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 are grateful for the referees' rigorous review of our manuscript and for their overall positive reception of our work. We have pasted below the entirety of the reviewers’ comments, interleaved with our responses.

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

In this manuscript, Gama et al. use a biophysical assay DAmFRET, structural analysis, and optogenetic tools to uncover the nucleation mechanism of CBM signalosome. They performed experiments first in yeast cells that lack death folds or related signaling networks, then confirmed their discoveries in human cells. The results presented here are clear and convincing. The paper is very well presented and clearly written.

They found it is the CARD domain of BCL10 that acts as a molecular switch that drives all-or-none activation of NF-kB. Monomeric BCL10 possesses an unfavorable conformation and serves as a nucleation barrier, keeping BCL10 in a supersaturated inactive state that allows for binary activation upon stimulation.

They also characterized CARD9 CARD domain and a coiled-coil region. They reasoned that CARD9CARD functions as a polymer seed to nucleate BCL10, and that the coiled-coil region has multimerization ability to facilitate nucleation. Furthermore, they characterized that MALT1 activation doesn't depend on BCL10 polymers but its own proximity. And MALT1 induces graded NF-kB activation, thus further demonstrating the binary activation is conferred by BCL10.

Major comments:

1. Fig S1D and E, the authors used TNF-a to activate NF-kB independent of CBM signalosome and found the activation in each cell increased with dose. In contrast, CBM activation led to bimodal cell activation. The authors claim that this is evidence that positive feedback upstream of NF-kB. We do not believe this claim can be made from this comparative experiment alone. We agree that positive feedback is important for activating an NF-kB response, but the comparison between CBM and TNFa is inaccurate and glosses over published data. Specifically, there is published data that TNF-a does activate a 'switch-like' or digital response, as defined by the translocation of p65 (see (Tay et al. 2010) among other studies that have examined p65 translocation at the single-cell level). The difference in T-sapphire expression between CBM and TNF activation is most likely due to TNFa induced oscillations of p65 translocation (although this is speculation on our part). Therefore we suggest to the authors that the TNF-a data (Fig S1D and E) should be omitted, as the claim of switch or not-switch as pertains to TNF signaling is more complex and nuanced than presented here. We believe omitting this data will strengthen the manuscript and avoid confusion in the field. The bimodal expression of the T-sapphire NF-kB reporter driven by the CBM signalosome activation is sufficient to claim an all-or-none response.

We thank the reviewer for this suggestion. We acknowledge that the activation of NF-κB by TNF-ɑ is more complex than we had presented, and agree that the differences in T-Sapphire reporter output could be attributed to p65 oscillations. We had not previously considered this interesting possibility -- which is not addressed by the present data -- believe it is worth future investigation. As suggested by the reviewer, we have now omitted the TNF-a data, and agree that this change does not impact the overall claims of the paper.

Fig 3B, the authors introduced CARD9CARD-µNS as a stable condensed seed for BLC10. However, considering CARD9CARD can form polymers at high concentration (Fig 3B and S3D), are these high expression levels of CARD9CARD able to induce BCL10-mEos3.1 assembly (as measured by DamFRET in yeast cells)? Can the authors examine BCL10 FRET at these high expression level of CARD9CARD? We assume that BCL10 will be assembled in these cells. This would provide a valuable control experiment and support the author's conclusions.

Indeed, this question is amenable to DAmFRET. Accordingly, we have now performed DAmFRET of yeast cells expressing Bc10-mEos3.1 in the presence of either CARD9CARD-mCardinal or mCardinal itself (see new Fig S6A and B, and associated results section). We confirmed that cells with high CARD9CARD-mCardinal expression had higher FRET on average than cells with low expression. Importantly, cells expressing high or low levels of mCardinal itself had the same FRET level (Fig S6).

Fig 3C, the text said "Whereas WT CARD9CARD assembled into polymers at high concentration, the pathogenic mutants R18W, R35Q, R57H, and G72S failed to do so (Fig 3C and S7B,C), explaining why they cannot nucleate BCL10". This claim that these mutants can not nucleate BCL10 does not have a figure call out or a reference. The authors then show the results in Fig 3E which supports this claim. Even though they were done in the context of full-length CARD, all proteins contain the I107E mutation that releases autoinhibition. For clarity, the authors should consider rearranging the text to avoid explaining a phenomenon and making conclusions before showing the results.

We have now rearranged this section to match the figures and claims.

Fig 4D, E and Video 1, the authors showed the nucleation of BCL10 into puncta within live cells is followed by p65 translocation to the nucleus. The authors claim that 'this result suggests that BCL10 is indeed supersaturated prior to stimulation' (paragraph 2 section titled BCL10 is endogenously supersaturated'). We fail to understand how this live-cell experiment leads to the conclusion BCL10 is supersaturated before stimulation. We think this text should be deleted from the text, or put into context with the DAmFRET data that lead the authors to make this claim. It would be interesting for the authors to define in discussion what are the golden criteria to claim a protein exists in a supersaturated state with live cells (by microscopy or other methods)? Adaptor protein assembly into puncta and the subsequent nuclear translocation of transcription factors is a common phenomenon across signalling pathways. Not all these pathways rely on signaling adaptors existing in a supersaturated state. The field of cell signaling (and cell biology in general) would benefit from a detailed definition of how these physical-chemical definitions of proteins are supported by experimental data. We believe that this paper will become a seminal paper in the field, and future work will benefit from a clear definition of how a claim of supersaturation is derived from the data.

