, failure to repair is more frequent in rqh1 mutants, which could reflect increased strand invasion at non-homologous sites.

Overall, I find the approach to investigate homology search and homology-directed repair using live cell imaging interesting and potentially very informative. The ability to observe the process in living cells, and with high temporal resolution, complements a variety of previous studies that employ more indirect approaches to invoke similar models. In particular, previous work by the Heyer, Lichten and Hunter laboratories, in budding yeast, has established that Sgs1 promotes non-crossover recombination by acting as a quality control in the maturation of HR intermediates. In this sense, while newly described here for fission yeast, it is not unexpected that homology-directed repair involves multiple strand invasion cycles. In my opinion, the strength of the work is the method/approach, rather than the specific conclusions made (even though I think that it is important to know how fission yeast cells perform homology search).*

We thank the Reviewer for their appreciation of the value that cell biology can bring to the study of homology-directed repair. We wholeheartedly agree that this work is consistent with prior work on Sgs1. With regards to multiple strand invasion cycles, while we agree that there may be many in the field who could be unsurprised by this result, we would argue that 1) demonstrating this by direct visualization of individual DNA repair invents has clear inherent value and 2) many studying homology search itself (or who have modeled homology search in silico, for example) do not incorporate multiple strand invasion cycles in their thinking. Thus, we would argue that this work goes beyond a technical feat and will have impact beyond the approach.

*However, for the reasons detailed below, my general impression is that it isn't clear how robust the method is at delivering unambiguous information on the important questions asked:

1) The authors state that they have developed a system to monitor the 'dynamics and kinetics' of an engineered, inter-homologue repair event. With this in mind, I was expecting a more detailed exploration of the process of homology search. For example, what happens at shorter time scales? Is it possible that by imaging at every 5 minutes many of the events are missed? Could the authors be missing very transient events (especially in rqh1 mutants) by using an inappropriate time scale? *

We acknowledge that it would be ideal to observe DSB repair across a range of time scales in our system. For practical reasons we found it most valuable to choose the 5 minute time window since it was most amenable to observing the entire course of repair as often as possible in an asynchronous cell population (see our response to Reviewer #1’s comment 3 above) while mitigating photobleaching. However, we recognize that we sacrificed time resolution between acquired frames in order to do this. Like the Reviewer, we were also concerned that we were missing transient events due to an inappropriate timescale.

To address this, we acquired additional data in WT cells with greater time resolution with a focus on encounter frequency rather than time to repair (as the overall length of the usable movie that we can obtain is shorter). When imaging WT cells with a site-specific DSB at 2 minute intervals (2.5 times more frequently), we do observe a shift (of ~ 1 encounter per 30 minute window) toward more colocalization events with the donor sequence. We also observe, however, that more sampling leads to an increase in random encounters as revealed by similar analysis of the two lacO control strain as described in the manuscript. These data will be included in the revision and suggest that we may be missing some transient encounter events while using 5 minute time points. As noted by Reviewer #2, this could account for repair in the subset of WT and Rqh1-null cells in which we observed no encounters. We will acknowledge these caveats in the revision but would argue that our data support the conclusion that loss of Rqh1 decreases the number and/or lifetime of strand invasion events.

2) Another point relates to the Rad52 signal/foci, which is central to the study. While it is clear to me what the authors consider to be a focus of Rad52, I am not sure how to interpret what has happens when Rad52 is as enriched throughout the entire nucleus as it is in the repair focus in the still before. For example, Figure 1C, 40 min vs 45 min. How do the authors interpret what is being visualised? Similarly, is the level of colocalization at 90 min really reflecting a specific enrichment of Rad52 at the DSB site? Much more of the Rad52 signal is away from the DSB. In other words, are quantitative criteria being used to assign colocalization events?

As described in our Methods and the text, we used specific criteria to define 1) whether DSBs are site-specific and 2) whether they are colocalized with the donor site. In the images indicated as “contrast adjusted” we have scaled each panel time point individually with respect to the pixel intensities (that is, the least and most intense pixels have been set the same value for each). This strategy allows us to convey relatively dim Rad52-mCherry foci, particularly early after DSB end resection. A consequence of this is that the apparent background for panels in which there is not a strong Rad52-mCherry focus will appear higher, while the background will appear relatively less at time points with a strong Rad52-mCherry focus. For this reason we also present the raw image (found above). It is important to emphasize that when we are applying co-localization criteria, we do so within a 3D stack of images to ensure that the Rad52-mCherry signal and lacO array GFP signal coincide. In 2D representation, however, we understand that this may appear less clear.

In the particular case of the colocalization in Figure 1C at 90 minutes that the Reviewer points out, it is more evident in the 3-D Z stacks that the surrounding mCherry signal apart from the colocalization with the lacO array is due to inhomogeneity in the background signal. Another contribution is that the lacO array signal often becomes delocalized during colocalization events (as evident in that 90 minute time point). Although this is an interesting observation, we are still investigating what activity may explain this response. We will address the caveats of our colocalization analysis more fully in the revision.

3) In the system described here, Rad52 foci form in only ~15% of cells. I think it would be important to rationalise this low number in the manuscript. Moreover, G2 Rad52 foci still form at considerable rates in cells without HO. I think it would be important that the authors provide some explanation on what this might reflect.

There are several considerations that we believe contribute to this observation, which we also documented previously in haploid cells (Leland et al., eLife, 2018). First and foremost, this assay is quite different from endpoint assays that involve induction of HO nuclease because we analyze only those events that happen immediately after additional of uracil to elevate HO endonuclease expression under the control of the urg1 promoter. Combined with the efficient repair of any DSB induced by leaky HO expression (taking less than an hour according to our data), we likely miss events that have already taken place or would take place later in other assay systems. Lastly, it is established that nucleosomes can prevent HO cleavage in its intrinsic role in budding yeast (Laurenson and Rine, Microbiol. Rev., 1992; Haber, Ann. Rev. Genet., 1998); we cannot rule out that cleavage at this particular site is less efficient due to intrinsic nucleosome stability. With respect to spontaneous DNA damage, most of this is short-lived and occurs in S-phase, likely due to replication stress, although we occasionally observe long-lived Rad52 foci in a sub-population of cells – this is in line with previous publications (Coulon et al., MBoC, 2006; Lorenz et al. Mol. Cell Biol., 2009; Sanchez et al., Mol. Cell Biol., 2012; Schonbrun et al., J Biol. Chem., 2013). We will provide a greater explanation of the observed induction rate in the revision.

\*Other issues to consider:**

4) In Figure 2D, the overlay does not show any green. It is possible that the green channel was not overlaid with the pink? *

We apologize for this error and very much appreciate the Reviewer noticing that it is missing from the merged image. This will be corrected.

5) In Figure 2D, the unadjusted images for Rad52 are very sharp. Did the authors perform contrast adjustment in the top panels? If so, this should be indicated. My current impression is that the data was duplicated by mistake.

The Rad52-mCherry data in Figure 2D was labelled correctly and not duplicated. Because cells lacking Rad51 accumulate extensively resected DSBs (and therefore abnormally high levels of Rad52 loading), the intensity of Rad52-mCherry is very high. For simplicity we will remove the contrast-adjusted Rad52-mCherry images in the revision.

6) I don't understand why is the time since nuclear division different is every single figure. For simplicity, it would be much better to start every figure at T=0.

We agree with the Reviewer. In the revision we will normalize all kymographs to begin at t=0 with the exception of the Fig. S1D (where we are visualizing the subsequent division).

*Reviewer #2 (Significance (Required)):

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

In this manuscript, the authors describe a system to monitor an inducible site-specific double-strand break (DSB) and the undamaged homologous locus during homology-directed repair in S. pombe cells. The authors show that the Rad52 focus on the induced DSB is more persistent than spontaneous Rad52 foci that form throughout the cell cycle. The persistent Rad52 focus intermittently colocalizes with the donor sequence labeled with LacI-GFP, reflecting multiple strand invasion events, and this colocalization requires the Rad51 recombinase. The authors report that the time to repair is dependent on the number of strand invasion events (colocalization of Rad52 and homolog), and that the initial distance between the induced DSB and the homolog predicts the time to their first contact, but does not predict the time to repair. Lastly, the authors claim that repair in rqh1Δ cells is bimodal, either failing to repair within the experimental time frame, or being more efficient than WT cells (which often involves a single colocalization event).

**These claims are supported by the data:**

1) Rad52 focus on the induced DSB is more persistent than spontaneous Rad52 foci that form throughout the cell cycle.

2) Multiple colocalization events between Rad52 focus and the donor sequence are frequent, and this colocalization is dependent on Rad51, which reflects multiple strand invasion events.

3) rqh1Δ cells have a lower rate of productive repair compared to WT cells. *

The key concern I have for this section is the noise in Rad52 images. For example, in Fig. 1C at 15 minutes, it looks like there is a Rad52 focus both before and after adjustment but the time point is labeled as not having a Rad52 focus. Conversely, in Fig. 2D at 60 minutes, it looks like there isn't a Rad52 focus but the time point is labeled as having a Rad52 focus. How did the authors determine the presence of a Rad52 focus? Additionally, it is difficult to assess colocalization of Rad52 and LacI-GFP in merged images (hard to see Rad52 focus in Fig. 1C merged and LacI-GFP in Fig. 2D merged).

The criteria that we established to indicate a Rad52-mCherry focus (as annotated by a pink circle and as explained in the Methods) is that it persists for at least three frames (>15 minutes). This was chosen because it is a characteristic of the HO-induced DSB but not of spontaneous DNA damage that occurs frequently during S-phase. Indeed, the numerous, small, and short-lived foci at the 15 minute time point in Fig. 1C referred to by the Reviewer occurs just 15 minutes after nuclear division and is perfectly characteristic of replication stress that is independent of HO endonuclease expression. Thus, the pink circles indicate a specific type of Rad52-mCherry focus that is relevant for the assay. We agree that the Rad52-mCherry focus in Fig. 2D at ~60 minutes is poorly visualized in the flattened image, but would like to emphasize that we assess the foci in the true 3D volume. With regards to the merged images, we will adjust the individual signals to make it easier for the reader to assess colocalization in the revision.

\*These claims are supported by weak data:**

1) The initial distance between the induced DSB and donor sequence predicts the time to their first physical encounter (Line 60). *

We agree with the Reviewer that our word choice (“predicts”) suggests a stronger relationship than is supported by the data. However, we also argue that there is nonetheless a meaningful correlation. We believe this is an important point to make because it supports prior work in budding yeast suggesting that relative position affects donor choice preference. We will edit this language in the revised text.

2) Repair efficiency is dictated by the number of strand invasion events (Line 61-62). Figures 3E and 3F technically have positive correlations that support the authors' claims but there is a lot of noise. I think the data needs to be more robust, especially considering the strong wording used to describe the data. A minor comment on Fig. 3F: why is there a data point with 3.5 encounters?

Again, we agree with the Reviewer that our word choice (“dictate”) is too strong given the data and we will edit the text accordingly. We thank the reviewer for noticing the error in Fig. 3F, which will be corrected.

\*These claims are not supported by the data:**

1) In the absence of Rqh1, successful repair requires a single strand invasion event (Line 63). *

We acknowledge that this is too strong a claim to make based on our data and will amend this language in the revision text. Specifically, and as outlined in our response to Reviewer #2 with regards to our imaging frequency, we will revise the manuscript to state that cells lacking Rqh1 are more likely to repair without a visualized colocalization event and/or they possess shorter lived strand invasion events. Importantly, repair outcome assays indicate that cells lacking Rqh1 display elevated gene conversion rates rather than non-HDR-mediated repair (Hope et al., PNAS, 2006). Thus, we do not expect that the lack of colocalization reflects NHEJ but rather our inability to “catch” the colocalization event with the temporal resolution we can achieve.

2) rqh1Δ cells that complete repair are more efficient than WT cells and often involve a single colocalization event (Line 178-179).

As for the above, we agree that our claim that rqh1Δ cells “often” involve a single colocalization event is too strong a claim based on our data. We will amend this language in the revised text.

Fig. 4A shows an example of a rqh1Δ cell with productive repair but without any colocalization with the homolog, which contradicts the statement that successful repair requires a single strand invasion event in the absence of Rqh1. If the authors interpreted the single continuous presence of Rad52 focus during time-lapse as evidence of a single strand invasion event, then it would nullify using multiple colocalization events as evidence for multiple strand invasion events. In other words, the data in Fig. 3D that clearly displays multiple colocalization events in individual cells during repair can no longer be evidence of multiple strand invasion events since those cells all had one continuous presence of Rad52 focus.

We believe that we understand the confusion that the Reviewer is articulating in their comment and apologize that we have not been clearer in explaining our interpretation. For this site-specific DSB to be repaired, we expect that it must either 1) engage with the homologous chromosome to be repaired by HR/BIR or 2) be repaired through an alternative pathway – at this non-repetitive, resected locus this would likely be a microhomology-mediated (alt-) NHEJ mechanism. However, prior analysis of repair outcome in a model of interhomologue repair in the absence of Rqh1 (Hope et al., PNAS, 2006) demonstrates an increase in cross-over HR events rather than end joining events, arguing that interhomologue HR still dominates (and with increased CO to NCO frequency). We interpret the continuous presence of a Rad52 focus to only reflect that a DSB has been subjected to resection and has not yet been repaired. Taking these two points together, within the lifetime of a Rad52-loaded DSB it can either 1) never colocalize with the donor sequence and fail to repair (as in cells lacking Rad51, Fig. 2D-F) or 2) undergo strand invasion (and therefore colocalization) at least one time (but possibly multiple times) to allow for HDR to occur. However, we agree (and must clarify in the revision) that we often infer that at least one strand invasion event has taken place to support successful HDR when we do not capture the event at our experimental time resolution. Based on the additional data at shorter timescales that we will add to the revised manuscript (as outlined in the response to Reviewer 2, point 1), which demonstrates that we may in some cases be undercounting relevant colocalization events that are too brief to be accurately captured with 5 minute time resolution, we think the most parsimonious explanation is that cells lacking Rqh1 spend less time with the DSB and donor sequence colocalized prior to repair. We agree with the Reviewer, however, that we cannot say whether this reflects a shorter duration of interactions and/or a fewer number of interactions. We will therefore revise the manuscript to acknowledge this point.

Regarding the second claim, I think Fig. 4D only shows rqh1Δ cells with successful repair (since the longest repair time is 55 minutes, but it is not clear from the figure legend). It is not shown how many colocalization events these cells had in Fig. 4D, but there are 16 cells in Fig. 4D while there are only 2 cells with a single encounter (shown in Fig. 4F). With these numbers, it seems like rqh1Δ cells that complete repair are more efficient than WT cells but only few of these cells involve a single colocalization event.

The Reviewer is correct, Figure 4D does indeed show only rqh1Δ cells with the site-specific DSB that successfully repair – this will be clarified in the revision text. As described above in our response to Reviewer #2’s comment 1, it may be that we are missing colocalization events in rqh1Δ DSB cells. However, we would argue that our data do support that, for cells lacking Rqh1 that execute repair, there are fewer and/or shorter-lived colocalization events. Again, this will be made clear in the revision.

Also, how often do Rad52 foci form spontaneously in rqh1Δ cells and what is the duration? This data was provided for WT but not for rqh1Δ.

We agree that increased levels of genome instability (and therefore Rad52 foci) would present an issue – and indeed this has prevented us from analyzing some genetic backgrounds. However, we do not observe a significant increase in spontaneous Rad52-mCherry focus formation in rqh1Δ cells. This data will be included in the revision.

All of the data would have been more supported if the homologous chromosome would have been tagged. Such a configuration would really have helped the interpretation of the rqh1∆ data.

We agree that in theory it would be advantageous to have both copies of the chromosome tagged. Indeed, we attempted to leverage a different version of this experimental system with lacO arrays on both copies while inducing a DSB. However, the complexity of monitoring (and keeping the identity clear) for the two copies presented major challenges. Better would be two distinct arrays – an approach that has been used in budding yeast. However, to date many groups, including ours, have been unable to get TetO-TetR arrays to perform well in fission yeast.

* Reviewer #3 (Significance (Required)):

The significance of this work is the conceptual advance in the field of DNA repair. Homology search is an important process in homology-directed repair and is not fully understood. This study reports time-lapse data on the interaction between a DSB and its donor template during repair and provides insight into the kinetics of homology search. The audience for this manuscript is the field of DNA repair, and to a lesser extent, field of live-cell imaging.*

---

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 describe a system to monitor an inducible site-specific double-strand break (DSB) and the undamaged homologous locus during homology-directed repair in S. pombe cells. The authors show that the Rad52 focus on the induced DSB is more persistent than spontaneous Rad52 foci that form throughout the cell cycle. The persistent Rad52 focus intermittently colocalizes with the donor sequence labeled with LacI-GFP, reflecting multiple strand invasion events, and this colocalization requires the Rad51 recombinase. The authors report that the time to repair is dependent on the number of strand invasion events (colocalization of Rad52 and homolog), and that the initial distance between the induced DSB and the homolog predicts the time to their first contact, but does not predict the time to repair. Lastly, the authors claim that repair in rqh1Δ cells is bimodal, either failing to repair within the experimental time frame, or being more efficient than WT cells (which often involves a single colocalization event).

These claims are supported by the data:

1) Rad52 focus on the induced DSB is more persistent than spontaneous Rad52 foci that form throughout the cell cycle.

2) Multiple colocalization events between Rad52 focus and the donor sequence are frequent, and this colocalization is dependent on Rad51, which reflects multiple strand invasion events.

3) rqh1Δ cells have a lower rate of productive repair compared to WT cells. The key concern I have for this section is the noise in Rad52 images. For example, in Fig. 1C at 15 minutes, it looks like there is a Rad52 focus both before and after adjustment but the time point is labeled as not having a Rad52 focus. Conversely, in Fig. 2D at 60 minutes, it looks like there isn't a Rad52 focus but the time point is labeled as having a Rad52 focus. How did the authors determine the presence of a Rad52 focus? Additionally, it is difficult to assess colocalization of Rad52 and LacI-GFP in merged images (hard to see Rad52 focus in Fig. 1C merged and LacI-GFP in Fig. 2D merged).

These claims are supported by weak data:

1) The initial distance between the induced DSB and donor sequence predicts the time to their first physical encounter (Line 60).

2) Repair efficiency is dictated by the number of strand invasion events (Line 61-62). Figures 3E and 3F technically have positive correlations that support the authors' claims but there is a lot of noise. I think the data needs to be more robust, especially considering the strong wording used to describe the data. A minor comment on Fig. 3F: why is there a data point with 3.5 encounters?

These claims are not supported by the data:

1) In the absence of Rqh1, successful repair requires a single strand invasion event (Line 63).

2) rqh1Δ cells that complete repair are more efficient than WT cells and often involve a single colocalization event (Line 178-179). Fig. 4A shows an example of a rqh1Δ cell with productive repair but without any colocalization with the homolog, which contradicts the statement that successful repair requires a single strand invasion event in the absence of Rqh1. If the authors interpreted the single continuous presence of Rad52 focus during time-lapse as evidence of a single strand invasion event, then it would nullify using multiple colocalization events as evidence for multiple strand invasion events. In other words, the data in Fig. 3D that clearly displays multiple colocalization events in individual cells during repair can no longer be evidence of multiple strand invasion events since those cells all had one continuous presence of Rad52 focus. Regarding the second claim, I think Fig. 4D only shows rqh1Δ cells with successful repair (since the longest repair time is 55 minutes, but it is not clear from the figure legend). It is not shown how many colocalization events these cells had in Fig. 4D, but there are 16 cells in Fig. 4D while there are only 2 cells with a single encounter (shown in Fig. 4F). With these numbers, it seems like rqh1Δ cells that complete repair are more efficient than WT cells but only few of these cells involve a single colocalization event. Also, how often do Rad52 foci form spontaneously in rqh1Δ cells and what is the duration? This data was provided for WT but not for rqh1Δ. All of the data would have been more supported if the homologous chromosome would have been tagged. Such a configuration would really have helped the interpretation of the rqh1∆ data.

Significance

The significance of this work is the conceptual advance in the field of DNA repair. Homology search is an important process in homology-directed repair and is not fully understood. This study reports time-lapse data on the interaction between a DSB and its donor template during repair and provides insight into the kinetics of homology search. The audience for this manuscript is the field of DNA repair, and to a lesser extent, field of live-cell imaging.

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

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

In this study, Vines et al developed a microscopy-based assay to determine the kinetics of a site-specific interhomolog repair event, in living fission yeast cells. They detect efficient homology search and homology-directed repair in the system. They also observe that repair is likely to involve multiple site-specific and Rad51-dependent co-localization events between the DSB and donor sequence, suggesting that efficient inter-homologue repair involves multiple strand invasion events. Loss of the RecQ helicase Rqh1 leads to repair through a single strand invasion event. However, failure to repair is more frequent in rqh1 mutants, which could reflect increased strand invasion at non-homologous sites.