We appreciate that the concept of supersaturation will be foreign to many biologists, and welcome this opportunity to elaborate. We have now rephrased the corresponding results section for figure 4D, E, and have added new evidence to support our claim that BCL10 is supersaturated, as had been requested by reviewer 2 (see below in response to point 1). Supersaturation, as we (correctly) use the term, occurs when the concentration of a protein in solution exceeds its equilibrium solubility for the given conditions. The term is also sometimes used to describe __global __protein “concentrations” in excess of the solubility limit, even if a dense phase has already formed and potentially depleted the effective concentration (in solution) to the solubility limit. This is a key distinction, as only the former implies a high-energy out-of-equilibrium scenario that predetermines a future change -- release of the excess energy via phase separation.

How does one experimentally determine if a protein is supersaturated? In theory, one may conclude that a protein is supersaturated if its assembly causes a net loss of energy from the system (i.e. exothermic). Unfortunately, it is likely not yet possible to perform such measurements with sufficient sensitivity inside a living cell. However, it is possible to infer that a protein is supersaturated if assembly can be shown to occur without a net input of energy to the system, i.e. without any change in thermodynamic control parameters such as temperature, pH, post-translational modifications, concentration of the protein, or concentration of any interacting factor. To do this, one introduces a substoichiometric amount of pre-assembled protein to the system. This manipulation will trigger assembly if the protein is supersaturated. If the protein is instead subsaturated, assembly will not occur and the exogenously added assemblies will simply dissolve. This phenomenon, known as “seeding” in the prion field, is considered a golden criterion sufficient to conclude that a protein has prion behavior. However, because bona fide prions additionally require a means for dissemination between cells, seeding analyzed at the cellular rather than population level is more appropriately considered a sufficient criterion for supersaturation (which is a prerequisite for classical prion behavior (Khan et al. 2018)). Our CARD9CARD-Cry2 experiment was designed to test this criterion. Specifically, it allowed us to introduce a seed independently of receptor activation, thereby precluding any orthogonal cellular response that might lower Bcl10 solubility through e.g. a post-translational change. That the seeds were substoichiometric is evidenced by the fact that Bcl10 polymerized homotypically following stimulation (i.e. it didn’t just bind to the CARD9CARD puncta, but went on to deposit onto itself).

How does assembly under this scenario differ in principle from the many examples of puncta formed by other signaling proteins that occur upon stimulation of their respective pathways? Puncta formation that is induced by a thermodynamic change in the cell cannot be said to have resulted from pre-existing supersaturation. Rather, the stimulus may have caused some change that either increases the effective concentration of the protein (e.g. upregulates its expression, induces a post-translational change that activates it, or releases an inhibitory factor) or reduces solvent activity (e.g. change in pH).

An additional requirement (necessary but not sufficient) is that the assembly must be regular with respect to some order parameter. That is to say, it must be a bona fide “phase”. At a minimum, this implies a uniform density. Additionally, for supersaturation to persist over biological timescales under physiological conditions and confinement volumes, the assembly (once formed) must also have structural repetition in at least two dimensions, i.e. crystallinity (Rodríguez Gama et al. 2021; Zhang and Schmit 2016). We know this to be true for Bcl10.

Rodríguez Gama A, Miller T, Halfmann R. 2021. Mechanics of a molecular mousetrap-nucleation-limited innate immune signaling. Biophys J 120:1150–1160. doi:10.1016/j.bpj.2021.01.007

Khan, T., Kandola, T.S., Wu, J., Venkatesan, S., Ketter, E., Lange, J.J., Rodríguez Gama, A., Box, A., Unruh, J.R., Cook, M., et al. (2018). Quantifying nucleation in vivo reveals the physical basis of prion-like phase behavior. Mol. Cell 71, 155-168.e7.

Zhang L, Schmit JD. 2016. Pseudo-one-dimensional nucleation in dilute polymer solutions. Phys Rev E 93:060401. doi:10.1103/PhysRevE.93.060401

Regarding the supersaturated state of BCL10, the authors convincingly use optogenetics to show how transient assemblies of CARD-Cry2 can template BCL10 assembly. This is a convincing experiment that shows templated nucleation of BCL10. To strengthen the claim that BCL10 is supersaturated endogenously we suggest the author quantify the expression of BCL10-mScarlet and CARD-Cry2 and ideally show that this phenomenon can be observed at expression levels equivalent to endogenous.