Overall, I find the approach to investigate homology search and homology-directed repair using live cell imaging interesting and potentially very informative. The ability to observe the process in living cells, and with high temporal resolution, complements a variety of previous studies that employ more indirect approaches to invoke similar models. In particular, previous work by the Heyer, Lichten and Hunter laboratories, in budding yeast, has established that Sgs1 promotes non-crossover recombination by acting as a quality control in the maturation of HR intermediates. In this sense, while newly described here for fission yeast, it is not unexpected that homology-directed repair involves multiple strand invasion cycles. In my opinion, the strength of the work is the method/approach, rather than the specific conclusions made (even though I think that it is important to know how fission yeast cells perform homology search). However, for the reasons detailed below, my general impression is that it isn't clear how robust the method is at delivering unambiguous information on the important questions asked:

1) The authors state that they have developed a system to monitor the 'dynamics and kinetics' of an engineered, inter-homologue repair event. With this in mind, I was expecting a more detailed exploration of the process of homology search. For example, what happens at shorter time scales? Is it possible that by imaging at every 5 minutes many of the events are missed? Could the authors be missing very transient events (especially in rqh1 mutants) by using an inappropriate time scale?

2) Another point relates to the Rad52 signal/foci, which is central to the study. While it is clear to me what the authors consider to be a focus of Rad52, I am not sure how to interpret what has happens when Rad52 is as enriched throughout the entire nucleus as it is in the repair focus in the still before. For example, Figure 1C, 40 min vs 45 min. How do the authors interpret what is being visualised? Similarly, is the level of colocalization at 90 min really reflecting a specific enrichment of Rad52 at the DSB site? Much more of the Rad52 signal is away from the DSB. In other words, are quantitative criteria being used to assign colocalization events?

3) In the system described here, Rad52 foci form in only ~15% of cells. I think it would be important to rationalise this low number in the manuscript. Moreover, G2 Rad52 foci still form at considerable rates in cells without HO. I think it would be important that the authors provide some explanation on what this might reflect.

Other issues to consider:

4) In Figure 2D, the overlay does not show any green. It is possible that the green channel was not overlaid with the pink?

5) In Figure 2D, the unadjusted images for Rad52 are very sharp. Did the authors perform contrast adjustment in the top panels? If so, this should be indicated. My current impression is that the data was duplicated by mistake.

6) I don't understand why is the time since nuclear division different is every single figure. For simplicity, it would be much better to start every figure at T=0.

Significance

see above.

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

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

Vines et al adapted a system that has been used in S. cerevisiae to study the homology search and homologous recombination repair events by live cell imaging. The authors utilized a system they set up in a fission yeast strain that has a fluorescently tagged endonuclease induced DSB site and monitored RAD52 focus formation in both haploid and diploid cells. The main findings presented are that multiple strand invasion events occur during DSB repair and the role of Rqh1 in promoting these multiple events. For example, cells with Rqh1 loss either have a single strand invasion event that quickly leads to repair or a very long extensive repair time. Overall the results are intriguing with new insight into DSB repair being presented. The manuscript would benefit from having another system to help to support or validate the key findings and/or the use of some mutants to help uncouple the different roles of Rad51 and/or Rqh1.

Major comment:

1) In Figure 1C, and also Figure 2D, the RAD52 focus observed does not appear in the same location as the LacO cassette. I assume this is because of the way the images are cropped. It would be nice if the authors are saying that the RAD52 focus co-localizes with the inducible DSB location for this to be more readily apparent in the representative images.

2) In Figure 3A, the authors claim that the mean time to repair an endonuclease induced DSB is 50 min +/- 20 min. It is unclear whether or not this experiment is done in a diploid strain.

3) In Figure 3, whether or not this experiment represents asynchronous cells can greatly influence the timing of DSB repair, as the cell cycle is a huge contributor to HDR repair.

4) In Figure 3D, since a major finding of the paper is that there are multiple invasion events, it would be nice to show some representative images of a few cells where multiple pairings occur.

5) It is known from Eric Greene's work that RAD51 mediated homology search can do multiple samplings of 8-9 nucleotide segments. Have the authors considered the area around the DSB site and how many potential pairing sites there might be in this region? Is it possible that having a LAC array with repeated segments might be influencing this the pairing since there would be multiple templates?

6) It would aid the reader if there were some picture schematics of what the authors think is occurring throughout the paper in the Figures. Since this is a results/discussion, this approach would be appropriate in lieu of a model figure at the end (which would also be very nice).

7) Since the multiple strand invasion events is a major finding of the paper, it is important to test the hypothesis that multiple strand invasion events are occurring a different way. A few ideas would be to examine Lorraine Symington's work on BIR where she observes multiple template switching events (Smith, CE, Llorente, B, Symington, LS (2007) Nature, 447(7140): 102-105) or something analogous to Wolf Heyer's recent study in Cell on template switching that the authors already cited. Another idea is to try a RAD51 mutant. For example, Doug Bishop's group has created a RAD51 mutant that uncouples the homology search from strand exchange, Rad51-II3A mutant (Cloud, V et al (2012) Science, 337(6099): 1222). Perhaps a mutant like this might be able to further support the key finding here.

8) It is surprising that Rqh1 doesn't have a role in DNA end resection since this is a conserved function from budding yeast to man. Would similar results to what is observed in Figure 4 be observed in a Dna2 or Exo1 mutant?

Minor comment:

1) As mentioned in the first line of the abstract, HDR is generally considered error-free as opposed to a pathway that "can be" error-free.

2) In Figure 2D, it is unclear whether this experiment is done in diploid cells. The rest of the figure is in diploid cells but two LacO cassette are not present past the first frame. Please clarify in the legend and/or figure panel. As mentioned above, this is also confusing in Figure 3.

Significance

The most important advancement in this paper is that multiple strand invasion events occur during homologous recombination and the role of the Rqh1 in this process. Rqh1 is important protein whose mutation is implicated in human disease such as Bloom syndrome and cancer. In addition, misregulation of double-strand break repair and particularly of Rad51 is associate with cancer. Therefore, understanding the basic mechanisms of how Rad51 mediates double-strand break repair and the role of Rqh1 in this process is critical for understanding fundamental aspects of cancer development.

Preprint Review

This preprint was reviewed using eLife’s Preprint Review service, which provides public peer reviews of manuscripts posted on bioRxiv for the benefit of the authors, readers, potential readers, and others interested in our assessment of the work. This review applies only to version 2 of the manuscript.

Summary:

VolcaNoseR is a plotting tool to make volcano plots from transcriptomics and proteomics datasets. It is easy to use, intuitive, interactive and reactive. However, VolcaNoseR does not provide any statistical analysis of the data, and it is therefore of limited applicability as it is likely that a user that is able to handle and analyse this type of data, is likely to be able to produce a scatter plot. The bigger hurdles with this type of analysis that could lead to new biological results are centered around the correct statistical analysis of the data, problems that the software does not address.

Preprint Review

This preprint was reviewed using eLife’s Preprint Review service, which provides public peer reviews of manuscripts posted on bioRxiv for the benefit of the authors, readers, potential readers, and others interested in our assessment of the work. This review applies only to version 1 of the manuscript.

Summary:

This manuscript describes the mechanism by which the transcriptional repressor ZBTB18 regulates the expression of sterol regulatory-element binding proteins (SREBP), transcription factors that control the expression of enzymes involved in fatty acid and cholesterol biosynthesis,in glioblastomas. The connection between core transcriptional regulation and tumor metabolism is an area of current interest. This manuscript uses immunoprecipitation and mass spectrometry to identify the well-characterized transcriptional corepressor CTBP2 as a new ZBTB18 interacting protein, which the authors show overlap in binding at many gene promoters. They further show that CTBP2 activates, while ZBTB represses the expression of some SREBP genes. The reciprocal regulation of ZBTB18 and CTBP2 has potential value in understanding the functional regulation of lipid biology. However, reviewers raised substantial concerns with the studies, as described below.

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 comments and outline below how we plan to address them.

---

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). The authors here describe a method to modify bacterial artificial chromosomes (BAC) harbouring gene loci from eukaryotes. When wanting to modify a BAC an antibiotic selection cassette is often included alongside the desired mutation/modification to increase the number of successful recombinants in E.coli. Traditionally, this is removed in a second recombination process to leave only the desired modification. The novelty in the procedure described herein is to add a synthetic intron consensus sequence around the selection cassette, which eliminates the need for the subsequent removal of the antibiotic cassette from the BAC before transfection into mammalian cells, saving time and resources. The technique is clever in its simplicity and appears to function for a number of gene loci. The authors validated the correct functioning of the modified BACs for a number of genes using three main assays - transcript level, protein level and localisation. **Major comments:** *Are the key conclusions convincing?* The conclusion that the method described generates functional modified BACs is valid. *Should the authors qualify some of their claims as preliminary or speculative, or remove them altogether?* While the method is successfully employed in this study, its efficiency is not quantified in relation to the state-of-the-art as described in the introduction. One assumes it would be more efficient, but this has not been tested empirically in the paper. Does the inclusion of the synthetic intron sequence have an effect on the efficiency of modifying BACs compared to a more typical two-step positive/negative antibiotic selection cassette? *

• *

This is a good point that we did not directly address. In general, the efficiency is similar to that of integrating any cassette with selectable marker, as has been published (Poser et al 2008), and therefore also higher than the two-step counterselection method, which requires such a cassette integration in the first step alone. We will include new data specifically addressing the efficiency of our new method (see specifics below)

The functionality of this approach rests entirely on the ability of the target cell to correctly splice out the synthetic intron. The authors are aware of this potential problem as highlighted in the lines below, but do not make efforts to explicitly test splicing. On lines 224-225, the authors state "We cannot exclude that a small portion of synthetic introns within individual cells are misspliced". On lines 230-231 it is stated that "mis-spliced mRNAs are probably minimal and degraded by nonsense-mediated decay". On lines 215-217, the authors describe an "investigation of transgenic lines at the single-cell level" that suggests "the synthetic intron is correctly spliced out in all the cells of the population". How do the authors reach this conclusion? U2OS and HeLa cells are considered very "robust" and may not show detectable consequences when stressed with an increased level of nonsense-mediated decay. Further, many genes maintain a high level of expression that buffers them against small changes in transcription/splicing. The synthetic intron might have a bigger impact on more tightly regulated genes, so assessing the splicing rate would be essential if the authors wish to advocate their technique as generally applicable.

• *

We will assay for splicing efficiency as outlined below.

The ability of the synthetic intron to be removed from final transcripts depends on functioning splicing machinery. The authors might emphasise this issue, as spliceosome mutations are important fields of study and might not be compatible with this method.

• *

We can add this in the text

The authors used un-directed integration of each BAC under study. Therefore, it is hard to assess what effect the synthetic intron has, as the authors only ever assess the downstream levels of the correctly spliced, translated and localised protein. The authors themselves state that this can lead to clonal variations in expression of up to 2-fold and on line 250 that this variation "could compensate for synthetic intron effects", but make no effort to test this. Again, lines 267-268 highlight the potential dangers of potential effects of the synthetic introns, but do not test these. \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.* If not already performed, a large number of bacterial colonies should be screened for the correct modification and frequency of correct ones reported. This frequency - reported for at least three different modifications - would estimate what sort of efficiency this method provides. The modified region of each BAC should be sequenced and the results reported. The rate of exactly modified clones is important, in case of spontaneous or low fidelity integration of the antibiotic cassette. The percentage of transcripts that have the synthetic intron correctly spliced out should be measured for some of the BAC constructs used in the study. A direct head-to-head comparison of this newer method compared to other techniques, or even the authors' own previous two-step approach is necessary to assess the benefits of this method. Preferably, the experiment would be run in parallel with and without antibiotic selection applied, to show that it drastically improves chances of finding a correct clone. *

We will generate 3 new mutations in BACs and analyze both the efficiency of integration by PCR and accuracy via sequencing. In practice, we have observed that the efficiency is similar to any other cassette integration, such as a GFP tag (Poser et al Nature Methods 2008) or a counterselection cassette (Bird et al Nature Methods 2012) (80-90%). Integrating a mutation via the second step of the counterselection method introduces a further 20% decrease in efficiencies on average.

\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.* Repeating the transformation of the BAC and targeting cassette and assessing the recombination efficiency and sequencing should only require existing reagents and take less than a week or two to complete. Quantitative RT-PCR to assess the percentage of transcripts that have the synthetic intron spliced out would take a little more work. However, this should not be a considerable investment in time or resources for a standard microbiology laboratory and could be completed within a few weeks using modern techniques, such as that described in Londoño et al. 2016. Repeating all the experiments in parallel would be considerable work and would only be strictly necessary if the authors wish to emphasise the benefits of their method over the many others already in wide use. *

• *

We will use quantitative PCR to estimate the fraction of transcripts that correctly splice out the artificial intron for two clonal cell lines characterized in the study: RNAi-resistant AurA-GFP (Fig 4), and GTSE1-14A (newly introduced; see below). While the exact method described in Londoño et al 2016 will not be applicable due to the larger size of the artificial intron, we believe we can adapt it to detect different splicing events.

\Are the data and the methods presented in such a way that they can be reproduced?* Barring the omission of Table S1, which presumably includes exact information on the BACs modified and sequences used etc., there is sufficient other data and methods to allow the experiments to be repeated. Targeting the ESI procedure to the middle of exons is likely to have a bigger impact for smaller exons as the authors mention on lines 99-100. Making it clear which exon sizes for each gene were successfully targeted in this study would help give some idea of how significant a problem this might be. Perhaps Table S1 contains this information, but it was not provided. It would also help reviewers check the design strategies. *

We apologize for inadvertently failing to upload Table S1 on bioRxiv. It has been uploaded now as part of this submission process. This table indeed contains BAC and target sequence information, including the size of the targeted exon (and the 2 “new” resulting exons). Targeted exons range in size from 138bp to 1537bp, and “new” exons are as small as 48bp.

\Are the experiments adequately replicated and statistical analysis adequate?* The replication and statistically analysis of the data as presented appear adequate. Figure Legends should state the statistic used to generate error bars. *

This will be updated

\*Minor comments:** Specific experimental issues that are easily addressable. Are the promoters used in the vectors described universally functional? For example, is the PGK promoter functional in yeast? *

• *

The PGK promoter contained in the cassettes is a mammalian promoter, which has also been reported to work in flies.

\Are prior studies referenced appropriately?* The manuscript may benefit from the referencing of BAC modification techniques from a wider variety of groups, such as those using CRISPR-guided recombineering (Pyne et al. 2015). *

We will add citations of more techniques

\Are the text and figures clear and accurate?* The body text is very clear save minor typographical or grammatical errors. Regarding figures, some of the coloured text in Figure 1 is somewhat illegible when printed in grayscale. Line 278 - The acronyms LAP and NLAP are not defined/explained. Antibody section starting Line 282 may fit better next to Western Blot section. Figure 2C - The blot images would benefit from arrows to indicate expected sizes of proteins. Figure 3A - the graph may benefit from a dashed line at 100% to highlight that values are normalised to controls. Figure 4 - The differences between panels B & C are unclear. Figure 4E - The legend could provide a little more detail on cell cycle stage/status of the captured cells. *

All of the above will be addressed accordingly

\Do you have suggestions that would help the authors improve the presentation of their data and conclusions?* Lines 23-27 are somewhat unclear and feel out of context. Perhaps the authors could clarify this as a further advantage of using BACs instead of endogenous gene modifications. *

Thanks for the input, we will clarify this.

While not affecting the factual content of the paper, I would advocate that the authors format the method described in Figure S3 into a more detailed text based layout similar to that seen in a typical Nature Methods article. However, this may depend on the format required by any eventual publishing journal.

• *

We prefer the graphical protocol, but will discuss whether to add a text protocol with the journal editor.

That all of the work the paper was carried out in human cell lines and using human genes is a further caveat, but the authors admit this in the discussion and one would assume that most mammalian cells would respond similarly in their ability to splice out the synthetic intron. Reviewer #1 (Significance (Required)): \Describe the nature and significance of the advance (e.g. conceptual, technical, clinical) for the field.* This work is a formal description of a newer method that could be useful for many of those employing bacterial artificial chromosomes in numerous studies, such as gene regulation. *Place the work in the context of the existing literature (provide references, where appropriate).* This work builds on methodology previously published by the authors - a counter-selection two-step procedure (Bird et al. 2011). It sets out to formally describe a method merely mentioned as "BAC intronization" in a later paper by some of the authors (Zheng et al. 2014). Other alternative one-step procedures are also available, but present a different set of challenges (Lyozin et al. 2014). Some newer approaches, such as those using CRISPR-guided recombineering (Pyne et al. 2015) or systems that combine CRISPR and positive/negative selection cassettes (Wang et al. 2016) may be slightly more efficient, but are also more complex in their design. Bird et al. 2011 DOI: 10/dv776q Pyne et al. 2015 DOI: 10/f7jx92 Wang et al. 2016 DOI: 10/f89db5 Zheng et al. 2014 DOI: 10/f5pkr6 *State what audience might be interested in and influenced by the reported findings.* As a technology paper this work should have interest from a broad field of research. While the use of BACs could sometimes be considered more traditional in light of the explosion in CRISPR-based genome editing capabilities, it is definitely seeing a resurgence as the limitations of CRISPR in modifying large regions of genome become more apparent. Therefore, technologies that accelerate the modification of BACs could prove increasingly useful. As category of audience, all those involved in significant recombineering or gene/genome engineering would potentially benefit. *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.* Synthetic genomics, synthetic biology, cancer cell biology, gene and genome engineering REFEREES CROSS COMMENTING I would agree with reviewer two's assessment that we both view the paper in a similar light. Reviewer #2 (Evidence, reproducibility and clarity (Required)): This is a methods-focused paper that presents a strategy to efficiently introduce mutations into a bacterial artificial transgene using synthetic introns. BAC-based methods have been an effective strategy for introducing trans genes into human cells to achieve near-endogenous expression, including extensive work from these authors. However, generating mutations and changes within the internal coding sequence presents some challenges for how to target these mutations and select for the mutated form. Here, the authors describe a way to overcome this by introducing synthetic introns into an adjacent sequence. This allows them to introduce a selectable marker and conduct the molecular biology without creating complications downstream for the functionality of the protein. This method is carefully described and presented. The authors also provide clear validation by using this to create RNAi-resistant versions of multiple different mitotic factors as well as creating targeted mutants that alter the functional properties of a protein. This work clearly takes advantage of other ongoing studies from these labs (including mutants and cell lines that appear to also have been described elsewhere), but the ability to combine these in a single paper and clearly describe the method provides a helpful advance and validation. Based on the description and data presented, I think that things are clear and carefully validated. As such, I do not have technical comments or concerns and I would be comfortable with this paper appearing in an appropriate journal in its present form. Reviewer #2 (Significance (Required)): This is a solid methods paper, but for considering the nature of the impact and significance of this paper, there are several things to note: 1.The BAC-based method does appear to be a powerful and effective strategy. However, beyond the work of Mitocheck and the authors that are part of this paper, this has not seen widespread adoption. It is possible that this current method may increase its usage due to the value of the targeted mutations within the coding sequence, but at present it is not a broadly used strategy. *

We agree that using BACs as transgenes has not seen widespread adoption as a tool on the broader cell biology community (although certainly beyond members of the Mitocheck consortium). This is likely because many erroneously think that it is a technique for specialist laboratories. We are trying to change this! For reasons outlined below, there is still an increasing desire for conditional analysis of mutated genes under physiological expression/regulation frequently not attainable via directed Cas9-based mutation. A major aim of this paper is thus to further simplify the methods for generating modified BAC transgenes.

2.This BAC-based approach (and also RNAi) are becoming increasingly replaced by the use of CRISPR/Cas9 genome editing. The absence of Cas9-based strategies in this paper limits the potential impact and reach of this paper. The authors do mention the possibility of using a similar synthetic intron strategy for use with Cas9 in the Discussion, and appear to have conducted some experiments. If possible, it would substantially increase the value of this paper if this data and strategy were also included in the Results section (acknowledging that this may still be a work in progress).

While some uses of BAC transgenes are in some cases better replaced by CRISPR/Cas9 techniques (i.e. GFP tagging), there are several occasions where using BACs are preferable: As stated in the text, RNAi-resistant BACs allow for conditional analysis of recessive mutations. Mutations in essential genes that are lethal will prevent growth and recovery of viable cells if integrated into the genome via Cas9. Additionally, deleterious mutations are prone to accumulate suppressive changes in chromosome integrity or gene expression during the procedure of selecting and expanding Cas9-modified cells for analysis, particularly in the genomically instable cancer cell lines frequently employed.

We use both BACs and CRISPR/Cas9 in our lab according to our needs.

We do have an ongoing project to apply this intronization technique to enable more efficient selection of CRISPR/Cas9 integrations. Preliminary results suggest that it works to allow selection of point mutations, but it is still being optimized, including a redesign of the cassette, and is not ready for publication.

3.The method is solid and well-validated, but there are no new results or insights presented in this paper from the work that is described (this is fine, just commenting for considering the right journal fit).

As “biological insights” gained as a result of this technique we had cited a couple studies that made use of the technique already (to functionally analyze a microcephaly-associated mutation in the centriolar protein CPAP at the single cell level in HeLa cells and neural progenitor cells (Zheng et al 2014, Gabirel et al 2016)). As a response to this critique to include “new biology” in this paper, we will add new unpublished data investigating a specific question: Is the cell-cycle-regulated disruption of the EB1-GTSE1 (microtubule plus-end tracking proteins) interaction in mitosis required for chromosome segregation fidelity? We have generated a GTSE1 mutant with 14 phosphosites mutated to alanine using this technique. We will present the effect on chromosome segregation.

REFEREES CROSS COMMENTING It appears that both reviewers are largely on the same page regarding this paper.

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

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

This is a methods-focused paper that presents a strategy to efficiently introduce mutations into a bacterial artificial transgene using synthetic introns. BAC-based methods have been an effective strategy for introducing trans genes into human cells to achieve near-endogenous expression, including extensive work from these authors. However, generating mutations and changes within the internal coding sequence presents some challenges for how to target these mutations and select for the mutated form. Here, the authors describe a way to overcome this by introducing synthetic introns into an adjacent sequence. This allows them to introduce a selectable marker and conduct the molecular biology without creating complications downstream for the functionality of the protein.

This method is carefully described and presented. The authors also provide clear validation by using this to create RNAi-resistant versions of multiple different mitotic factors as well as creating targeted mutants that alter the functional properties of a protein. This work clearly takes advantage of other ongoing studies from these labs (including mutants and cell lines that appear to also have been described elsewhere), but the ability to combine these in a single paper and clearly describe the method provides a helpful advance and validation.

Based on the description and data presented, I think that things are clear and carefully validated. As such, I do not have technical comments or concerns and I would be comfortable with this paper appearing in an appropriate journal in its present form.

Significance

This is a solid methods paper, but for considering the nature of the impact and significance of this paper, there are several things to note:

1.The BAC-based method does appear to be a powerful and effective strategy. However, beyond the work of Mitocheck and the authors that are part of this paper, this has not seen widespread adoption. It is possible that this current method may increase its usage due to the value of the targeted mutations within the coding sequence, but at present it is not a broadly used strategy.

2.This BAC-based approach (and also RNAi) are becoming increasingly replaced by the use of CRISPR/Cas9 genome editing. The absence of Cas9-based strategies in this paper limits the potential impact and reach of this paper. The authors do mention the possibility of using a similar synthetic intron strategy for use with Cas9 in the Discussion, and appear to have conducted some experiments. If possible, it would substantially increase the value of this paper if this data and strategy were also included in the Results section (acknowledging that this may still be a work in progress).

3.The method is solid and well-validated, but there are no new results or insights presented in this paper from the work that is described (this is fine, just commenting for considering the right journal fit).

REFEREES CROSS COMMENTING

It appears that both reviewers are largely on the same page regarding this paper.

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

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

Summary:

Provide a short summary of the findings and key conclusions (including methodology and model system(s) where appropriate). The authors here describe a method to modify bacterial artificial chromosomes (BAC) harbouring gene loci from eukaryotes. When wanting to modify a BAC an antibiotic selection cassette is often included alongside the desired mutation/modification to increase the number of successful recombinants in E.coli. Traditionally, this is removed in a second recombination process to leave only the desired modification. The novelty in the procedure described herein is to add a synthetic intron consensus sequence around the selection cassette, which eliminates the need for the subsequent removal of the antibiotic cassette from the BAC before transfection into mammalian cells, saving time and resources. The technique is clever in its simplicity and appears to function for a number of gene loci. The authors validated the correct functioning of the modified BACs for a number of genes using three main assays - transcript level, protein level and localisation.

Major comments:

Are the key conclusions convincing?

The conclusion that the method described generates functional modified BACs is valid.

Should the authors qualify some of their claims as preliminary or speculative, or remove them altogether?

While the method is successfully employed in this study, its efficiency is not quantified in relation to the state-of-the-art as described in the introduction. One assumes it would be more efficient, but this has not been tested empirically in the paper. Does the inclusion of the synthetic intron sequence have an effect on the efficiency of modifying BACs compared to a more typical two-step positive/negative antibiotic selection cassette? The functionality of this approach rests entirely on the ability of the target cell to correctly splice out the synthetic intron. The authors are aware of this potential problem as highlighted in the lines below, but do not make efforts to explicitly test splicing. On lines 224-225, the authors state "We cannot exclude that a small portion of synthetic introns within individual cells are misspliced". On lines 230-231 it is stated that "mis-spliced mRNAs are probably minimal and degraded by nonsense-mediated decay". On lines 215-217, the authors describe an "investigation of transgenic lines at the single-cell level" that suggests "the synthetic intron is correctly spliced out in all the cells of the population". How do the authors reach this conclusion? U2OS and HeLa cells are considered very "robust" and may not show detectable consequences when stressed with an increased level of nonsense-mediated decay. Further, many genes maintain a high level of expression that buffers them against small changes in transcription/splicing. The synthetic intron might have a bigger impact on more tightly regulated genes, so assessing the splicing rate would be essential if the authors wish to advocate their technique as generally applicable. The ability of the synthetic intron to be removed from final transcripts depends on functioning splicing machinery. The authors might emphasise this issue, as spliceosome mutations are important fields of study and might not be compatible with this method. The authors used un-directed integration of each BAC under study. Therefore, it is hard to assess what effect the synthetic intron has, as the authors only ever assess the downstream levels of the correctly spliced, translated and localised protein. The authors themselves state that this can lead to clonal variations in expression of up to 2-fold and on line 250 that this variation "could compensate for synthetic intron effects", but make no effort to test this. Again, lines 267-268 highlight the potential dangers of potential effects of the synthetic introns, but do not test these.

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.

If not already performed, a large number of bacterial colonies should be screened for the correct modification and frequency of correct ones reported. This frequency - reported for at least three different modifications - would estimate what sort of efficiency this method provides. The modified region of each BAC should be sequenced and the results reported. The rate of exactly modified clones is important, in case of spontaneous or low fidelity integration of the antibiotic cassette. The percentage of transcripts that have the synthetic intron correctly spliced out should be measured for some of the BAC constructs used in the study. A direct head-to-head comparison of this newer method compared to other techniques, or even the authors' own previous two-step approach is necessary to assess the benefits of this method. Preferably, the experiment would be run in parallel with and without antibiotic selection applied, to show that it drastically improves chances of finding a correct clone.

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.

Repeating the transformation of the BAC and targeting cassette and assessing the recombination efficiency and sequencing should only require existing reagents and take less than a week or two to complete. Quantitative RT-PCR to assess the percentage of transcripts that have the synthetic intron spliced out would take a little more work. However, this should not be a considerable investment in time or resources for a standard microbiology laboratory and could be completed within a few weeks using modern techniques, such as that described in Londoño et al. 2016. Repeating all the experiments in parallel would be considerable work and would only be strictly necessary if the authors wish to emphasise the benefits of their method over the many others already in wide use.

Are the data and the methods presented in such a way that they can be reproduced?

Barring the omission of Table S1, which presumably includes exact information on the BACs modified and sequences used etc., there is sufficient other data and methods to allow the experiments to be repeated. Targeting the ESI procedure to the middle of exons is likely to have a bigger impact for smaller exons as the authors mention on lines 99-100. Making it clear which exon sizes for each gene were successfully targeted in this study would help give some idea of how significant a problem this might be. Perhaps Table S1 contains this information, but it was not provided. It would also help reviewers check the design strategies.

Are the experiments adequately replicated and statistical analysis adequate?

The replication and statistically analysis of the data as presented appear adequate. Figure Legends should state the statistic used to generate error bars.

Minor comments:

Specific experimental issues that are easily addressable. Are the promoters used in the vectors described universally functional? For example, is the PGK promoter functional in yeast?

Are prior studies referenced appropriately?

The manuscript may benefit from the referencing of BAC modification techniques from a wider variety of groups, such as those using CRISPR-guided recombineering (Pyne et al. 2015).

Are the text and figures clear and accurate?

The body text is very clear save minor typographical or grammatical errors. Regarding figures, some of the coloured text in Figure 1 is somewhat illegible when printed in grayscale.

Line 278 - The acronyms LAP and NLAP are not defined/explained.

Antibody section starting Line 282 may fit better next to Western Blot section.

Figure 2C - The blot images would benefit from arrows to indicate expected sizes of proteins.

Figure 3A - the graph may benefit from a dashed line at 100% to highlight that values are normalised to controls.

Figure 4 - The differences between panels B & C are unclear.

Figure 4E - The legend could provide a little more detail on cell cycle stage/status of the captured cells.

Do you have suggestions that would help the authors improve the presentation of their data and conclusions?

Lines 23-27 are somewhat unclear and feel out of context. Perhaps the authors could clarify this as a further advantage of using BACs instead of endogenous gene modifications.

While not affecting the factual content of the paper, I would advocate that the authors format the method described in Figure S3 into a more detailed text based layout similar to that seen in a typical Nature Methods article. However, this may depend on the format required by any eventual publishing journal. That all of the work the paper was carried out in human cell lines and using human genes is a further caveat, but the authors admit this in the discussion and one would assume that most mammalian cells would respond similarly in their ability to splice out the synthetic intron.

Significance

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

This work is a formal description of a newer method that could be useful for many of those employing bacterial artificial chromosomes in numerous studies, such as gene regulation.

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

This work builds on methodology previously published by the authors - a counter-selection two-step procedure (Bird et al. 2011). It sets out to formally describe a method merely mentioned as "BAC intronization" in a later paper by some of the authors (Zheng et al. 2014). Other alternative one-step procedures are also available, but present a different set of challenges (Lyozin et al. 2014). Some newer approaches, such as those using CRISPR-guided recombineering (Pyne et al. 2015) or systems that combine CRISPR and positive/negative selection cassettes (Wang et al. 2016) may be slightly more efficient, but are also more complex in their design.

Bird et al. 2011 DOI: 10/dv776q

Pyne et al. 2015 DOI: 10/f7jx92

Wang et al. 2016 DOI: 10/f89db5

Zheng et al. 2014 DOI: 10/f5pkr6

State what audience might be interested in and influenced by the reported findings.

As a technology paper this work should have interest from a broad field of research. While the use of BACs could sometimes be considered more traditional in light of the explosion in CRISPR-based genome editing capabilities, it is definitely seeing a resurgence as the limitations of CRISPR in modifying large regions of genome become more apparent. Therefore, technologies that accelerate the modification of BACs could prove increasingly useful. As category of audience, all those involved in significant recombineering or gene/genome engineering would potentially benefit.

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.

Synthetic genomics, synthetic biology, cancer cell biology, gene and genome engineering

Reviewer #3

The authors address an important open question in decision neuroscience: how do perceptual and metacognitive choices arise from the same sensory input? They perform an experiment that asks observers to report both their percept and associated confidence, and then investigate the role of history biases in both perceptual and metacognitive decisions. They find support for (well-established, e.g. https://elifesciences.org/articles/11946 ) second-order theories of metacognition, which state that metacognitive reports arise from more than just the information used to guide first-order perceptual decisions.

The methods and statistics are rigorous, and the authors have thoroughly reviewed the relevant literature on the topic. The mutual information analysis is a nice way to quantify the size of different biasing factors. However, the paper presents only limited mechanistic insight into history biases, decision processes or metacognition. It would help to clarify what the main insights are, and how they change our understanding of decision-making and its underlying computational nature.

Technical suggestions/comments:

-A recent study (https://elifesciences.org/articles/49834 ) shows the importance of correcting for slow fluctuations within a session, to accurately estimate history biases that are due to the previous trial itself. It would be great to see if the results hold after such a correction, to ensure that the effects are not due to slow fluctuations in choice and metacognitive reports within a session.

-The answer to whether perceptual and metacognitive decisions arise from the same or distinct processes may depend strongly on the nature of the confidence report. Would the same conclusions hold if choice and confidence were indicated with the same action (e.g. https://elifesciences.org/articles/12192 )?

-Half of participants base their confidence ratings mostly on confidence history, rather than on evidence. Does this just mean people were not very motivated to do well on the confidence report? Were they rewarded to perform well on both tasks?

-The authors test the population-level prevalence and effect sizes of perceptual and metacognitive history biases? However, with n = 37 their size does not seem large or diverse enough to infer properties of the general population.

Reviewer #2

In this paper, Benwell and colleagues present a perceptual decision-making task with confidence ratings in human subjects to investigate choice and metacognitive history biases. The statistical analyses are solid and thorough. The mutual information analysis is a valuable complement. The paper is comprehensive, clearly written and the figures are informative. It would help if the authors could emphasise their reasoning across the paper: why selecting this particular paradigm to study history biases, and what is the added value of their findings beyond previous work? Many similar data sets (perceptual decision & confidence) have been published and are open. Moreover, it seems important to better justify their hypotheses and motivate the analyses, particularly the final focus on repeat vs. alternation in a paradigm in which these were not manipulated explicitly by the experimenters.

The authors suggest that history biases are adaptive (building on the stability of real-life environments), whereas at other times, that they are maladaptive ("irrelevant factors such as previous confidence reports"). It would be helpful to explicit the arguments in favour of either interpretation, and to clarify what computational interpretation the present findings favour, if any. There is already a little bit about why it may be advantageous to assume stability. Are there other reasons to think of choice/metacognitive bias as helpful vs. maladaptive? In which contexts? If repeating were a more prevalent bias than alternating in the population, why would this be useful? Relatedly, it would be helpful to further clarify for readers why it is relevant to study choice and confidence history biases, i.e. explain why it is not a simple by-product of experimental designs where experimenters artificially present multiple times very similar decisions.

It is interesting that the authors comment on the prevalence of each result in their sample, instead of simply reporting statistics on group means, to get a better sense of the strength of the findings. However, it is a bit difficult to generalise about "population prevalence" unless larger samples than the current n=37 are used. Because the experimental design overlaps with previous work, most of these analyses could be re-done on other datasets to address discrepancies between the present findings and that of Urai et al. The confidence database (Rahnev et al., 2020) may provide a useful resource for future work (especially for drawing conclusions about population prevalence based on the current sample of 37 subjects).

Technical suggestions/comments:

Could the authors indicate the proportion of errors vs. correct trials on previous repeating vs. alternative trials? If there were more errors on alternating trials, could it be that due to a post error slowing mechanism whereby subjects became more accurate after an alternate, hence the increase in d'? Regarding Fig S1 and the comparison with previous studies, it would be worth discussing the results in relation to a related study in rats (Hermoso, Hyafil et al., 2020 ncomms), for instance examining if the choice bias was overall driven by correct trials and absent after errors (Fig 2A). Finally, in Fig S1 the authors show a choice bias present for both high and low confidence at t-1. Would it be more precise, for concluding about a lack of an influence of confidence, to perform an ordinal regression analysis using the 4 levels of confidence available?

In Fig 1F, 1G, 1H, could the authors perform psychometric and statistical analyses to actually demonstrate the findings that the authors describe, or rephrase that the confirmation of model predictions are qualitative only? (For instance, showing quantitatively that the slopes are different for high and low confidence in Fig. 1F)

In particular, when comparing Fig 1F and 1G, are the patterns for confidence and RTs identical with respect to absolute orientation? If so, is this an issue for interpreting confidence data (supposedly not only a strict reflection of RTs but also incorporating information about accuracy)?

Could the authors comment on (even briefly) the other psychometric parameters (stimulus independent lapse rate and slope)? Why do the stimulus-independent lapses fixed and not fitted with the two other psychometric parameters? Does it change the conclusions if they are fitted? It would be worth checking parameters of the sigmoid psychometric function in Fig S1 (right panel red curve), because the psychometric function looks unusual with an increase at highly positive orientations.

It is reassuring to see that the correlation results in Fig. S4 are reproduced using an alternative metric of metacognitive efficiency (meta-d'/d'). However, could the authors provide this measure for all other analyses based on meta-d'-d'? I am not asking for a detailed breakdown or new figures, but at least in the text specify whether findings are maintained using this alternative metric.

Could the authors argue that here we have a true metacognitive history bias, and not a bias due to low-level effects e.g. motor anchoring, use of scale? (See e.g. Foda, H., Barger, K., Navajas, J., & Bahrami, B. (2017). Domain-general idiosyncratic anchoring of metacognition.)

Reviewer #1

This preprint by Benwell and colleagues aims at studying 'history' biases during perceptual and metacognitive decision-making. The authors rely on a canonical, well-controlled two-alternative orientation discrimination task with confidence reports on a 4-point scale, tested in 37 human observers. Based on both model-free (mutual information) and model-based (type-2 signal detection theory) analyses, the authors report dissociations between perceptual history biases (related to previous perceptual decisions) and metacognitive history biases (related to previous confidence reports). They interpret these two forms of history biases as 'heuristics' informing type-1 (perceptual) and type-2 (confidence) decisions in the absence of sufficient evidence in the current trial.

History biases have generated a lot of interest over recent years, and the comparison of perceptual and metacognitive history biases in a single behavioral modeling study is interesting. The standard 2-AFC + 4-point confidence paradigm used by the authors is well suited to address the research question outlined by the authors in the introduction. However, I am not entirely satisfied by the definition of metacognitive history biases used by the authors throughout the manuscript – something which is likely to impact some of the dissociations reported by the authors.

Indeed, the authors report no effect of confidence at trial t-1 on perceptual bias at trial t (lines 147-149). However, this apparent dissociation appears unsurprising if the response at trial t-1 is not considered in the analysis (as shown in Figure 4). Indeed, a high confidence report in favor of a Left response at trial t-1 is likely to bias the perceptual response at trial t in the opposite direction to a high confidence report in favor of a Right response at trial t-1. In other words, when considering metacognitive history biases, the authors should consider four (not two) types of trials, based on the previous confidence report (high/low) but also the previous response (Left/Right). A true dissociation between previous confidence and the current response would imply that the perceptual history bias (i.e., the effect of the previous response on the current response) is of the same magnitude following low confidence reports and high confidence reports. It is very important that the authors re-compute the effect of metacognitive history biases in this response-dependent fashion. Indeed, Figure 4 will show no effect (i.e., an apparent dissociation) even if the perceptual history bias is strongly increased following high confidence reports than low confidence reports. The manuscript should always consider the direction of the previous response when assessing metacognitive history biases, and only claim dissociations if these previous response-dependent analyses reveal no interaction between the size of perceptual history biases and the previous confidence report (high vs. low).

Along the same lines, Figure 2C should be modified to plot confidence not as a function of the absolute orientation, but as a function of the orientation signed by the provided response (positive for a stimulus orientation consistent with the provided response, and negative for a stimulus orientation inconsistent with the provided response). It is highly likely that confidence does not scale with the absolute orientation as suggested by Figure 2C, but as a function of the consistency of the stimulus orientation with the provided response (highest confidence for an easy stimulus orientation consistent with the provided response, lowest confidence for an easy stimulus orientation inconsistent with the provided response).

Regarding the use of mutual information (MI) as a metric for comparing all kinds of effects, I wonder whether the authors could explain more extensively their reasoning. I was under the impression that MI is bounded differently for different kinds of variables (e.g., a binary effect is bounded at 1 bit). Comparing binary variables (Left/Right responses) with non-binary variables (1-4 confidence reports) may thus be problematic, unless I have missed something. To validate the use of MI for claiming "sub-optimal metacognitive performance" (lines 193-195) would require to perform sanity checks such as model simulations with optimal metacognitive performance. The goal of these simulations is to show the difference between the pattern of MI obtained for simulations with genuinely optimal metacognitive performance (and matched perceptual performance to the human subjects) and the pattern of MI obtained for human subjects.

In the type-2 SDT modeling, the authors now separate Left and Right responses when computing their abs(meta-c minus c) measure. As mentioned above, I believe that the authors should always split Left and Right responses when assessing metacognitive history biases and claiming dissociations between type-1 (perceptual) and type-2 (metacognitive) processes. As above, also, model simulations with optimal metacognitive performance but no history biases (neither perceptual nor metacognitive) may be useful to validate the analyses carried out by the authors on their behavioral data.

In Figure 4, as in Figure 2C, post high confidence and post low confidence should be further split between post high/low confidence in a Left response and post high/low confidence in a Right response. While a high confidence report in the previous trial may not potentiate perceptual sensitivity (d') on the current trial, it is possible that a high confidence report for a Left response in the previous trial would bias perception in favor of a Left response in the current trial more than a low confidence report for a Left response (and vice versa for a Right response).

Finally, the repeat vs. alternate effect shown in Figure 5 could be driven by at least two very different mechanisms that cannot be teased apart from presented analyses: 1) the reduced orientation sensitivity in repeat trials proposed by the authors, and 2) a non-zero 'repetition lapse' rate (i.e., a fraction of trials in which the subject blindly repeats his/her previous response). To distinguish between these two accounts, it would be very useful to re-plot Figure 5C. Instead of the absolute orientation, one would use as x-axis the orientation signed by the previous response (positive if the current orientation is in the same direction as the previous response, negative otherwise) and plot on the y-axis the fraction of repeat decisions in the current trial. Fitting a sigmoid function to this curve with a non-zero probability of a blind repetition should afford to tell whether the lower apparent sensitivity in repeat trials shown on Figure 5C is actually triggered by blind repetitions – instead of a lower sensitivity to orientation.

Preprint Review

This preprint was reviewed using eLife’s Preprint Review service, which provides public peer reviews of manuscripts posted on bioRxiv for the benefit of the authors, readers, potential readers, and others interested in our assessment of the work. This review applies only to version 2 of the manuscript.

Summary

This preprint describes a behavioral study of 'history biases' during perceptual and metacognitive decisions. The preprint reports dissociations between history biases across successive perceptual decisions and history biases across successive confidence reports. The question of how perceptual decisions and confidence reports arise from the same sensory input is important, and the statistical analyses of the behavioral data are rigorous and thorough. In particular, the proposed mutual information analysis appears valuable to quantify and compare different biasing effects. However, despite these methodological strengths, the preprint does not reveal substantive new insights into the computational processes which may give rise to the observed effects. Also, the preprint reads at times more like a collection of findings than a coherent study of a targeted research question. In this respect, it would be useful to emphasize the added value of the findings beyond the extensive previous work on history biases and on the relation between type-1 (perceptual) and type-2 (metacognitive) decisions. An explicit motivation of the performed analyses, particularly the focus on the repetition vs. alternation contrast at the end of the results, would be very useful to clarify the reasoning followed by the authors throughout the study.

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 all reviewers for their comments and suggestions, which will make our manuscript a much better one. Accordingly, we have already made changes to the manuscript (marked in yellow) and we will perform all the experiments requested. Below, we answer the reviewers point by point.

Reviewer #1 (Evidence, reproducibility and clarity (Required)): This study provides solid evidences showing a role for the spectraplakin Short-stop (Shot) in subcellular lumen formation in the Drosophila embryonic and larval trachea. This subcellular morphogenetic process relies on an inward membrane growth that depends on the proper organization of actin and microtubules (MTs) in terminal cells (TCs). Shot depletion leads to a defective or absent lumen while conversely, Shot overexpression promotes excessive branching, independently on the regulation of centrosome numbers previously shown to be important for the regulation of the lumen formation process (Ricolo, D., Deligiannaki, M., Casanova, J. & Araújo, S. J. Centrosome Amplification Increases Single-Cell Branching in Post-mitotic Cells. Current Biology 26, 2805-2813 (2016)). Shot is rather important to regulate the organization of the cytoskeleton by crosslinking MTs and actin. Shot expression in TCs is controlled by the Drosophila Serum Response Factor (DSRF) transcription factor. Finally Shot functionally overlaps with the MT-stabilizing protein Tau to promote lumen morphogenesis. The figures are clear and the questions well addressed with carefully designed and controlled experiments. However, I would have few suggestions that will hopefully make some points clearer. **Major comments:** -Statistical analyses should be added for comparisons of proportions, including Fig. 1E, 1L, Fig. 2G-I, Fig. 6L, Fig. 7K, Fig. 8C-D and Fig. 9G.

We agree with this and have now redone all graphs and revised all quantifications from this study. We have added error bars in all above mentioned graphs and have provided statistical analysis where appropriate. We have also redone all graphics and phenotype reporting, which is done now in relation to total TCs (rather than embryos or GBs and DBs TCs). This was suggested also by reviewer #2 and we agree because this is a more stringent and comparable way of quantifying our results.

-It is not always clear what genotype has been used as the "wt" genotype, as in Fig. S2 or Fig. 3 for example, this should be added to figure legends.

We have now clarified which flies are used as controls in each experiment throughout the paper. We have left wt where flies were wt, and changed all other cases to either the genotype or “control”.

-Live imaging of Shot has been performed with ShotC-GFP, that cannot bind actin. Don't the authors think ShotA-GFP would reflect more accurately Shot endogenous behavior as it interacts both with actin and MTs? It would be better to show this, even if the results shown here tend to be consistent with Shot endogenous localization shown with Shot antibody staining.

We agree and we will analyse movies with both ShotC and ShotA and present them in the revised version.

-It is of course not possible to generate CRISPR mutant flies with mutations in putative DSRF binding sites in a reasonable amount of time, to confirm that Shot transcription is controlled by DSRF. It would thus be nice to reveal shot mRNA expression with in situ hybridization experiments in wt vs. bs embryos. This would confirm that Shot mRNA is downregulated upon DSRF inhibition and rule out a possible indirect effect on Shot protein stability for example.

We believe the presented 3-way approach (in silico, protein quantification and phenotype rescue) is sufficient to show that Shot expression is regulated by DSRF. It is unlikely that we are dealing with protein stability or other issues, because we can rescue the lumen elongation phenotype by solely expressing Shot in TCs. However, we agree it would be nice to show this in an in situ hybridization experiment, and we will try to provide a conclusive one for resubmission. In situ detection methods, however, may not be accurate enough to detect such differences in single-cells.

-In the same figure, it would also be interesting to show what happens to actin and MTs in bs TCs and to which extent their organization is rescued by Shot overexpression.

We are working on this for resubmission. These experiments were frozen by the current COVID-19 pandemic and this is why they were not submitted with the first version.

-UAS-EB1GFP does not seem to be an appropriate control in Figure 9 (A and B) since it can affect MT dynamics (Vitre, B. et al. EB1 regulates microtubule dynamics and tubulin sheet closure in vitro. Nat. Cell Biol. 10, 415-421 (2008)). Why not simply use an UAS-GFP?

We have not detected any notorious larval TC phenotypes by overexpressing UASEB1GFP in TCs. Their branching is comparable to that in previous studies (for example Schotenfeld-Roames, et al Current Biology 2014) and there were no detectable luminal branching phenotypes. However, we agree it is more correct to analyse cells with a plain GFP and have repeated the controls for this experiment using DSRFGAL4UASGFP. This is now shown in figure 9.

-Shot and probably Tau crosslinking activities are important for lumen morphogenesis with a striking increase in the number of embryos without lumen in shot3 and shot3 tauMR22 mutant embryos. The rescue experiments clearly show that Shot binding to both MT and actin is essential for efficient rescue. The same might apply to Tau since it is able to crosslink actin and MTs (Elie, A. et al. Tau co-organizes dynamic microtubule and actin networks. Sci Rep 5, 1-10 (2015)). I believe showing actin and MTs organization in these rescue experiments would be necessary.

We agree and we will provide these experiments upon resubmission.

Second, the overexpression experiments indicate that Shot is able to induce extra lumen formation even when unable to bind actin as shown with the increase in the number of supernumerary lumina (ESLs) under overexpression of ShotC and ShotCtail to a lesser extent. This phenotype is also observed under Tau overexpression. This suggest that not crosslinking anymore but rather making MTs more stable could be sufficient to promote extra lumen formation in a wt context. Stabilising MTs by treatment with Taxol might thus be sufficient to promote ESL formation. I am fully aware of the difficulty of treating Drosophila embryos with drugs, making this experiment hard to do, but I think this dual function of Shot and Tau (crosslinking actin and MTs to promote branching vs. stabilizing MTs leading to excessive branching) should be discussed.

In Figure 2 we show not just that UASShotC is able to induce ESl but also that UAS-ShotCtail containing only the MT binding domain of Shot is enough to induce ESLs in TCs, whereas UAS-deltaCtail is not. We agree Taxol treatment would be a nice experiment to do, however we also think we provide enough evidence that MT stability is enough for ESL whereas de novo lumen formation requires crosslinking of MTs to actin. As advised, we will discuss better both Shot and Tau dual function in ESL generation and de novo lumen formation for resubmission.

**Minor comments:**

We have already addressed most these minor comments in the manuscript (text revised and changes in yellow). And we provide answers to some of the comments below.

-p2 line 1: 'acentrosomal luminal branching points' may be better than 'acentrosomal branching points' to describe the phenotype. -p4, line 16: the reference 23 is not properly inserted (should be after 'closure'). -p5, line 16: Please mention what the abbreviations Bnl and Btn stand for. -p5, line 20: these 80% of TCs cells with defects in subcellular lumen formation should appear on the graph in Fig. 1E (as shown in graph 1L).

We have added shot RNAi results to graph E in figure 1.

-p5, line 26: this 36% value does not seem to correspond to anything on the graph in Fig. 1N. According to the figure legend, 20% of TCs did not elongate at all and the lumen was completely absent (class IV), which is consistent with the result shown in Fig. 1L. Also, I am not sure why only 25 TCs were analysed in Fig. 1N while there are the data to analyse more as shown in Fig. 1E (400 TCs), this would make the graph more representative.

Figure 1 N represents a detail of the different phenotypes present in shot mutant embryos. Whereas for most of the paper we consider only complete lack of TC lumen, here we show the different types of affected TCs and not just the ones with a complete lack of subcellular lumen. We apologise because it was not explained in the original manuscript that types III and IV are the “no lumen” class (they were subdivided into 2 classes because they have different cell enlongation phenotypes). 36% of the total of affected TCs displayed the lack of lumen phenotype (this means a 22,5 % of the total number of TCs, because total affected TCs are 62,5% only). Numbers are similar but not exactly the same because this analysis was done using confocal microscopy and cells analysed one by one in detail, which is not possible using colorimetric methods and only luminal markers. This is also the reason we only analysed 25 TCs in this case. We thank the reviewer for pointing this out and have better described it in the manuscript.

-p6, line 8: ShotA-GFP is indeed a long isoform but is not the full-length Shot, as it does not contain the plakin repeat exon which would add another ~3000aa.

We have corrected this.

-p6, lines 21-23: ShotA-GFP localisation is not shown in FigS1. The authors should refer to Fig. 2. Enlarged areas/arrows might help the reader to better visualise the different localisations of ShotA-GFP and ShotC-GFP.

We thank the reviewer for this request and we will change the figure providing enlarged areas upon resubmission. In this version of the manuscript we have already changed the error in figure referral in the text.

-p7, line 23: Rca1 mutants should be better introduced here.

We have added one sentence of introduction to the Rca1 phenotype.

-p8, line 6: Shot colocalizes/associates with stable MTs and actin would be a more appropriate title for this paragraph.

We thank the reviewer for this alternative, and we have changed this title in the manuscript.

-p16, line 18: 'Shot is able to mediate crosstalk' would be better than 'Shot is able to crosstalk'. -p40, lines 6 and 7: L, M and N should be K', K' and K' respectively. -p41, Fig 10D: It is quite hard to see on the cartoon what the phenotype is for Shot OE.

We will make this clearer for resubmission.

-The following reference shows an important role for Shot in crosslinking actin and MTs during morphogenesis of the Drosophila embryo and should be cited in this manuscript (Booth, A. J. R., Blanchard, G. B., Adams, R. J. & Röper, K. A Dynamic Microtubule Cytoskeleton Directs Medial Actomyosin Function during Tube Formation. Developmental Cell 29, 562-576 (2014)).

We thank the reviewer for pointing this out, because this is of course an important reference known to us, which we forgot to add. We have now added this to the manuscript.

-FigS3. It would be good to add the labels on the figure (ShotC-GFP in green, and MoeRFP/lifeActinRFP in Magenta).

We will do this for resubmission.

Reviewer #1 (Significance (Required)): The findings shown in this manuscript shed an important light on the way subcellular morphogenesis occurs. It was known that both actin and MTs were required in this process, particularly during the formation of Drosophila trachea (JayaNandanan, N., Mathew, R. & Leptin, M. Guidance of subcellular tubulogenesis by actin under the control of a synaptotagmin-like protein and Moesin. Nature Communications 1-10 (2019). doi:10.1038/ncomms4036; Gervais, L. & Casanova, J. In Vivo Coupling of Cell Elongation and Lumen Formation in a Single Cell. Current Biology 20, 359-366 (2010)). This work provides additional molecular insights into the way branching morphogenesis from a single cell occurs in vivo, clearly demonstrating a requirement for actin-MT crosslinking mediated by Shot and Tau. This could be of great interest in the field of branching morphogenesis and lumen formation, not only in invertebrates but also in vertebrates where such a crosslinking might occur in the vasculature, the lung, the kidney or the mammary gland for example (Ochoa-Espinosa, A. & Affolter, M. Branching Morphogenesis: From Cells to Organs and Back. Cold Spring Harb Perspect Biol 4, a008243-a008243 (2012)). *Field of expertise:* morphogenesis, Drosophila, cytoskeleton, microtubules. Reviewer #2 (Evidence, reproducibility and clarity (Required)): **Summary:** The development of branched structures with intracellular lumen is widely observed in single cells of circulatory systems. However the molecular and cellular mechanisms of this complex morphogenesis are largely unknown. In previous study, the authors revealed that centrosome as a microtubule organizing center (MTOC) located at the apical junction contributes subcellular lumen formation in the terminal cells of Drosophila tracheal system. The microtubule bundles organized by MTOC are suggested to serve as trafficking mediators and structural stabilizers for the newly elongated lumen. In this manuscript, they focused on a Drosophila spectraplakin, Shot, which have been reported to crosslink MT minus-ends to actin network, in the subcellular lumen formation. The paper started by description of lumen elongation defect of the tracheal terminal cells in the shot[3] null mutant. The overexpression of full-length and series of truncated form of shot exhibited extra-subcellular lumina (ESL) in TCs, suggesting that Shot is required for the lumen formation in dose dependent manner. They next addressed whether Shot overexpression induces ESL through the supernumerary centrosomes as in Rca1 mutant, however the number of centrosomes was not affected. Moreover, the ESL were sprouted distally from the apical junction, suggesting that Shot operate in different way from the Rca1-dependent microtubule organization. To get mechanistic insight of Shot in the luminal formation, they checked localization of the Shot and found it localized with stable MTs around the nascent lumen and with the F-actin at the tip of the cell during the cell elongation and subcellular lumen formation. In shot[3] mutant, the MT-bundles were no longer localized to apical region and the actin accumulation at the tip of the cell was also reduced. The rescue experiments using several truncated forms of Shot, and well-designed genetic analysis using various shot mutants revealed that both MT binding domain and actin binding domains are needed to develop the lumen. The expression of shot was under the regulation by terminal cell-specific transcription factor bs/DSRF, and the overexpression of shot in bs LOF mutant suppressed its phenotype, indicated that part of the luminal phenotype of bs mutant in terminal cells are due to lower levels of the activity of shot. Finally, they checked whether Tau can compensate the function of shot in the subcellular lumen formation. The lumen elongation defect in shot mutant was suppressed by tau expression, and tau overexpression phenocopied the shot overexpression-induced ESL. Although tau mutant did not show the lumen formation defects, the double mutant of shot and tau exhibited synergistic effect. Shot was also required for subcellular luminal branching at larval stages. Overall, this work highlighted the importance of Shot as a crosslinker between MT and actin that acts in downstream of the FGF signaling-induced bs/DSRF expression for the subcellular lumen formation. An excess of Shot is sufficient for ESL formation from ectopic acentrosomal branching points. Furthermore, the Tau protein can functionally replace Shot in this context. **Major comments:** *- Are the key conclusions convincing?* *- Should the authors qualify some of their claims as preliminary or speculative, or remove them altogether?* The conclusions were basically supported by the set of data presented in this article, but following points need to be clarified. The truncated form ShotC lacks only half of calponin domain that are essential for the actin binding, thus it is still possible to bind actin to some extent. Although the actin binding activity is reported as "very weak" in the cited references, the quantitative analysis has not been done. Thus, the interpretation and claims based on the experiments using ShotC should be reviewed carefully.

We agree with the reviewer and will revise all the text for resubmission in order to make this unambiguous. However, we would like to remark that our claims are not only based on UAS-ShotC but also in the shotkakP2 allele, which does not contain one of the calponin domains and in isoforms such UAS-Shot C-tail which do not have any ABD.

Data set in some places seems fragmented. For example, overexpression study of shot constructs (Fig. 2) lacks phenotypic comparison of control (btl Gal4 driven control FP) to compare if phenotypes of shot constructs expression are different from control. Different methods of phenotypic quantification are employed. One was counting embryo number with at least one abnormality among 20 TCs of DB or GB, or the other counting every TC for the presence of lumen/branching conditions. The latter is more stringent measure and is more appropriate for the study of single cell morphogenesis.

We totally agree with the reviewer. We have now revised all quantifications and graphs:

1) We have used btl>GFP as control to all overexpression experiments in embryos and DSRFGAL4UASGFP in control larvae.

2) We have made the paper uniform regarding quantifications, which are now all done in relation to total TCs and not embryos.

For this reason, many of the graphs, figure legends and quantification values in the the manuscript text are now changed.

*- 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.* The all movies were using ShotC isoform which lacks half of the actin binding domain. The truncated isoform is not suitable to observe the localization, especially the colocalization with actin. The movies need to be retaken using full-length Shot at the dosage that does not interfere with normal TC development.

We agree and we will analyse movies with both ShotC and ShotA for resubmission.

Some statements on Moesin and Tau localization sound as if the authors studied Shot interaction with nascent Moe and Tau molecules. This is confusing because fragments of Moe and Tau, but not functional full length proteins, were used.

We will revise the text to make this unambiguous fir resubmission.

*- 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. Because the transgenic fly is already present, we assume it would be done in 4 weeks. However, it would be influnced under social circumstances whether the lab facilities are able to access or not. *- Are the data and the methods presented in such a way that they can be reproduced?* *- Are the experiments adequately replicated and statistical analysis adequate?* The methods provided seem to be sufficient for reproducing the data by competent researchers, and most of the data are solid and the sample numbers are sufficient for the claims. However, the criteria for phenotypic evaluation differs among graphs and figures, that possibly confuse the readers. Standardized measurement methods are desirable. **Minor comments:** *- Specific experimental issues that are easily addressable.* In the rescue experiments shown in Figure 6, only full-length Shot rescued the subcellular lumen formation, but either of truncated Shot did not. The localization study of MT and actin in those conditions will reveal whether proper localizations of actin and MT are critical for the lumen formation.

We are working on this for resubmission. These experiments were stalled by the current COVID-19 pandemic and this is why they were not submitted with the first version. We will provide MT and actin localization for the rescue experiments with ShotA and ShotC.

*- Are prior studies referenced appropriately?* The references are cited appropriately. *- Are the text and figures clear and accurate?* There are several typos: Remodelling -> remodeling, signalling -> signaling. In the figure 2, G and H seem redundant. Scale bars are missing in Fig1 F-K, Fig2 K-L, Fig6 A-I, Fig7 E-J and Fig8 E-J.

We have changed the graphs in figure 2. Typos have been corrected. We will provide errors bars for resubmission.

The author often called shot+ genotype as "wild type". They are transgenic strains with some mutations, and cannot be found in the wild. They should be simply called with genotype or "control" for experiments.

We thank the reviewer for pointing these typos and incoherences with control genotypes. We have partly revise the text and figures and will finish for resubmission.

*- Do you have suggestions that would help the authors improve the presentation of their data and conclusions?* In Figure 4, as the localization of Shot is difficult to see in detail, enlarged insets might help. In addition, the green and cyan in C'-E' is difficult to distinguish.

We will change this for resubmission.

With Figure 5, the authors claimed that Shot LOF leads to disorganized MT-bundles and actin localization. We feel this is an overstatement and the Figure should be backed up with better data, or removed. F-actin and microtubule localizations are highly dynamic and the snapshot pictures are insufficient for demonstrating defective localization. It is also possible that (potential) difference in the marker localization is due to indirect effect of Shot LOF in cell shape.

We agree with the reviewer that fixed samples are not the best to analyse cytoskeletal components, but we observe clear differences in MT bundles and specially in actin localization in shot mutants as compared to controls and we believe it is important to show these results. Cell shape might of course alter the analysis which is why we present 3 different cell shapes in Figure 5. In addition, there are many previous studies where localization of MTs and actin was done in fixed mutant embryos, where cell shape is also affected, and revealed important steps in TC formation (Gervais and Casanova, 2010; JayanNadanan et al. 2014).Nonetheless, we have revised the text in order to avoid overstatements.

Reviewer #2 (Significance (Required)): *- Describe the nature and significance of the advance (e.g. conceptual, technical, clinical) for the field.* *- Place the work in the context of the existing literature (provide references, where appropriate).* In blood capillary and insect trachea, the branching process of single vessel cells involves sprouting of cell protrusions, followed by the lumen extension from the main vessels. The lumen formation involves assembly of plasma membrane components inside of the cytoplasm. Since the luminal membrane is associated with protein complexes common to apical cell membrane, lumen formation is believed to involve redirection of apical trafficking of membranes to intracellular sites (Sigurbjörnsdóttir, Mathew, Leptin 2014, 10.1038/nrm3871). The authors previously demonstrated that centrosome is an important link of preexisting lumen to de novo lumen formation, leading to the hypothesis that centrosome-derived microtubules organize lumen membrane assembly. *- State what audience might be interested in and influenced by the reported findings.* In this manuscript, the authors addressed this issue by looking at the function of Shot/Plakin that has both microtubule and actin binding activities. Shot is an ideal candidate for linking actin-rich cell protrusions in the leading edge to centrosome- associated lumen tip. Indeed the authors clearly showed that shot is required for lumen extension and overexpressed shot protein associates with intracellular tract rich in microtubules and F-actin. Their findings are definitely a progress in the field of Drosophila tracheal development. Having said that, how Shot links leading edge protrusions and centrosomes, how it is organized into pre-lumen tract, and how it contribute to further assembly of luminal membrane and directed secretion, are not well understood yet. Without clues to those fundamental questions, I believe this paper is most appropriate for experts readers of Drosophila cell biology and tracheal development. Finally I feel that the paper include many data sets and some pictures are not easy to grasp essential points, such as three movies showing localization of overexpressed shot-C, RFP-moesin, and Lifeact. *- Define your field of expertise with a few keywords to help the authors contextualize your point of view.* Drosophila, tracheal cell biology. *- Indicate if there are any parts of the paper that you do not have sufficient expertise to evaluate.* No Reviewer #3 (Evidence, reproducibility and clarity (Required)): **Summary** In their manuscript entitled "Coordinated crosstalk between microtubules and actin by a spectraplakin regulates lumen formation and branching" Ricolo and Araujo characterize the requirement for Short Stop (Shot) in the formation of subcellular tubes in tracheal terminal cells. The authors examined embryos homozygous for shot3, a presumed null allele of shot. They found an 80% penetrant defect in seamless tube formation or growth. The phenotype resembles that reported for mutations in blistered, which encodes the Drosophila SRF ortholog. The authors find that expression of SRF is not blocked by mutations in shot and later find that bs mutants have decreased levels of shot expression and that shot overexpression can partly suppress the bs tube formation defects. The authors then examine whether the requirement for shot is autonomous to the trachea and find that it is, as pan-tracheal shot RNAi replicates the seamless tube defects. The authors find that overexpression of various Shot isoforms results in the formation of ectopic seamless tubes within terminal cells. Using the various transgenic constructs available for shot, the authors show that the overexpression phenotype is dependent upon the interaction between Shot and microtubules, and is dose-dependent. Previous work had shown that ectopic terminal cell tubes also can arise due to increased centrosome number; the authors show that centrosome number is not altered in shot mutants. Shot has well characterized actin and microtubule binding functions, and the authors show that Shot localization overlaps both with microtubules and with actin, and that both cytoskeletal elements are aberrant in shot mutant cells. In a series of experiments utilizing various shot mutant backgrounds and shot transgenes, the authors identify requirements for both Shot-cytoskeleton interactions in the formation and branching of seamless tubes in terminal cells. Finally, the authors examine the requirement for Tau in the same processes. Tau and Shot had previously been found to work together in neurons, and this seems to be true in terminal cells as well. Tau overexpression induces ectopic seamless tubes and can partially suppress shot loss of function. Embryos mutant for tau showed seamless tube directionality defects, but not lumen formation or branching. Embryos doubly mutant for tau and shot showed a more severe seamless tube defect than shot mutants alone - an increase in terminal cells with no lumen from 22% to 85%. Authors also examined terminal cells in larval stages using dsrf-Gal4 to knockdown shot in terminal cells (rather than pan-tracheal knockdown with breathless). The authors conclude from their studies that Shot, through its interactions with microtubules and the actin cytoskeleton coordinate the outgrowth and branching of subcellular tubes. Overlapping function of Tau and possibly other additional MAPs also act in these processes. The work is largely well done and the conclusions are supported by the data. **Minor concerns:** -If one were to start this work today, crispr knockout and knockins would be preferred. While shot^3 is widely considered a null allele, there are indications that some shot function is still present in shot^3 embryos. This would also be relevant to the penetrance of the defects. The transgenes are useful, but given the dosage effects noted in various of the authors experiments, interpretation of some experiments is complicated as compared to a knockin. For overexpression experiments, landing site constructs would be preferable. I do not mean to suggest that the authors necessarily go this route, but am just pointing out a limitation of the approach.

We agree, but we also think that with the amount of data and tools generated by other labs over recent years, regarding shot function in the nervous system (Voelzmann et al 2017), we are in a position to be able to take the conclusions of this work based on these transgenic and different shot alleles.

-Insight into function at higher resolution than altered microtubule and actin organization would significantly increase the impact. -cell autonomy (line 19, p5) is not the correct term. Pan-tracheal knockdown tests tissue autonomy. Mosaic analysis or terminal cell specific knockdown would address cell autonomy.

We have changed the manuscript accordingly.

-line 14 p6 acting should be actin -dsrf-Gal4 transgenes were made by Mark Metzstein

We have corrected these.

-there also appears to be rescue of the fusion cell defects of shot by Tau overexpression. Authors should comment on this and what it means for the seamless tubulogenesis program in terminal cells vs fusion cells.

We will reanalyse shot rescued with tau embryos focusing on fusion phenotypes and discuss this in the revised version.

Reviewer #3 (Significance (Required)): The findings will be of interest to a broad cell biology community as they provide a conceptual advance and may help to focus future work on seamless tubulogenesis. The authors do a good job of placing the results in the context of previous studies. *Field of expertise:* Drosophila, tracheal tubulogenesis, developmental 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 #3

Evidence, reproducibility and clarity

Summary

In their manuscript entitled "Coordinated crosstalk between microtubules and actin by a spectraplakin regulates lumen formation and branching" Ricolo and Araujo characterize the requirement for Short Stop (Shot) in the formation of subcellular tubes in tracheal terminal cells.

The authors examined embryos homozygous for shot3, a presumed null allele of shot. They found an 80% penetrant defect in seamless tube formation or growth. The phenotype resembles that reported for mutations in blistered, which encodes the Drosophila SRF ortholog. The authors find that expression of SRF is not blocked by mutations in shot and later find that bs mutants have decreased levels of shot expression and that shot overexpression can partly suppress the bs tube formation defects.

The authors then examine whether the requirement for shot is autonomous to the trachea and find that it is, as pan-tracheal shot RNAi replicates the seamless tube defects.

The authors find that overexpression of various Shot isoforms results in the formation of ectopic seamless tubes within terminal cells. Using the various transgenic constructs available for shot, the authors show that the overexpression phenotype is dependent upon the interaction between Shot and microtubules, and is dose-dependent.

Previous work had shown that ectopic terminal cell tubes also can arise due to increased centrosome number; the authors show that centrosome number is not altered in shot mutants.

Shot has well characterized actin and microtubule binding functions, and the authors show that Shot localization overlaps both with microtubules and with actin, and that both cytoskeletal elements are aberrant in shot mutant cells. In a series of experiments utilizing various shot mutant backgrounds and shot transgenes, the authors identify requirements for both Shot-cytoskeleton interactions in the formation and branching of seamless tubes in terminal cells.

Finally, the authors examine the requirement for Tau in the same processes. Tau and Shot had previously been found to work together in neurons, and this seems to be true in terminal cells as well. Tau overexpression induces ectopic seamless tubes and can partially suppress shot loss of function. Embryos mutant for tau showed seamless tube directionality defects, but not lumen formation or branching. Embryos doubly mutant for tau and shot showed a more severe seamless tube defect than shot mutants alone - an increase in terminal cells with no lumen from 22% to 85%.

Authors also examined terminal cells in larval stages using dsrf-Gal4 to knockdown shot in terminal cells (rather than pan-tracheal knockdown with breathless).

The authors conclude from their studies that Shot, through its interactions with microtubules and the actin cytoskeleton coordinate the outgrowth and branching of subcellular tubes. Overlapping function of Tau and possibly other additional MAPs also act in these processes.

The work is largely well done and the conclusions are supported by the data.

Minor concerns:

-If one were to start this work today, crispr knockout and knockins would be preferred. While shot^3 is widely considered a null allele, there are indications that some shot function is still present in shot^3 embryos. This would also be relevant to the penetrance of the defects. The transgenes are useful, but given the dosage effects noted in various of the authors experiments, interpretation of some experiments is complicated as compared to a knockin. For overexpression experiments, landing site constructs would be preferable. I do not mean to suggest that the authors necessarily go this route, but am just pointing out a limitation of the approach.

-Insight into function at higher resolution than altered microtubule and actin organization would significantly increase the impact.

-cell autonomy (line 19, p5) is not the correct term. Pan-tracheal knockdown tests tissue autonomy. Mosaic analysis or terminal cell specific knockdown would address cell autonomy.

-line 14 p6 acting should be actin

-dsrf-Gal4 transgenes were made by Mark Metzstein

-there also appears to be rescue of the fusion cell defects of shot by Tau overexpression. Authors should comment on this and what it means for the seamless tubulogenesis program in terminal cells vs fusion cells.

Significance

The findings will be of interest to a broad cell biology community as they provide a conceptual advance and may help to focus future work on seamless tubulogenesis. The authors do a good job of placing the results in the context of previous studies.

Field of expertise: Drosophila, tracheal tubulogenesis, developmental 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

Summary:

The development of branched structures with intracellular lumen is widely observed in single cells of circulatory systems. However the molecular and cellular mechanisms of this complex morphogenesis are largely unknown. In previous study, the authors revealed that centrosome as a microtubule organizing center (MTOC) located at the apical junction contributes subcellular lumen formation in the terminal cells of Drosophila tracheal system. The microtubule bundles organized by MTOC are suggested to serve as trafficking mediators and structural stabilizers for the newly elongated lumen.

In this manuscript, they focused on a Drosophila spectraplakin, Shot, which have been reported to crosslink MT minus-ends to actin network, in the subcellular lumen formation. The paper started by description of lumen elongation defect of the tracheal terminal cells in the shot[3] null mutant. The overexpression of full-length and series of truncated form of shot exhibited extra-subcellular lumina (ESL) in TCs, suggesting that Shot is required for the lumen formation in dose dependent manner. They next addressed whether Shot overexpression induces ESL through the supernumerary centrosomes as in Rca1 mutant, however the number of centrosomes was not affected. Moreover, the ESL were sprouted distally from the apical junction, suggesting that Shot operate in different way from the Rca1-dependent microtubule organization. To get mechanistic insight of Shot in the luminal formation, they checked localization of the Shot and found it localized with stable MTs around the nascent lumen and with the F-actin at the tip of the cell during the cell elongation and subcellular lumen formation. In shot[3] mutant, the MT-bundles were no longer localized to apical region and the actin accumulation at the tip of the cell was also reduced. The rescue experiments using several truncated forms of Shot, and well-designed genetic analysis using various shot mutants revealed that both MT binding domain and actin binding domains are needed to develop the lumen. The expression of shot was under the regulation by terminal cell-specific transcription factor bs/DSRF, and the overexpression of shot in bs LOF mutant suppressed its phenotype, indicated that part of the luminal phenotype of bs mutant in terminal cells are due to lower levels of the activity of shot. Finally, they checked whether Tau can compensate the function of shot in the subcellular lumen formation. The lumen elongation defect in shot mutant was suppressed by tau expression, and tau overexpression phenocopied the shot overexpression-induced ESL. Although tau mutant did not show the lumen formation defects, the double mutant of shot and tau exhibited synergistic effect. Shot was also required for subcellular luminal branching at larval stages.

Overall, this work highlighted the importance of Shot as a crosslinker between MT and actin that acts in downstream of the FGF signaling-induced bs/DSRF expression for the subcellular lumen formation. An excess of Shot is sufficient for ESL formation from ectopic acentrosomal branching points. Furthermore, the Tau protein can functionally replace Shot in this context.

Major comments:

- Are the key conclusions convincing? - Should the authors qualify some of their claims as preliminary or speculative, or remove them altogether?

The conclusions were basically supported by the set of data presented in this article, but following points need to be clarified.

The truncated form ShotC lacks only half of calponin domain that are essential for the actin binding, thus it is still possible to bind actin to some extent. Although the actin binding activity is reported as "very weak" in the cited references, the quantitative analysis has not been done. Thus, the interpretation and claims based on the experiments using ShotC should be reviewed carefully.

Data set in some places seems fragmented. For example, overexpression study of shot constructs (Fig. 2) lacks phenotypic comparison of control (btl Gal4 driven control FP) to compare if phenotypes of shot constructs expression are different from control. Different methods of phenotypic quantification are employed. One was counting embryo number with at least one abnormality among 20 TCs of DB or GB, or the other counting every TC for the presence of lumen/branching conditions. The latter is more stringent measure and is more appropriate for the study of single cell morphogenesis.

- 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.

The all movies were using ShotC isoform which lacks half of the actin binding domain. The truncated isoform is not suitable to observe the localization, especially the colocalization with actin. The movies need to be retaken using full-length Shot at the dosage that does not interfere with normal TC development.

Some statements on Moesin and Tau localization sound as if the authors studied Shot interaction with nascent Moe and Tau molecules. This is confusing because fragments of Moe and Tau, but not functional full length proteins, were used.

- 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.

Because the transgenic fly is already present, we assume it would be done in 4 weeks. However, it would be influnced under social circumstances whether the lab facilities are able to access or not.

- Are the data and the methods presented in such a way that they can be reproduced? - Are the experiments adequately replicated and statistical analysis adequate?

The methods provided seem to be sufficient for reproducing the data by competent researchers, and most of the data are solid and the sample numbers are sufficient for the claims. However, the criteria for phenotypic evaluation differs among graphs and figures, that possibly confuse the readers. Standardized measurement methods are desirable.

Minor comments:

- Specific experimental issues that are easily addressable.

In the rescue experiments shown in Figure 6, only full-length Shot rescued the subcellular lumen formation, but either of truncated Shot did not. The localization study of MT and actin in those conditions will reveal whether proper localizations of actin and MT are critical for the lumen formation.

- Are prior studies referenced appropriately?

The references are cited appropriately.

- Are the text and figures clear and accurate?

There are several typos: Remodelling -> remodeling, signalling -> signaling. In the figure 2, G and H seem redundant. Scale bars are missing in Fig1 F-K, Fig2 K-L, Fig6 A-I, Fig7 E-J and Fig8 E-J.

The author often called shot+ genotype as "wild type". They are transgenic strains with some mutations, and cannot be found in the wild. They should be simply called with genotype or "control" for experiments.

- Do you have suggestions that would help the authors improve the presentation of their data and conclusions?

In Figure 4, as the localization of Shot is difficult to see in detail, enlarged insets might help. In addition, the green and cyan in C'-E' is difficult to distinguish.

With Figure 5, the authors claimed that Shot LOF leads to disorganized MT-bundles and actin localization. We feel this is an overstatement and the Figure should be backed up with better data, or removed. F-actin and microtubule localizations are highly dynamic and the snapshot pictures are insufficient for demonstrating defective localization. It is also possible that (potential) difference in the marker localization is due to indirect effect of Shot LOF in cell shape.

Significance

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

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

In blood capillary and insect trachea, the branching process of single vessel cells involves sprouting of cell protrusions, followed by the lumen extension from the main vessels. The lumen formation involves assembly of plasma membrane components inside of the cytoplasm. Since the luminal membrane is associated with protein complexes common to apical cell membrane, lumen formation is believed to involve redirection of apical trafficking of membranes to intracellular sites (Sigurbjörnsdóttir, Mathew, Leptin 2014, 10.1038/nrm3871). The authors previously demonstrated that centrosome is an important link of preexisting lumen to de novo lumen formation, leading to the hypothesis that centrosome-derived microtubules organize lumen membrane assembly.

- State what audience might be interested in and influenced by the reported findings.

In this manuscript, the authors addressed this issue by looking at the function of Shot/Plakin that has both microtubule and actin binding activities. Shot is an ideal candidate for linking actin-rich cell protrusions in the leading edge to centrosome- associated lumen tip. Indeed the authors clearly showed that shot is required for lumen extension and overexpressed shot protein associates with intracellular tract rich in microtubules and F-actin. Their findings are definitely a progress in the field of Drosophila tracheal development. Having said that, how Shot links leading edge protrusions and centrosomes, how it is organized into pre-lumen tract, and how it contribute to further assembly of luminal membrane and directed secretion, are not well understood yet. Without clues to those fundamental questions, I believe this paper is most appropriate for experts readers of Drosophila cell biology and tracheal development.

Finally I feel that the paper include many data sets and some pictures are not easy to grasp essential points, such as three movies showing localization of overexpressed shot-C, RFP-moesin, and Lifeact.

- Define your field of expertise with a few keywords to help the authors contextualize your point of view.

Drosophila, tracheal cell biology.

- Indicate if there are any parts of the paper that you do not have sufficient expertise to evaluate.

No

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 provides solid evidences showing a role for the spectraplakin Short-stop (Shot) in subcellular lumen formation in the Drosophila embryonic and larval trachea. This subcellular morphogenetic process relies on an inward membrane growth that depends on the proper organization of actin and microtubules (MTs) in terminal cells (TCs). Shot depletion leads to a defective or absent lumen while conversely, Shot overexpression promotes excessive branching, independently on the regulation of centrosome numbers previously shown to be important for the regulation of the lumen formation process (Ricolo, D., Deligiannaki, M., Casanova, J. & Araújo, S. J. Centrosome Amplification Increases Single-Cell Branching in Post-mitotic Cells. Current Biology 26, 2805-2813 (2016)). Shot is rather important to regulate the organization of the cytoskeleton by crosslinking MTs and actin. Shot expression in TCs is controlled by the Drosophila Serum Response Factor (DSRF) transcription factor. Finally Shot functionally overlaps with the MT-stabilizing protein Tau to promote lumen morphogenesis.

The figures are clear and the questions well addressed with carefully designed and controlled experiments. However, I would have few suggestions that will hopefully make some points clearer.

Major comments:

-Statistical analyses should be added for comparisons of proportions, including Fig. 1E, 1L, Fig. 2G-I, Fig. 6L, Fig. 7K, Fig. 8C-D and Fig. 9G.

-It is not always clear what genotype has been used as the "wt" genotype, as in Fig. S2 or Fig. 3 for example, this should be added to figure legends.

-Live imaging of Shot has been performed with ShotC-GFP, that cannot bind actin. Don't the authors think ShotA-GFP would reflect more accurately Shot endogenous behavior as it interacts both with actin and MTs? It would be better to show this, even if the results shown here tend to be consistent with Shot endogenous localization shown with Shot antibody staining.

-It is of course not possible to generate CRISPR mutant flies with mutations in putative DSRF binding sites in a reasonable amount of time, to confirm that Shot transcription is controlled by DSRF. It would thus be nice to reveal shot mRNA expression with in situ hybridization experiments in wt vs. bs embryos. This would confirm that Shot mRNA is downregulated upon DSRF inhibition and rule out a possible indirect effect on Shot protein stability for example.

-In the same figure, it would also be interesting to show what happens to actin and MTs in bs TCs and to which extent their organization is rescued by Shot overexpression.

-UAS-EB1GFP does not seem to be an appropriate control in Figure 9 (A and B) since it can affect MT dynamics (Vitre, B. et al. EB1 regulates microtubule dynamics and tubulin sheet closure in vitro. Nat. Cell Biol. 10, 415-421 (2008)). Why not simply use an UAS-GFP?

-Shot and probably Tau crosslinking activities are important for lumen morphogenesis with a striking increase in the number of embryos without lumen in shot3 and shot3 tauMR22 mutant embryos. The rescue experiments clearly show that Shot binding to both MT and actin is essential for efficient rescue. The same might apply to Tau since it is able to crosslink actin and MTs (Elie, A. et al. Tau co-organizes dynamic microtubule and actin networks. Sci Rep 5, 1-10 (2015)). I believe showing actin and MTs organization in these rescue experiments would be necessary.

Second, the overexpression experiments indicate that Shot is able to induce extra lumen formation even when unable to bind actin as shown with the increase in the number of supernumerary lumina (ESLs) under overexpression of ShotC and ShotCtail to a lesser extent. This phenotype is also observed under Tau overexpression. This suggest that not crosslinking anymore but rather making MTs more stable could be sufficient to promote extra lumen formation in a wt context. Stabilising MTs by treatment with Taxol might thus be sufficient to promote ESL formation. I am fully aware of the difficulty of treating Drosophila embryos with drugs, making this experiment hard to do, but I think this dual function of Shot and Tau (crosslinking actin and MTs to promote branching vs. stabilizing MTs leading to excessive branching) should be discussed.

Minor comments:

-p2 line 1: 'acentrosomal luminal branching points' may be better than 'acentrosomal branching points' to describe the phenotype.

-p4, line 16: the reference 23 is not properly inserted (should be after 'closure').

-p5, line 16: Please mention what the abbreviations Bnl and Btn stand for.

-p5, line 20: these 80% of TCs cells with defects in subcellular lumen formation should appear on the graph in Fig. 1E (as shown in graph 1L).

-p5, line 26: this 36% value does not seem to correspond to anything on the graph in Fig. 1N. According to the figure legend, 20% of TCs did not elongate at all and the lumen was completely absent (class IV), which is consistent with the result shown in Fig. 1L.

Also, I am not sure why only 25 TCs were analysed in Fig. 1N while there are the data to analyse more as shown in Fig. 1E (400 TCs), this would make the graph more representative.

-p6, line 8: ShotA-GFP is indeed a long isoform but is not the full-length Shot, as it does not contain the plakin repeat exon which would add another ~3000aa.

-p6, lines 21-23: ShotA-GFP localisation is not shown in FigS1. The authors should refer to Fig. 2. Enlarged areas/arrows might help the reader to better visualise the different localisations of ShotA-GFP and ShotC-GFP.

-p7, line 23: Rca1 mutants should be better introduced here.

-p8, line 6: Shot colocalizes/associates with stable MTs and actin would be a more appropriate title for this paragraph.

-p16, line 18: 'Shot is able to mediate crosstalk' would be better than 'Shot is able to crosstalk'.

-p40, lines 6 and 7: L, M and N should be K', K' and K' respectively.

-p41, Fig 10D: It is quite hard to see on the cartoon what the phenotype is for Shot OE.

-The following reference shows an important role for Shot in crosslinking actin and MTs during morphogenesis of the Drosophila embryo and should be cited in this manuscript (Booth, A. J. R., Blanchard, G. B., Adams, R. J. & Röper, K. A Dynamic Microtubule Cytoskeleton Directs Medial Actomyosin Function during Tube Formation. Developmental Cell 29, 562-576 (2014)).

-FigS3. It would be good to add the labels on the figure (ShotC-GFP in green, and MoeRFP/lifeActinRFP in Magenta).

Significance

The findings shown in this manuscript shed an important light on the way subcellular morphogenesis occurs. It was known that both actin and MTs were required in this process, particularly during the formation of Drosophila trachea (JayaNandanan, N., Mathew, R. & Leptin, M. Guidance of subcellular tubulogenesis by actin under the control of a synaptotagmin-like protein and Moesin. Nature Communications 1-10 (2019). doi:10.1038/ncomms4036; Gervais, L. & Casanova, J. In Vivo Coupling of Cell Elongation and Lumen Formation in a Single Cell. Current Biology 20, 359-366 (2010)). This work provides additional molecular insights into the way branching morphogenesis from a single cell occurs in vivo, clearly demonstrating a requirement for actin-MT crosslinking mediated by Shot and Tau.

This could be of great interest in the field of branching morphogenesis and lumen formation, not only in invertebrates but also in vertebrates where such a crosslinking might occur in the vasculature, the lung, the kidney or the mammary gland for example (Ochoa-Espinosa, A. & Affolter, M. Branching Morphogenesis: From Cells to Organs and Back. Cold Spring Harb Perspect Biol 4, a008243-a008243 (2012)).

Field of expertise: morphogenesis, Drosophila, cytoskeleton, microtubules.

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 response to reviewers consists of three parts:

1. A summary of the main points from the two reviews, and the authors' response to these points.
2. A detailed revision plan for the preprint, taking into account both the main points of the reviews, and other comments made by the reviewers.
3. A point-by-point response to the reviewers.

For figure citations, OV = old version, i.e. bioRxiv preprint 2019-826180v2, and NV = new version, i.e. revised and re-submitted version.

1. Summary of main points by the reviewers, and authors’ responses:

• Both reviewers felt that the manuscript was overlong; Reviewer 1 recommended either shortening it or splitting it into two stories, while Reviewer 2 recommended cutting down the text.
  • We have considerably shortened the manuscript in accordance with this request (see revision plan below). We had already considered splitting the manuscript into two parts during the drafting stage, and had rejected this possibility as the data are intertwined - the retroactive validation of the dimer interface by the mutagenesis constructs (OV Fig. S3 [NV Fig. S4]) being a good example.
  • The revised manuscript features 7 main figures and 13 supplementals.
• Both reviewers felt too much text and figure space was allocated to negative data, specifically the investigation of potential lipid binding by the TbMORN1 protein, and that there should be more focus on the positive parts of the story.
  • A key part of shortening the manuscript has been moving most of the negative data on lipid binding into the supplemental figures, and considerably shortening the associated text. This has allowed the main figures and associated text to focus more on the positive elements of the project, while still ensuring publication of all the data.
• The reviewers appear to be in slight disagreement concerning discussion of the data. Reviewer 1 has encouraged more speculation on the physiological role of PE binding, a potential lipid transfer function, a role for calcium ions, the relevance of the observed disulphide bond, and the role of zinc ions in apicomplexan proteins; Reviewer 2 has recommended avoiding excessive speculation or inference.
  • Given that both reviewers have agreed that the original manuscript was overlong, we have implemented Reviewer 2's suggestion here and reduced the amount of speculation in the revised text.
• The reviewers agreed that the technical quality of the data was high and that the conclusions drawn were robust.
  • We are glad that the reviewers were appreciative of the data quality. For this reason, we were reluctant to remove any of the data from the manuscript and would prefer instead to transfer it to the supplementals. We feel that the negative data still have considerable community value, given that they show that MORN repeats are not automatically lipid binding modules and can thus act as a caveat to other researchers.

2. Detailed revision plan for the preprint:

• We have implemented the reviewers' suggestions and substantially shortened the manuscript, primarily by trimming the (phospho)lipid-binding section, which contains a large amount of negative data. The following main figures have been moved into the supplemental section:
  • OV Fig. 2 ("TbMORN1 interacts with phospholipids but not liposomes") has become NV Fig. S2
  • OV Fig. 4 ("TbMORN1(2-15) does not bind to liposomes in vitro") has become NV Fig. S6
  • OV Fig. 8 ("Conservation and properties of residues in TbMORN1(7- 15)") has become NV Fig. S11
• This has left a total of 7 main figures and 13 supplementals.
• The text associated with the entirety of the lipid-binding part (OV lines 210- 530, OV Figs. 2-6 [NV Figs. 2-4, S2, S6], OV Supplemental Figs. 2-6 [NV Supplemental Figs. S3-S5, S7, S8]) has been condensed. The focus of this section is now on the positive parts of the data: the PE association (OV Fig. 3 [NV Fig. 2]) and the in vivo work (OV Figs. 5, 6 [NV Figs. 3, 4]).
• We have additionally limited the amount of inference and speculation in the manuscript.

3. Point-by-point responses to the reviewers

Reviewer #1 (Evidence, reproducibility and clarity):

MORN (membrane occupation and recognition nexus) repeat proteins are found in prokaryotes and eukaryotes. They feature characteristic repeats in their primary sequence, have been assumed to play a role in lipid binding, but remain poorly characterized on the functional and structural level. This manuscript tries to address both these questions and is organized in major parts. In the first part the authors characterize a putative role of MORN repeat proteins in lipid binding and membrane association. In the second part, the authors use X-ray crystallography to establish the structure of MORN repeat proteins and to investigate the dimerization.

As a cleverly chosen point of departure, they focus their study particularly on MORN1 from Trypanosoma brucei (TbMORN1), which is composed solely on MORN repeats. The structures of MORN repeats (from several species) in part two provide interesting insights into their mode of homotypic interactions and their role as dimerization or oligomerization devices. The lipid binding and membrane association of MORN proteins in the first part remains somewhat confusing and unclear, despite the use of a whole battery of techniques.

We anticipate that the shortening and refocusing of the lipid binding data has addressed this issue.

It is questionably, why the authors invest so many figures and words to inform the reader on negative results.

We have chosen to publicise our negative data in full because, as noted in the manuscript, there is a widespread and erroneous assumption that MORN repeats are lipid binding modules. We feel that publishing these data will allow them to act as a caveat to other researchers working on MORN repeat proteins. We have, however, addressed the reviewer's request in that we have considerably shortened the text associated with these data and have moved the corresponding figures into the supplementals.

The authors suggest that MORN proteins can bind to lipids via their hydrophobic acyl chainswhich is 'very hard to imagine under physiological conditions unless TbMORN1 is a lipid carrier and not a membrane-binding proteins. Unfortunately, a role as lipid carrier has not been rigorously tested.

The reviewer is correct that we have not specifically tested for a function as a lipid carrier protein and although this was only speculation, it has been toned down accordingly.

In this sense the first part remains somewhat immature and incoherent. Furthermore, they suggest based on the lack-of-evidence that MORN proteins do not bind membranes in vivo and in vitro.

We are not clear where this suggestion was made. Our data indicate that TbMORN1 does not directly bind membranes in vivo or in vitro, and we therefore noted that putative lipid binding by other MORN repeat proteins should be viewed with caution. Specifically, we stated in the Discussion (OV lines 955-956) that "the presence of MORN repeats in a protein should not be taken as indicative of lipid binding or lipid membrane binding without experimental evidence". Again, our expectation is that the major changes planned for the data presentation in this section will make it more coherent.

The main issue of this manuscript is, in my view, the way the data were presented.The manuscript is generally well-written, but much too long. The structural work is important and concise.

We have considerably shortened the manuscript as per the reviewer's request, and especially the section on lipid binding.

The first part, however, reports in five separate figures on a lack of membrane binding by a MORN protein and its ability to bind individual lipids. The physiologically relevance of this lipid binding is questionable as acknowledged by the authors.

We have moved two of these figures (OV Figs. 2, 4) into the supplementals section [NV Figs. S2, S6], shortened the associated text, and limited the amount of speculation.

Even though I find it important that the membrane/lipid binding ability of MORN proteins is rigorously tested, I would highly recommend to separate the current manuscript in two independent stories. Alternatively, I would recommend to reduce the first part into a single figure and to remove the most artifactual assays.

We have implemented the second of these two suggestions for the manuscript. We had already considered splitting the manuscript during the drafting stage, but rejected this possibility as the data were too intertwined. Consequently, we have opted to considerably reduce the first part, and moved OV Figs. 2 and 4 into the supplementals [NV Figs. S2, S6]. We would prefer not to remove data altogether as they are likely to have community value even if they are negative and as noted, they are of good quality.

In the current form, the first part and the second part of the manuscript remain somewhat detached from each other. The characterization of the lipid binding/membrane binding properties has a number of substantial weaknesses (e.g. use of quite different, nonphysiological buffers for membrane binding assays; use of deletion mutants for the binding assays, which do not show the full potential of oligomerization). This which makes it hard to read and confuses the reader. Even though I have no reason to doubt the conclusions by the authors, I do not think that all necessary caution has been invested to rule out other possibilities.

We believe that the shortening and refocusing of the manuscript should address these issues. For consideration of the buffer and deletion mutant points, please see responses to Major Points below.

In summary, even though the technical quality of the individual performed assays is high, there are some conceptual issues that make it hard to make a strong case based on a collection of individual, clear datasets. Even though I find the structures of the MORN proteins important, timely, and interesting, I would not recommend this study for publication in its current form. The manuscript would be more fun to read if both of the parts would be shortened substantially and more focused.

We have implemented this suggestion: the manuscript has been considerably shortened (from 20,489/135,073 to 18,555/103,988 characters/words, focused on reducing the negative lipid-binding results).

While I agree that most evidence provided on lipid/membrane binding of TbMORN1 argue against a direct role of MORN proteins in membrane binding, I feel that the experimental approach is not coherent enough. See a few major points of criticism below.

Major Points:

1. The authors decide to characterize the membrane binding of a MORN repeat protein using a deletion variant that lacks the N-terminal repeat. However, in Figure 1B they show that the N-terminal repeat is important for the formation of higher-order oligomers. While I fully understand that the presence of the most N-terminal repeat does hamper the structural work, I find it problematic to remove it for the lipid/membrane-binding assays. The formation of higher oligomeric species beyond the dimer, may be important for membrane binding/recruitment (avidity effects).

As we explained in the manuscript, the reason for not using the full-length protein for in vitro work was because it was polydisperse, and that the yields were extremely low. See OV lines 178-179 ("The yields of TbMORN1(1-15) were always very low, making this construct not generally suitable for in vitro assays".) and OV lines 411-414 ("...TbMORN1(1-15), which was polydisperse in vitro and formed large oligomers (Fig. 1B). The membrane-binding activity of these polydisperse oligomers was not possible to test in vitro, as the purification yields of TbMORN1(1-15) were always low."). Consequently, we used the longest construct that was suitable in terms of chemical and oligomeric homogeneity. Using the full-length protein would have had inherent problems with aggregation, and consequently would have compromised the data and derived results. In order to make this clear in the manuscript we edited the sentence mentioned above as follows:

“It was not possible to test the membrane-binding activity of these polydisperse oligomers in vitro however, as the purification yields of TbMORN1(1-15) were always low. As an alternative, the possible membrane association of TbMORN1(1-15) was examined in vivo."

2) (Related to point 1) I do not understand the choice of the buffers used for some of the assays. The use of pH 8.5 and NaCl concentrations of 200 mM are non-physiological.

These were the buffer conditions required to retain the protein in a monodisperse state, suitable for in vitro assays.

For CD spectroscopy, a high ionic strength was obtained by the use of 200 mM NaF. If a high ionic strength is required to prevent the formation of higher oligomers of MORN, it raises the question if the formation of higher oligomers (under physiological conditions) may also contribute to their function.

The oligomers of TbMORN1 may indeed be the most functionally relevant form of TbMORN1 but we do not currently have a means of testing this in vitro, as acknowledged in the text (OV lines 411-414, quoted above). The aim of CD spectroscopy was to assess fold integrity and stability of different constructs; we used buffers as recommended for the CD spectroscopy experiments by Kelly et al, 2005 (doi:10.1016/j.bbapap.2005.06.005) (Table 1 and section 4.2). Furthermore, the CD spectra of TbMORN(1-15) and TbMORN(2-15) (OV Fig. S1E [NV Fig. S1E]) are basically superimposable, suggesting identical secondary structure content at the concentration used for these experiments.

It is unclear, in which buffer the fluorescence anisotropy measurements were performed.

We have provided details on the buffer conditions for the fluorescence anisotropy experiments in the Materials and Methods section, NV page 23, lines 962-963.

The sucrose-loaded vesicles were hydrated in a 20 mM HEPES pH 7.4, 0.3 M Sucrose. The composition of the buffer after the addition of MORN proteins is not clear.

The Materials and Methods are now unambiguous on this point. Please see NV lines 1036- 1046: "6 μM Rhodamine B dihexadecanoyl phosphoethanolamine (Rh-DHPE) was added to all lipid mixtures to facilitate the visualisation of the SLVs. The lipid mixtures were dried under a nitrogen stream, and the lipid films hydrated in 20 mM HEPES pH 7.4; 0.3 M sucrose. The lipid mixtures were subjected to 4 cycles of freezing in liquid nitrogen followed by thawing in a sonicating water bath at RT. The vesicles were pelleted by centrifugation (250,000 × g, 30 min, RT) and resuspended in 20 mM HEPES pH 7.4, 100 mM KCl to a total lipid concentration of 1 mM. SLVs were incubated with 1.5 μM purified TbMORN1(2-15) in gel filtration buffer (20 mM Tris-HCl pH 8.5, 200 mM NaCl, 2% glycerol, 1 mM DTT) at a 1:1 ratio (30 min, RT)." The liposomes were at physiological pH and close to physiological ionic strength.

Despite the use of an impressive array of techniques, this first part of the manuscript remains somewhat immature and incoherent. Due to the use of constructs that have not the full ability to oligomerize (point 1) and due to the inconsistent use of experimental conditions, it is hard to draw firm conclusions from this first part.

Any biochemical study is conducted within the constraints of the choice of construct and the choice of buffer conditions, and the data are valid within those parameters. This applies as much to positive data as to negative data, so we are not clear why the reviewer is placing such emphasis on this point. In the case of the LiMA data, which are the most unbiased and comprehensive dataset in the manuscript, these experiments were well-controlled and there were also domains present that were recruited to membranes under the buffer conditions, allowing us to rule out that the assay conditions were completely unsuitable. Validating negative results should be done as carefully and with as many orthogonal approaches as the validation of positive results. The reviewer acknowledges below that "the data point in the direction that MORN proteins (or at least TbMORN1) does not directly bind to membranes". This is the conclusion that we wanted to communicate.

For example: In Figure 2E TbMORN(2-15) does show some concentration-dependent binding, which -however- is interpreted as background binding. What are the results using this assay (or better: a liposome floatation assay) when using full-length TbMORN(1-15) in a more physiological buffer?

As noted already, it is not possible to use the TbMORN1(1-15) construct for in vitro assays owing to the extremely low yields and polydisperse nature of the protein. The excess fulllength protein was associated with the cytosolic fraction and not the membrane fraction in vivo (OV Fig. 6B [NV Fig. 4B]).

The statement that MORN proteins bind to lipids, but not to liposomes/membranes is -in my view- not sufficiently addressed to make a strong case.

At no point do we suggest that MORN repeat proteins in general bind to lipids and not to liposomes/membranes. On the contrary, and as detailed in the manuscript, we set out to assay the lipid binding activity of TbMORN1, found that it appears to bind to lipids but not to liposomes/membranes, and have therefore cautioned that lipid or liposome/membrane binding of other MORN repeat proteins must be tested experimentally before claims of function are made.

3) The physiological relevance of lipid binding to MORN proteins remains obscure (as also acknowledged by the authors). Does the binding of PE lipids to the MORN protein have a physiological role? Does the binding of fluorescent PI(4,5)P2 point to a physiological role of MORN proteins?

These are interesting questions that we would like to address in future work.

4) In light of recent data from the Chris Stefan lab (PMID: 31402097) a co-incidence detection of PI(4,5)P2, PS, and cholesterol seems possible. Can the authors address this possibility?

Again, the involvement of cholesterol, PS, and PI(4,5)P2 would be interesting questions for subsequent work but are beyond the scope of the present study. We did partially address this issue in our use of PI(4,5)P2, POPC and cholesterol containing liposomes in liposome cosedimentation assays, which showed no binding (OV Fig. S3A [NV Fig. S4A]).

Furthermore, the role of Ca2+ signaling / Ca2+ ions has not been addressed. In light of the important role of Ca2+ for the recognition of PI(4,5)P2 (PMID: 28177616), this point should be addressed.

We carried out liposome pelleting assays in the presence of Ca2+ and Mg2+, and saw no binding by TbMORN1(2-15) in either condition (see data below). These data were not included in the MS because of the insufficient number of technical replicates available.

5) For characterizing the binding of lipids to MORN proteins, the authors use nonphysiological fluorescent and short-chain lipid analogues at concentrations, which are unlikely to occur for endogenous PIPs in the cytosol of cells. Why choosing such an artificial system? Why introducing this system at length, if other -less artifact-prone- assays are available? I would recommend to not feature this assay as prominently as it was in the current study.

Our aim was to stick to using the same fluorophore throughout all the experiments. The choice of short-chain lipids was constrained by what was commercially available with the BODIPY TMR fluorophore. We have implemented the reviewer's suggestion in the manuscript, and the text associated with the fluorescence anisotropy assays has been considerably shortened. We are aware that the chosen concentration of the fluorescent lipids was out of physiological range, but the requirements of the fluorescence anisotropy itself necessitated a compromise. The possible shortcomings of the fluorescence anisotropy assays are, we believe, more than amply compensated by the LiMA data.

6) How would PE find its way to the lipid binding region in MORN? Would it diffuse to the MORN protein via the aqueous phase or would the MORN protein pickup PE form membranes up collision? The authors should address this point, by separating the lipiddepleted MORN protein from donor-vesicles containing PE by a dialysis membrane. If PE would not find its way to the lipid binding site of MORN, this would imply that MORN protein can extract lipids only upon colliding with the membrane. What is the stoichiometry of PE to MORN?

These are all interesting questions that we would like to pursue in subsequent work, but we feel that they are beyond the scope of the present study. Until we have conditions suitable for obtaining high yields and monodisperse populations of the full-length protein, which probably also necessitates developing conditions for controlled oligomerisation, it would be premature to start this. As to how it picks up PE: it is well known that specific lipid binding/chaperoning proteins can deliver their lipid cargo to other proteins. Additionally, proteins that bind lipids use hydrophobic domains to both interact with and sequester fatty acids and/or lipids from membranes. The literature is populated with lots of such examples. https://www.sciencedirect.com/science/article/pii/S0092867416310765.

Despite my critique raised above, I agree with the authors that the data point in the direction that MORN proteins (or at least TbMORN1) does not directly bind to membranes. Their data, however, would still be consistent with a role as lipid transfer protein and a recruitment of MORN proteins to the membrane by other proteins. Have the authors performed any additional experiments in this direction? Also, the potential role of palmitoylation is only mentioned in the discussion (page 22), while palmitoylation would provide a simple means for membrane recruitment.

We are glad that the reviewer concurs with our main conclusion. We agree, as noted in the discussion, that a role as a lipid transfer protein might still be possible, and this is something that we would like to pursue in follow-up work. We have not yet performed any additional experiments in this direction. Concerning palmitoylation, the predictions using the CSS-Palm software were always weak and ambiguous, and in addition the best candidate cysteine residue was Cys351, which is in our structure engaged in the disulphide bond observed in the C2 crystal form. We feel that this is something to keep in mind, but is not yet a strong enough hypothesis to pursue intensively.

Minor Points:

Figure 1B: The authors should provide information on the void volume of the column.

Implemented in the figure legend (7.2 ml).

Page 17, line 696-701: The authors point out that the C2 crystal form is stabilized by two disulfide bridges. The authors should comment on the physiological relevance of these disulfide bridges.

Given the reducing environment of the cytosol, it is an open question as to whether these disulphide bridges exist in vivo. We would prefer not to speculate on this point, as we do not feel it would be productive.

Page 18, line 734-740: The authors should provide data on the potential role of Zn2+ on MORN function in a physiological context. The section describing that the dimer is stabilized by Zn2+ ions (pages 18 and 19) lacks a discussion if Zn2+ are functionally relevant. There is only a beautiful sequence analysis and a discussion of the conservation of the Zn2+ coordinating residues. Can the authors perform Zn2+ titrations and SEC-MALS experiments (or alternatives such as SAXS) to show that Zn2+ indeed affects the oligomeric state of only the PfMORN, but not the other MORN proteins that form alternative dimers?

The known requirement for zinc ions in Plasmodium growth was already noted (OV lines 992- 993, Marvin et al., 2012), and is, we believe, sufficient to address the issue of physiological relevance at this stage. The zinc ions are predicted to affect the architecture of the apicomplexan (Plasmodium, Toxoplasma) MORN1 protein dimers, not their oligomeric state. For PfMORN1, SEC-MALS and SAXS were carried out in 20 mM Tris-HCl pH 7.5, 100 mM NaCl with no zinc present. When EDTA was added, no change in behaviour of the protein was seen by SEC-MALS. When “TPEN”, a strong zinc chelator, was added, the protein precipitated in SEC-MALS experiments.

Reviewer #1 (Significance):

A putative role of MORN proteins in membrane and lipid binding is addressed. The view the MORN proteins bind directly to membranes is challenged. Structures of dimeric MORN proteins provide important insight into the modes of dimerization.

There is a recent structure of MORN proteins (which is referenced by the authors), but I feel that additional structural work is important and justified. The work on membrane vs. lipid binding is important, but not sufficiently addressed in the current manuscript.

We are glad that the reviewer finds the structural work important and justified, although we disagree with the reviewer’s assessment of the lipid binding. As noted in the previous paragraph, our data challenge the assumption that MORN repeat proteins directly bind membranes, and we feel that this alone is a significant conceptual advance.

I would recommend to separate the study in two parts. The audience is likely to confused (or bored) by the lengthy discussion on whether or not MORN proteins bind lipids and or membrane or not.

We would prefer to implement the reviewer's other suggestion, namely that the manuscript is considerably shortened and less focus given to the negative data on lipid binding.

I am not an expert in structural biology, but have a fair understanding of structural biology. I have worked on lipid binding proteins and have a very good understanding of lipid/membrane-binding assays.

---

Reviewer #2 (Evidence, reproducibility and clarity):

Summary

The manuscript describes an extensive and detailed investigation into the structure and function(s) of MORN domains. It has to be acknowledged that, despite the considerable amount of work reported, the conclusions are rather limited. From a technical viewpoint, the experiments have been appropriately executed and, generally, I concur with the conclusions drawn. However, the manuscript is over-long: in general, I would recommend concentrating on positive conclusions which can be drawn from the data and avoid excessive speculation or inference (some examples given below).

We are glad that the reviewer is satisfied with the technical quality of the work and (in general) the validity of the conclusions. We acknowledge that the original submission was fairly long, and have considerably shortened the revised manuscript and focused more on the positive conclusions in order to implement this suggestion.

Major Comments

There are three general- perhaps rather obvious- points to make. First, there is no particular reason to think that conservation of structure necessarily indicates conservation of a particular function. There seems to be an implicit assumption that MORN domains are associated with a specific, well-defined biological function. Given their diversity, are there particular reasons to think that this is the case?

The reviewer is exactly right that there is an implicit assumption that MORN domains are associated with a specific, well-defined function: specifically, lipid binding. It is this assumption, which has been widely circulated in the almost complete absence of experimental evidence, that we are challenging. We agree that MORN repeats are likely to be capable of multiple functions, and protein-protein interactions are now better supported than protein-lipid interactions.

Second, a strategy which examines the properties of just the recombinant MORN domains in vitro, removed from the context of the whole protein (eg junctophilin) or- importantly- its interacting partners in vivo, has obvious limitations. Frequently a reductionist approach is successful; however, in this case, MORN domains appear to be less tractable to that kind of approach. For all the in vitro binding and structural experiments presented, there is always a concern that the absence of other parts of the relevant MORN-containing protein or its partners could explain failure or inconsistency of in vitro biological activity measurements.

Again, the reviewer is right that there is an inherent contextual limitation to any in vitro work that utilises a single protein, but this is a concern that - by definition - could be raised about any in vitro study utilising a single protein. It should be noted that we have also carried out in vivo experiments using TbMORN1 (OV Figs. 5, 6 [NV Figs. 3, 4]).

Third, the possibility that MORN domains might mediate interactions with other proteins seems to be given little consideration, in spite of the Li et al (2019) paper. An experimental strategy which looked for binding partners (eg by pulldown assay) might have provided more insight.

These data are already in the literature. A previous study by the same team (Morriswood et al., 2013) used proximity-dependent biotin identification to identify candidate binding partners and near neighbours of TbMORN1.

In order to stress this point we added the following sentence in the discussion section, NV pages 18-19, lines 774-778.

“The concluding data presented here suggest that TbMORN1 utilises this oligomerisation capacity to build mesh-like assemblies, which can reach considerable size in vitro (Fig. 7G). These mesh-like assemblies may reflect the endogenous organisation of the protein in vivo, where a number of binding partners have already been identified (Morriswood et al., 2013)”.

Minor Comments

1. In the abstract and elsewhere the authors refer to a possible function of MORN domains as 'dimerisation and oligomerisation devices' (line 53). What is the evidence that dimer formation is important for function in vivo?

This is an interesting and important question and one that we would like to address in future work. We did attempt to generate trypanosome cell lines that inducibly expressed monomeric TbMORN1 (the double mutant, where the point mutations were simultaneously introduced in the dimerisation interface in repeats 13 and 14), but no expression of the ectopic protein was ever observed (9 separate clones obtained in 3 independent transfections). This might indicate the importance of the dimeric state in vivo, perhaps hinting that dimerisation is important for protection from degradation. In general, proteins assuming higher oligomeric states in homo- or heteromeric assemblies benefit from increased robustness in the cellular environment and optimised activity by the following means:

• Increased stability by decreasing the surface area/volume ratio
• Simple construction of larger complexes
• Allosteric regulation
• Co-localisation of distinct biological functions
• Substrate channelling
• Protection from aggregation or degradation

Which or which combination of the factors is relevant for TbMORN1 being a functional dimer in vivo is difficult to say at this point.

1. Did the authors attempt to co-crystallize TbMORN1(7-15) with PI(4,5)P2?

No. For crystallisation, we used lysine methylated samples, and by doing this we neutralised positively-charged potential binding sites which would have interacted with the negatively charged lipid headgroup. We did not observe any bound lipids in the electron density maps obtained from the crystals.

1. Fig 2C: did the authors also estimate binding stoichiometry as well as the equilibrium binding constants for these data? This should be determined by fitting a single binding site model to the data. Other methods (eg ITC) can probably determine this with more accuracy. The value of stoichiometry is sometimes forgotten in such binding measurements- is one ligand bound per monomer or dimer, for example?

We discussed estimation of the binding stoichiometry in the fluorescence anisotropy assays at some length, but the conclusion was that the required experiments would contain too many approximations to provide high-confidence data. We did use ITC and also MST, but did not observe any binding with these assays.

1. Lines 674-678 I found it hard to work out whether these constructs harbour the natural C-terminal sequence without truncation or addition of an affinity tag. I think the answer is 'yes' but it was difficult working this out from the details in M&M.

TbMORN1(7-15) crystallisation was with a C-terminal Strep tag; TgMORN1(7-15) and PfMORN1(7-15) had their affinity tags removed by protease treatment prior to crystallisation. We have clarified this point in the M&M, page 29, lines 1189-1192: “Crystallisation of TbMORN1(7-15) (with a C-terminal Strep tag), TgMORN1(7-15) and PfMORN1(7-15) (both with affinity tags removed) was performed at 22 °C using a sitting-drop vapour diffusion technique and micro-dispensing liquid handling robots (Phoenix RE (Art Robbins Instruments) and Mosquito (TTP labtech).”

1. Lines 688-694 The PISA interface analysis is useful here in distinguishing crystal contacts from those which persist in solution. The discussion of the results is unclear, however, on this critical point: were the dimer interfaces the only contacts which were significant in the various crystal forms?

Yes, correct. PISA showed that the described dimerisation contacts were the only significant ones in the various crystal forms. Other crystals contacts had typically low P-values and poor ΔG and small “radar” surface in the complexive PISA analysis.

In the case of both TbMORN1 crystal forms and in the case of the TgMORN1 P43212 crystal form we have a dimer in the asymmetric unit, while in the case of the PfMORN1 and TgMORN1 P6222 form we have one molecule in the asymmetric unit, and the dimer is created by the crystallographic twofold axis. In the latter cases the quaternary structure resulting from the symmetry operations was the top-scoring one considering either P-values and/or the number of stabilising interactions buried surface area.

1. Lines 754-763 This paragraph seems rather speculative and is a good example where the text could be cut down.

If the line citation is correct, then we disagree with this assessment and would prefer not to implement it. The paragraph in question concerns a detailed and very precise discussion of the side chain interactions that stabilise the V-shaped forms of TgMORN1 and PfMORN1.

1. Line 765-788 This section is also rather overdone: such observations are only useful if they are subsequently tested by recording dimer conformation for a representative selection of MORN dimers from different species.

Again, we disagree with the reviewer's assessment of this analysis. The analysis has considerable predictive power and already has some experimental validation via the SAXS observation that PfMORN1 is capable of forming extended dimers in solution (OV Fig. 10C [NV Fig. 7C]).

1. Lines 800-801 I don't think this statement is strictly correct. The SAXS data show that PfMORN1(7-15) adopts an extended conformation, with no evidence of the 'V' shaped structure. Related to that point, from what I could glean from the SAXS Methods section, all solution conditions for these experiments were conducted without Zn2+? If some dimer interfaces require Zn2+, should it not be included?

We have clarified this statement. The SAXS experiments were conducted without zinc, and, as we have stressed, the V-shaped form of TgMORN1 and PfMORN1 was only ever observed in the crystals. For PfMORN1, SEC-MALS and SAXS were carried out in 20 mM Tris-HCl pH 7.5, 100 mM NaCl with no zinc present. When EDTA was added, no change in behaviour of the protein was seen by SEC-MALS. When “TPEN”, a strong zinc chelator, was added, the protein precipitated in SEC-MALS experiments.

Reviewer #2 (Significance):

There is certainly value in establishing that MORN domains do not, in vitro, appear to bind to lipid vesicles, and to define their lipid binding capability (although it is rather complex). The crystal structures and SAXS data extend the rather limited structural data on MORN domains. Despite the effort involved, conclusions about likely functions of MORN domains in vivo are rather limited.

We are glad that the reviewer acknowledges the value in challenging the assumption that MORN repeats are lipid binding devices, and that the structural data are important for expanding the knowledge base on this class of repeat motif proteins. In vivo functional work is being actively pursued at present.

My expertise lies in X-ray crystallography and protein biochemistry.

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 describes an extensive and detailed investigation into the structure and function(s) of MORN domains. It has to be acknowledged that, despite the considerable amount of work reported, the conclusions are rather limited. From a technical viewpoint, the experiments have been appropriately executed and, generally, I concur with the conclusions drawn. However, the manuscript is over-long: in general, I would recommend concentrating on positive conclusions which can be drawn from the data and avoid excessive speculation or inference (some examples given below).

Major Comments

There are three general- perhaps rather obvious- points to make. First, there is no particular reason to think that conservation of structure necessarily indicates conservation of a particular function. There seems to be an implicit assumption that MORN domains are associated with a specific, well-defined biological function. Given their diversity, are there particular reasons to think that this is the case? Second, a strategy which examines the properties of just the recombinant MORN domains in vitro, removed from the context of the whole protein (eg junctophilin) or- importantly- its interacting partners in vivo, has obvious limitations. Frequently a reductionist approach is successful; however, in this case, MORN domains appear to be less tractable to that kind of approach. For all the in vitro binding and structural experiments presented, there is always a concern that the absence of other parts of the relevant MORN-containing protein or its partners could explain failure or inconsistency of in vitro biological activity measurements. Third, the possibility that MORN domains might mediate interactions with other proteins seems to be given little consideration, in spite of the Li et al (2019) paper. An experimental strategy which looked for binding partners (eg by pulldown assay) might have provided more insight.

Minor Comments

1. In the abstract and elsewhere the authors refer to a possible function of MORN domains as 'dimerisation and oligomerisation devices' (line 53). What is the evidence that dimer formation is important for function in vivo?
2. Did the authors attempt to co-crystallize TbMORN1(7-15) with PI(4,5)P2?
3. Fig 2C: did the authors also estimate binding stoichiometry as well as the equilibrium binding constants for these data? This should be determined by fitting a single binding site model to the data. Other methods (eg ITC) can probably determine this with more accuracy. The value of stoichiometry is sometimes forgotten in such binding measurements- is one ligand bound per monomer or dimer, for example?
4. Lines 674-678 I found it hard to work out whether these constructs harbour the natural C-terminal sequence without truncation or addition of an affinity tag. I think the answer is 'yes' but it was difficult working this out from the details in M&M.
5. Lines 688-694 The PISA interface analysis is useful here in distinguishing crystal contacts from those which persist in solution. The discussion of the results is unclear, however, on this critical point: were the dimer interfaces the only contacts which were significant in the various crystal forms?
6. Lines 754-763 This paragraph seems rather speculative and is a good example where the text could be cut down.
7. Line 765-788 This section is also rather overdone: such observations are only useful if they are subsequently tested by recording dimer conformation for a representative selection of MORN dimers from different species.
8. Lines 800-801 I don't think this statement is strictly correct. The SAXS data show that PfMORN1(7-15) adopts an extended conformation, with no evidence of the 'V' shaped structure. Related to that point, from what I could glean from the SAXS Methods section, all solution conditions for these experiments were conducted without Zn2+? If some dimer interfaces require Zn2+, should it not be included?

Significance

There is certainly value in establishing that MORN domains do not, in vitro, appear to bind to lipid vesicles, and to define their lipid binding capability (although it is rather complex). The crystal structures and SAXS data extend the rather limited structural data on MORN domains. Despite the effort involved, conclusions about likely functions of MORN domains in vivo are rather limited. My expertise lies in X-ray crystallography and protein biochemistry.

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

MORN (membrane occupation and recognition nexus) repeat proteins are found in prokaryotes and eukaryotes. They feature characteristic repeats in their primary sequence, have been assumed to play a role in lipid binding, but remain poorly characterized on the functional and structural level. This manuscript tries to address both these questions and is organized in major parts. In the first part the authors characterize a putative role of MORN repeat proteins in lipid binding and membrane association. In the second part, the authors use X-ray crystallography to establish the structure of MORN repeat proteins and to investigate the dimerization.

As a cleverly chosen point of departure, they focus their study particularly on MORN1 from Trypanosoma brucei (TbMORN1), which is composed solely on MORN repeats. The structures of MORN repeats (from several species) in part two provide interesting insights into their mode of homotypic interactions and their role as dimerization or oligomerization devices. The lipid binding and membrane association of MORN proteins in the first part remains somewhat confusing and unclear, despite the use of a whole battery of techniques. It is questionably, why the authors invest so many figures and words to inform the reader on negative results. The authors suggest that MORN proteins can bind to lipids via their hydrophobic acyl chains- which is 'very hard to imagine under physiological conditions unless TbMORN1 is a lipid carrier and not a membrane-binding proteins.' Unfortunately, a role as lipid carrier has not been rigorously tested. In this sense the first part remains somewhat immature and incoherent. Furthermore, they suggest based on the lack-of-evidence that MORN proteins do not bind membranes in vivo and in vitro.

The main issue of this manuscript is, in my view, the way the data were presented.The manuscript is generally well-written, but much too long. The structural work is important and concise. The first part, however, reports in five separate figures on a lack of membrane binding by a MORN protein and its ability to bind individual lipids. The physiologically relevance of this lipid binding is questionable as acknowledged by the authors. Even though I find it important that the membrane/lipid binding ability of MORN proteins is rigorously tested, I would highly recommend to separate the current manuscript in two independent stories. Alternatively, I would recommend to reduce the first part into a single figure and to remove the most artifactual assays. In the current form, the first part and the second part of the manuscript remain somewhat detached from each other. The characterization of the lipid binding/membrane binding properties has a number of substantial weaknesses (e.g. use of quite different, non-physiological buffers for membrane binding assays; use of deletion mutants for the binding assays, which do not show the full potential of oligomerization). This which makes it hard to read and confuses the reader. Even though I have no reason to doubt the conclusions by the authors, I do not think that all necessary caution has been invested to rule out other possibilities.

In summary, even though the technical quality of the individual performed assays is high, there are some conceptual issues that make it hard to make a strong case based on a collection of individual, clear datasets. Even though I find the structures of the MORN proteins important, timely, and interesting, I would not recommend this study for publication in its current form. The manuscript would be more fun to read if both of the parts would be shortened substantially and more focused. While I agree that most evidence provided on lipid/membrane binding of TbMORN1 argue against a direct role of MORN proteins in membrane binding, I feel that the experimental approach is not coherent enough. See a few major points of criticism below.

Major Points:

1) The authors decide to characterize the membrane binding of a MORN repeat protein using a deletion variant that lacks the N-terminal repeat. However, in Figure 1B they show that the N-terminal repeat is important for the formation of higher-order oligomers. While I fully understand that the presence of the most N-terminal repeat does hamper the structural work, I find it problematic to remove it for the lipid/membrane-binding assays. The formation of higher oligomeric species beyond the dimer, may be important for membrane binding/recruitment (avidity effects).

2) (Related to point 1) I do not understand the choice of the buffers used for some of the assays. The use of pH 8.5 and NaCl concentrations of 200 mM are non-physiological. For CD spectroscopy, a high ionic strength was obtained by the use of 200 mM NaF. If a high ionic strength is required to prevent the formation of higher oligomers of MORN, it raises the question if the formation of higher oligomers (under physiological conditions) may also contribute to their function. It is unclear, in which buffer the fluorescence anisotropy measurements were performed. The sucrose-loaded vesicles were hydrated in a 20 mM HEPES pH 7.4, 0.3 M Sucrose. The composition of the buffer after the addition of MORN proteins is not clear. Despite the use of an impressive array of techniques, this first part of the manuscript remains somewhat immature and incoherent. Due to the use of constructs that have not the full ability to oligomerize (point 1) and due to the inconsistent use of experimental conditions, it is hard to draw firm conclusions from this first part. For example: In Figure 2E TbMORN(2-15) does show some concentration-dependent binding, which -however- is interpreted as background binding. What are the results using this assay (or better: a liposome floatation assay) when using full-length TbMORN(1-15) in a more physiological buffer? The statement that MORN proteins bind to lipids, but not to liposomes/membranes is -in my view- not sufficiently addressed to make a strong case.

3) The physiological relevance of lipid binding to MORN proteins remains obscure (as also acknowledged by the authors). Does the binding of PE lipids to the MORN protein have a physiological role? Does the binding of fluorescent PI(4,5)P2 point to a physiological role of MORN proteins?

4) In light of recent data from the Chris Stefan lab (PMID: 31402097) a co-incidence detection of PI(4,5)P2, PS, and cholesterol seems possible. Can the authors address this possibility? Furthermore, the role of Ca2+ signaling / Ca2+ ions has not been addressed. In light of the important role of Ca2+ for the recognition of PI(4,5)P2 (PMID: 28177616), this point should be addressed.

5) For characterizing the binding of lipids to MORN proteins, the authors use non-physiological fluorescent and short-chain lipid analogues at concentrations, which are unlikely to occur for endogenous PIPs in the cytosol of cells. Why choosing such an artificial system? Why introducing this system at length, if other -less artifact-prone- assays are available? I would recommend to not feature this assay as prominently as it was in the current study.

6) How would PE find its way to the lipid binding region in MORN? Would it diffuse to the MORN protein via the aqueous phase or would the MORN protein pickup PE form membranes up collision? The authors should address this point, by separating the lipid-depleted MORN protein from donor-vesicles containing PE by a dialysis membrane. If PE would not find its way to the lipid binding site of MORN, this would imply that MORN protein can extract lipids only upon colliding with the membrane. What is the stoichiometry of PE to MORN?

Despite my critique raised above, I agree with the authors that the data point in the direction that MORN proteins (or at least TbMORN1) does not directly bind to membranes. Their data, however, would still be consistent with a role as lipid transfer protein and a recruitment of MORN proteins to the membrane by other proteins. Have the authors performed any additional experiments in this direction? Also, the potential role of palmitoylation is only mentioned in the discussion (page 22), while palmitoylation would provide a simple means for membrane recruitment.

Minor Points:

Figure 1B: The authors should provide information on the void volume of the column.

Page 17, line 696-701: The authors point out that the C2 crystal form is stabilized by two disulfide bridges. The authors should comment on the physiological relevance of these disulfide bridges.

Page 18, line 734-740: The authors should provide data on the potential role of Zn2+ on MORN function in a physiological context. The section describing that the dimer is stabilized by Zn2+ ions (pages 18 and 19) lacks a discussion if Zn2+ are functionally relevant. There is only a beautiful sequence analysis and a discussion of the conservation of the Zn2+ coordinating residues. Can the authors perform Zn2+ titrations and SEC-MALS experiments (or alternatives such as SAXS) to show that Zn2+ indeed affects the oligomeric state of only the PfMORN, but not the other MORN proteins that form alternative dimers?

Significance

A putative role of MORN proteins in membrane and lipid binding is addressed. The view the MORN proteins bind directly to membranes is challenged. Structures of dimeric MORN proteins provide important insight into the modes of dimerization.

There is a recent structure of MORN proteins (which is referenced by the authors), but I feel that additional structural work is important and justified. The work on membrane vs. lipid binding is important, but not sufficiently addressed in the current manuscript.

I would recommend to separate the study in two parts. The audience is likely to confused (or bored) by the lengthy discussion on whether or not MORN proteins bind lipids and or membrane or not.

I am not an expert in structural biology, but have a fair understanding of structural biology. I have worked on lipid binding proteins and have a very good understanding of lipid/membrane-binding assays.

Preprint Review

This preprint was reviewed using eLife’s Preprint Review service, which provides public peer reviews of manuscripts posted on bioRxiv for the benefit of the authors, readers, potential readers, and others interested in our assessment of the work. This review applies only to version 1 of the manuscript.

Summary

This paper analyzes the evolution of the KRAB-containing zinc finger protein (KZFP) family of proteins. While the reviewers were all interested in the topic, several major concerns came up during review. These include technical limitations of the methods chosen to analyze this challenging protein family (e.g., determination of orthology, selection analysis, and so on), and that new ideas, including claims about non-coding evolution and positive selection, are not convincingly supported by the analysis presented.

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

R-We would like to thank the reviewers for their constructive feedback. We respond to all the reviewers points below. We highlighted major changes introduced to the manuscript in response to both reviewers’ comments in the attached revised version of the manuscript.

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

The work described in this manuscript by Tan and Marques aims to address if the splicing of enhancer-associated long noncoding RNAs (elncRNAs) has a direct impact on enhancer activity or just reflects their cognate's enhancer's high activity.

For this purpose, the authors started by integrating RNA-seq data for human lymphoblastoid cell lines with ENCODE enhancer annotations and ChIP-seq data for enhancer function-associated chromatin modifications to show that multi-exonic elncRNAs are more transcriptionally active than single-exonic elncRNAs and eRNAs. They then show that regions flanking elncRNA splice sites are enriched in splicing-associated sequence elements and that these are under stronger purifying selection (suggesting some functional relevance), both when compared to promoter-associated lncRNAs. They also show the concomitance of cis-disrupted splicing in elncRNAs and drops in expression in their target genes. Finally, they use causal inference analysis of joint seQTLs and joint scQTLs to investigate the causal relationship between splicing of elncRNAs and expression of putative gene targets and chromatin states at elncRNA cognate enhancers, respectively. They conclude that, in both cases, most associations are causally mediated by splicing of elncRNAs and therefore that this contributes to their enhancer activity.

This manuscript is generally well written, targets an original question and potentially sets the seeds for a new exciting line of research on transcriptional regulation, by providing some evidence for the functional relevance of the splicing of elncRNAs. However, the overlooking of some important aspects of the regulation of RNA splicing led to biases in the design of data analyses and in the interpretation of the biological implications of some results that need to be dealt with before the described work can be considered sound enough for publication.

R:We would like to thank the reviewer for taking the time to assess our manuscript and for the constructive comments.

**Essential revisions:**

1.Results, page 10, lines 21-24: The statement that "the impact of SS variants on gene splicing efficiency depends on the total number of alternative transcripts and exons" is not properly substantiated. The four examples given in Figure S3 do not illustrate any dependence or trend. If such dependence is "expected" the underlying concept must be explained, i.e. why the impact of SS variants on splicing efficiency should depend on the number of alternative transcripts and exons. The reader is unrealistically expected to be used to the chosen splicing efficiency metric to intuit its dependence on the number of exons. Moreover, it is not obvious where the dependence on the number of alternative transcripts comes from, particularly given that alternative splicing (e.g. the skipping of a neighbouring exon, if internal) is not profiled.

R-In the previous version of the manuscript, we estimated gene level changes in splicing, which includes all alternative splicing events within a gene. Therefore, the more exons an elncRNA has, the more “diluted” we expect the overall impact of SS variant on elncRNA splicing efficiency to be. After considering the reviewer’s comment, we realized that only splicing events directly impacted by the SS variant should be considered in this analysis.

In the revised version of the manuscript, we considered only alternative splicing events that include the splice donor acceptor site changed by the SS variant, and are therefore a direct consequence, of the SS variant. As suggested by the reviewer, for these splicing events, we report the fold difference in Percentage-Spliced-In (PSI) (estimated by Leafcutter (Li et al. 2018)) between samples that carry reference and alternative alleles at these SS variants. To further illustrate these changes, we now include a diagram, for each SS variant, with differential splicing information and the fold difference in PSI for each affected splicing events (Figure 3B,C, Supplementary Figure S3). In addition, the overall change across all affected splicing events is also plotted in Figure 3D and Supplementary Figure S4.

We have modified this section to account for this and the next comment from the reviewer.

“To estimate the impact of SS variant on splicing efficiency, we calculated the Percentage-Spliced-In (PSI) (Li et al. 2018) per individual and for each elncRNA splicing event involving the splice donor or acceptor site disrupted by the SS variants (Figure 3B,C, Supplementary Figure S3). PSI measures exon inclusion and considers spliced reads spanning exon junctions (Li et al. 2018). We compared the average difference in PSI, as a proxy for change in splicing efficiency, of all affected splicing events between individuals that carry the reference and alternative canonical splice donor/acceptor sites (GT-AG). Alongside decreased exon inclusion, SS variants can also promote exon skipping events (Figure 3B,C, Figure S3). Despite some increase in exon skipping, SS variants are associated with an overall decrease in splicing efficiency (Figure 3D and Supplementary Figure S4).” (Page 10).

Along these same lines, and more importantly, why haven't the authors looked at the possibility that a variant disrupting a splice site would lead to skipping of the neighbouring exon (if internal)? Given how the spliceosome operates (in terms of intron and exon recognition), wouldn't this be the most likely scenario? When calculating the splicing index, are reads spanning junctions between non-consecutive exons considered? Otherwise, not profiling alternative isoforms generated by exon skipping will necessarily bias splicing efficiency quantifications by overlooking fully efficient splicing associated with such isoforms. Similarly, how did the authors make sure that splicing changes did not bias elncRNA expression estimates? How was the effective transcript length determined for the calculation of RPKMs? The authors need to make these methodological clarifications, as well as why exon skipping was not considered as a splicing disruption with potential functional implications. Calculating the percent spliced-in (PSI) for all internal exons would be much informative.

R-Regarding the methodology, what we refer to as splicing efficiency is Percentage-Spliced-In (PSI). We calculated PSI for all, including alternative, splicing events. We now make this clearer throughout the manuscript and in the figure axis/legends.

As detailed in the methods section, to minimize the impact of alternative splicing on gene expression estimates, we quantified expression at the gene, and not at the transcript, level using HTSeq across all annotated exons. This approach allows us to assess elncRNA and target gene expression while masking differences in alternative transcript abundance, which are not relevant in the context of this analysis.

As suggested by the reviewer, instead of considering PSI of all possible splicing events of the gene, in the revised version of the manuscript, we considered only splicing events that are directly impacted by the SS variant. This change does not impact our conclusions, but certainly provides a better understanding of how SS variants impact splicing and we would like to thank the reviewer for raising this point. As predicted by the reviewer and as expected given how the spliceosome operates, exon skipping is a frequent outcome of SS variants. However, the increase in exon skipping is not sufficient to compensate for the decrease in the inclusion of these exons, which is directly impacted by the SS variants. This is demonstrated by the lower overall splicing efficiency for each elncRNAs in individuals that carry SS variants that disrupt canonical splice/donor acceptor sites (Figure 3D and Supplementary Figure S4).

3.All results in panels 3B-F are presented as fold differences. It is actually not clear what those differences refer to. For instance, the grey boxes are the distributions of the fold differences in splicing index / expression between individuals carrying reference alleles and what?

R-The boxplots represent the distribution of the fold difference in PSI or expression for each individual relative to the median PSI or expression in individuals with the reference genotype. As expected, the distribution of log fold difference in either PSI or expression for individuals carrying the reference allele is centered at 0.

We have clarified this in the methods section and figure legends.

4.It is expectable that most joint seQTLs result from variants directly impacting splicing in cis. As the quantification of splicing is noisier than that of expression, a stronger effect is required for the detection of an sQTL than an eQTL. In other words, joint seQTLs are essentially sQTLs. This illustrated by the example in Figure 4A, with the SNP in an intronic region of the elncRNA being associated with strong differences in splicing and tiny (R-We agree with the reviewer that the quantification of splicing is noisier than that of expression. However, and in contrast with the reviewer’s hypothesis, higher “noise” in splicing quantification compared to expression led to weaker associations between splicing and seQTLs, as illustrated in the figure below. This is in line with splicing being measured with higher error rate, which would ultimately lead to smaller detectable sQTL effect than what they would be with perfect measurements. This also demonstrates that since a priori, eQTLs association are stronger, if a bias exists in the causal inference analysis, it should favour detection of non-causal associations. Therefore, our approach is not biased in detecting causal seQTLs.

We agree with the reviewer that this potential bias may be a concern to readers and should be addressed. We have added this analysis to the text (Supplementary Figure S7E) and explained why the causal inference testing approach is not biased in detecting causal seQTLs.

“To assess whether this approach was biased towards the detection of causal seQTLs we compared the slope and adjusted p-value of the associations between the variant and splicing or expression for all causal seQTLs. As illustrated in Supplementary Figure S7E this analysis revealed there is no evidence that stronger sQTLs would favour causal model predictions.” (Page 16).

5.It is not totally clear what message the authors intend to convey with the result of panel 4D. Are they talking about the relative position of the variant to the elncRNA transcript or the target transcript? If the former, shouldn't the known synergy between transcription and 5´ end splicing reflect on elncRNA expression? If the latter, it is not obvious how the result connects to the mentioned synergy.

R:In Figure 4D, we show the relative position, within a transcript, of the exonic splicing junction which is associated with causal seQTLs. The enrichment in associations with splicing junctions located at 5’ end of elncRNAs is consistent with the synergy between 5’ end splicing and transcription. We clarify this in the text:

“Importantly, 90% of seQTL associations that support elncRNA splicing as a mediator of target expression are associated with splicing junctions located at the 5´ end of the transcript, which is consistent with the known synergy between transcription and 5´ end splicing (Furger et al. 2002; Damgaard et al. 2008)(Figure 4D).” (Page 17).

**Proofreading edits:**

R:We would like to thank the reviewer for identifying all the typos listed below. We have corrected them in the revised version of the manuscript.

6.Introduction, page 3, line 13: double "in".

7.Figure S2A, leftmost panel X-axis label: "intrno" instead of "intron".

8.Results, page 10, line 30: remove "of".

9.It is 5´ and 3´(prime) not 5' and 3' (apostrophe).

**Other suggestions:**

10.Violin plots (with included boxplots) would more comprehensively convey the differences in distributions than the chosen notched boxplots.

R-We thank the reviewer for this suggestion. Although we appreciate the added information a violin plot can provide, this also renders, in our opinion, their interpretation less intuitive. Because boxplots are simpler and easier to interpret, after consideration, we decided to continue using these to represent the distribution of the data.

Reviewer #1 (Significance (Required)):

It is hard for me to assess the significance of this work (beyond some evidence for the potential functional relevance of the splicing of elncRNAs) until the aforementioned concerns are addressed but it is of potential interest to the broad RNA research community.

I am a computational biologist with experience in the analysis of high-throughput transcriptomic data and a focus on transcriptional and alternative splicing regulation.

========================================================================

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

This manuscript addresses an interesting question - whether the splicing of transcriptional enhancer-associated RNAs influences their transcriptional enhancement activity. The analyses appear carefully done, using appropriate datasets and statistical methods. The authors find, for example, that marks of active chromatin are enriched near spliced elncRNAs, that splicing-related motifs of elncRNAs are under selective constraint, and that splicing of elncRNAs is associated with higher elncRNA expression and to very slightly higher expression of target genes.

R:We thank the reviewer for the constructive feedback on our manuscript. We have extended our analysis to address the reviewers concerns that we detail in our response to the comments below.

However, I did not find the main results convincing of the main conclusion for the following reasons:

1.The most direct evidence is shown in Figure 3, where SNPs that occur in 3' splice sites of elncRNA introns are explored, and it is shown that variants predicted to disrupt splicing of elncRNA introns are associated with reduced expression of target but not non-target genes. But the fold difference in expression of target genes is extremely small - a few percent - and is actually less than the fold difference in expression of the elncRNAs themselves (which appears closer to 10%), raising the question of whether elncRNA expression rather than splicing may be more important for activity. Furthermore, the entire analysis has an anecdotal quality, being based on only 4 splice-disrupting elncRNA variants. I did not find the figure at all convincing of the conclusion the authors draw from it.

R:We agree with the reviewer that our analysis is limited by the available genotyping data that is restricted to common genetic variants. Our evolutionary constraint analysis (Figure 2) indicates that variants that disrupt elncRNA splicing are depleted by natural selection and so we expected to identify a relatively small number of elncRNAs (n=4) suitable for this analysis. Despite the anticipated challenges in identifying elncRNA splice site mutations, we nevertheless believe this unbiased natural mutational analysis is analogous to experimentally disrupting splice sites of these 4 elncRNA candidates.

Regarding the strength of impact of splicing on target expression: in the absence of a comparable experiment, we could not anticipate the magnitude of the effect. We acknowledge that previous studies, which sought to completely remove splicing by either deleting all elncRNA introns (Yin et al. 2015) or terminating transcription after its 1st exon (Engreitz et al. 2016), were both associated with significantly stronger impact on elincRNA splicing and target expression than what we report here. The analysis we present here involves single nucleotide polymorphisms and so it is not surprising to have resulted in more moderate impact on overall splicing. Furthermore, whether the differences in the impact on target expression between this and previous analysis is the result of stronger effect of complete removal of splicing or a consequence of the genetic changes introduced remains unclear. The small yet consistent decrease in target expression we observed, even with minimal changes in splicing of an unbiased set of 4 candidates, is in our opinion strong evidence that modulation in elncRNA splicing is sufficient to impact, albeit moderately, target expression.

Importantly, we replicated the impact of decreased splicing on target expression of the 4 elncRNA candidates using 89 samples of Yoruba (YRI) population from the Geuvadis dataset (Supplementary Figure S5). The robustness of the mutational study consistently supports the physiologically relevant effect of elncRNA splicing on cognate enhancer function.

As pointed out by the reviewer, elncRNA SS variants led to stronger impact on the expression of the elncRNAs compared to that of their targets (Figure 3F,H and Supplementary Figure S4), which suggests that target expression regulation is likely a consequence of changes in elncRNA expression as a result of changes in its splicing. This is described as our working model in the discussion section of the manuscript.

2.Figure 4 uses a causal inference approach and involves larger datasets. While causal inference can be a useful tool to identify candidate causal relationships, it does not prove causality, which still requires some sort of experimental perturbation. Thus, I found these results suggestive but still not satisfying to justify, e.g., the title of the paper or claims made in the abstract. As in Figure 3, the specific example shown in Fig. 4A again shows a relatively tiny effect on target gene expression, which again appears to be a few percent at most.

R:For the reasons explained above, we had no expectation that the effect size of the association between elncRNA splicing and target expression would be high. It is nevertheless key that these associations are robust, which would provide reliable support for our hypothesis. To assess this, we used 2 independent datasets to replicate elncRNA target associations with sQTL variants associated with elncRNA splicing: 1) 147 LCL samples from GTEx and 2) 31,684 blood samples from eQTLgen. Using these datasets, we replicated the association between sQTL and target expression for targets of up to 77% of elncRNAs. As expected, replicated associations have significantly higher effect size in both datasets (Supplementary Figure S9) and 1.2 times more associations can be replicated in the eQTLgen blood samples with a larger cohort size. LCL-specific effect of elncRNA splicing likely explains why not all associations are replicated in these blood samples. We report these analyses in the manuscript (Supplementary Figure S9).

We agree with the reviewer that the causal inference analysis is only suggestive per se. However, we would argue that conclusions of the present manuscript do not rely on this analysis alone, but instead on the combined evidence of several experiments, including the natural mutational analysis that is analogous to the experiment the reviewer proposes.

Considering the reviewers concern, we realized that previous version of Figure 4A did not reflect the average strength of the association between seQTL variant and target expression (median=0.319, ranging from 0.16 to 0.81, Rebuttal Figure 1). For this reason, we replaced the previous illustration by a more representative example (Figure 4A).

The text illustrating reproducibility of our results in GTEx and eQTLgen have been added to Page 17 of the manuscript.

“We used two independent datasets to assess the robustness of elncRNA target association with sQTL variants we predict to be associated with the splicing of these elncRNAs in LCLs. Using a smaller cohort of LCLs (n=147 (GTEx Consortium 2013)), we found a significant association in the same direction between sQTL and target expression for targets of 70% of elncRNAs (45% of variants). A larger fraction of associations (77% of elncRNAs and 52% of variants) could be replicated in a larger cohort of blood samples (n=31,684 (Võsa et al. 2018)). The difference in size between these two cohorts is likely to explain the difference in replication rate. The association between elncRNA splicing variants and target expression that were replicated have significantly higher effect size relative to non-replicated associations (Supplementary Figure S9). Furthermore, LCL-specific effect also likely explains why not all associations can be replicated in the large blood cohort.” (Page 17).

3.Figure 2 shows that splicing-related signals are under selective constraint in spliced elncrRNAs, which is convincing. But this does not prove that splicing of elncRNAs is directly related to enhancer activity. It is equally plausible that elncRNA expression directly impacts enhancer activity and that elncRNA splicing is conserved because it boost elncRNA expression, for example.

R:The reviewer is right and the sentence “If splicing of elncRNAs is important for enhancer function, …” does not faithfully describe the conclusions that can be drawn from the analysis reported in Figure 2. This portion of the text now reads: “If splicing of elncRNAs is functionally relevant, one would expect selection to have prevented the accumulation of deleterious mutations in their splicing-associated motifs during evolution” (Page 8). We would like to thank the reviewer for pointing this out.

Other points:

4.Are the ChiP profiles in Figures 1A-E significantly different from each other in a statistical sense? Probably yes, but a specific test should be done.

R:We now added boxplots representing the distribution of read density centered at transcript promoters. Statistical difference in the distribution is also tested. We show this in the revised Figure 1A-E and Supplementary Figure S1B-C.

5.This sentence (p. 10) was hard to follow and should be clarified: "As expected, the impact of SS variants on gene splicing efficiency depends on the total number of alternative transcripts and exons and ranges from 11% to 24% for elncRNA with 6 to 2 number of exons, respectively (Supplementary Figure S3)."

R:We had previously estimated the average amount of change in splicing for all alternative splicing events at each elncRNA candidate. To calculate this, we considered the difference in Percentage-Spliced-In (PSI) for all splicing events and divided this by the total number of considered events. Given that only a subset of events is affected by a splice site variant, the more exons an elncRNA has, the more alternative splicing events are likely to occur and the lower the average impact of a SS variant on overall gene splicing efficiency is expected to be. Following a comment from reviewer 1 (comment 1), we now only consider splicing events directly disrupted by the SS variant. We agree this sentence was not clear and have removed it from the manuscript.

6.Related to point 5 above, Supplementary Figure 3 is somewhat confusing because two splicing change and three expression change plots are shown for each locus, without labels of what each one is, or explanation of what the red and green colors mean.

R:We apologize for the confusion and thank the reviewer for pointing this out. In the figure, we plot the fold difference in elncRNA splicing, target gene splicing, target gene expression, non-target gene expression, and elncRNA expression. elncRNA features are plotted in red and target gene features are plotted in green. We have added labels to clarify the relevant plots (Figure 3D-H, Supplementary Figure S4,5).

**Minor points:**

1.Top of p. 16: "90% of those that support elncRNA splicing as a mediator 3 of target expression are located at the 5' end of the transcript, which is 4 consistent with the known synergy between transcription and 5' end splicing" - a reference is needed

R:We thank the reviewer for pointing this out and we have now added the appropriate reference.

“Importantly, 90% of seQTL associations that support elncRNA splicing as a mediator of target expression are associated with splicing junctions located at the 5´ end of the transcript, which is consistent with the known synergy between transcription and 5´ end splicing (Furger et al. 2002; Damgaard et al. 2008)(Figure 4D).” (Page 17)

2.Figure 5B,C y-axes indicate "fold difference", but scales include negative numbers, which is confusing. Probably should redo the analysis showing log of fold difference.

R:We thank the reviewer for the suggestion. Since the fold difference in Percentage-Spliced-In (PSI) used to estimate the amount of splicing at each exon junction can be of both positive and negative values, we now plot the log modulus transformation (John and Draper, 1980) of the data, which is equivalent to a log transformation while preserving the sign of the data. The analysis has been redone for Figure 3D-H, 5B,C, and Supplementary Figure S4, S5. This change does not impact the conclusions and makes the interpretation of the results more intuitive.

3.p. 20 Describes U1 snRNP as "a protein essential for the 4 recognition of nascent RNA 5' splice site and assembly of the spliceosome". U1 is a large RNA-protein complex, not a protein.

R:We thank the reviewer for pointing this out and this has now been corrected.

“Chromatin-bound lncRNAs have been recently shown to be enriched in U1 small nuclear ribonucleoprotein (snRNP) RNA-protein complex, a protein essential for the recognition of nascent RNA 5´ splice site and assembly of the spliceosome (Yin et al. 2020).” (Page 22)

4.Typo: p. 11, l. 2 missing word (genes): "expression levels of other nearby was unaffected"

R:This has been corrected.

Reviewer #2 (Significance (Required)):

The question addressed is very interesting, given recent work the significance of transcription from enhancers, and work addressing functional relationships between splicing and expression. The work is suggestive of effects of enhancer splicing on expression but I did not find it fully convincing as the effects observed are extremely small, and other explanations are not ruled out, as discussed above.

Prior literature has shown that many active enhancers are transcribed, that enhancer transcription can preced and is positively correlated with target gene expression, and work from both Ulitsky and from the authors indicates that splicing of enhancer-associated lncRNAs is positively correlated with enhancer activity. A variety of studies have also shown that splicing of protein-coding genes generally has a strong positive effect on gene expression. Here, the authors attempt to go further and show that splicing of enhancers causes increased transcriptional enhancement of target genes. A variety of public genotype, expression, chromatin and other types of data are analyzed to address this question. The statistical genetics crowd may find the work of interest, but molecular biologists will not be convinced of the conclusions. My expertise is in computational biology, genomics and RNA biology.

Engreitz JM, Haines JE, Perez EM, Munson G, Chen J, Kane M, McDonel PE, Guttman M, Lander ES. 2016. Local regulation of gene expression by lncRNA promoters, transcription and splicing. Nature 539: 452-455.

John, J., & Draper, N. (1980). An Alternative Family of Transformations. Journal of the Royal Statistical Society. Series C (Applied Statistics), 29(2), 190-197. doi:10.2307/2986305

Li YI, Knowles DA, Humphrey J, Barbeira AN, Dickinson SP, Im HK, Pritchard JK. 2018. Annotation-free quantification of RNA splicing using LeafCutter. Nature genetics 50: 151-158.

Yin Y, Yan P, Lu J, Song G, Zhu Y, Li Z, Zhao Y, Shen B, Huang X, Zhu H et al. 2015. Opposing Roles for the lncRNA Haunt and Its Genomic Locus in Regulating HOXA Gene Activation during Embryonic Stem Cell Differentiation. Cell stem cell 16: 504-516.