As stated above, that BCL10-mScarlet formed polymers that we observed to elongate homotypically off of the CARD9CARD seeds indicates that the protein was supersaturated under the conditions of the experiment. The concentration of CARD9 is not a relevant parameter in this case. We had already compared the expression of BCL10-mScarlet to endogenous BCL10 in 293T, THP-1, and human fibroblast cells by quantitative immunodetection (Fig. S10D), revealing that the expression level of our BCL10-mScarlet constructs matched that of endogenous BCL10, which was approximately the same in all cell lines. We also compared the distribution of expression levels of BCL10-mScarlet versus that of endogenous BCL10 using antibody staining followed by flow cytometry, which confirmed that the range of expression levels of BCL10-mScarlet falls within that of endogenous BCL10 in 293T cells (Fig. S10F). Hence, we believe our data suffice to conclude that Bcl10 is supersaturated at endogenous levels of expression.

Minor comments:

1. Special character "delta" is not displayed in the text (instead only a space).

This error occurred upon exporting the manuscript from our text editor to a PDF. We now have made sure all special characters are present in the PDF version.

Several cell lines including mouse, human, and yeast lines were used across this manuscript. It would be clearer and more helpful if the exact cell type of the line could be indicated. Such as, "BCL10-mEos3.1 yeast cells" instead of "BCL10-mEos3.1 cells", "BCL10-mScarlet HEK293T cells" instead of "BCL10-mScarlet cells".

We have now modified all instances to indicate the origin of the cell lines tested.

Fig 5B, the authors indicated that BCL10 colocalized with CARD9CARD, then please show the merged image as well.

We have now included the merged image to indicate colocalization in the inset images.

Fig 6E, authors claimed that cells were stimulated with blue light for the indicated durations. The longest duration is 12 hours. Please specify if it was continuous exposure or several rounds of exposure in the indicated durations.

We have now specified in the figure legends, text, and methods section, that this specific experiment used a continuous exposure of blue light.

Reviewer #1 (Significance (Required)):

This work used a combination of FRET and optogenetic tools to engineer CBM signaling and visualize the effects. They incorporated knowledge from structure biology, together with their results from mutations and truncations, dissected the significance of each protein in CBM signalosome, and demonstrated in detail how higher-order assemblies make all-or-none cellular decisions. We believe this paper will be a seminal paper in the field of cell signalling and cytoplasmic organization. It defines a new paradigm of macromolecules assembly of signalling complexes as being dependent on protein existing in a supersaturated state. Importantly this paper opens up new questions regarding macromolecular signaling complexes (found in many innate immune signaling pathways): How is protein supersaturation maintained and used throughout evolution to construct biochemical signalling switches?

This paper will be of particular interest to scientists working on immunity and cell signalling, especially in the field of higher-order assemblies. However, we feel the impact of this paper goes beyond these fields, and we believe this manuscript will be of broad interest to the cell biology and biophysics communities. For reference, our expertise is in innate immunity and cell biology.

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

In their manuscript entitled "A nucleation barrier springloads..." Rodriguez-Gama et al. dissect the assembly mechanism of the signalosome, composed of the proteins CARD9, BCL10 and MALT1, using a novel in-cell biophysical approach (DAmFRET). They first overexpressed fluorescently tagged versions of the proteins to promote their assembly in yeast and mammalian cells, finding that CARD9 forms higher order assemblies across a wide range of concentrations with no discontinuity in the DAmFRET profile. In contrast, the DAmFRET profile of BCL10 showed a clear separation between monomers and higher order assemblies, which started to form spontaneously only at higher BCL10 concentrations. Furthermore, at the two states of the proteins co-exist at all concentrations. These observations imply that there is a nucleation barrier to forming BCL10 assemblies. MALT1 showed no change in FRET regardless of its expression level. These observations, alongside fluorescence microscopy of the assemblies, and previous structural studies, suggest that BCL10 forms self-templating polymers that act as a switch for an all-or-nothing immune response, assayed in this case by monitoring the nuclear translocation of the NF-kB subunit p65. The authors also assessed the effects of known disease-causing mutations on the nucleation barrier, showing that changes in the strength of the nucleation barrier can have major effects on signalosome function. Finally, they used optogenetic methods to trigger assembly of individual signalosome components, providing insight into the minimal components/conditions required for signalosomes to work.

Major comments

Overall, the experiments by Rodriguez-Gama et al. offer convincing evidence that there is a nucleation barrier to BCL10 polymerisation, and that a CARD9 template is sufficient to overcome the barrier. Although the existence of a nucleation barrier had already been postulated, based on structural and other studies (referenced by the authors), it had lacked a rigorous demonstration. This work provides that demonstration, which is important for the signalosome field and more broadly applicable to researchers studying cellular decision making. The study further demonstrates that DaMFRET is an excellent to study protein assembly processes in their native environment, allowing the authors to tackle a question that would have been technically very difficult to address otherwise. The optogenetic experiments are a nice sufficiency test for their ideas.

We feel there are a few key points to address before publication.

1) One of the main conclusions is that spring-loading the nucleation barrier with high super-saturating BCL10 concentrations allows a decisive response. Although much of the data strongly imply this conclusion, the dependence of the immune response on BCL10 concentration was not tested directly. A key prediction of the nucleation barrier is that at concentrations below saturation, BCL10 should not be able to induce an all-or-nothing response when stimulated. At saturated/super-saturated concentrations BCL10 should be able to induce a response. At deeply super-saturated concentrations the response should start to be activated spontaneously in the absence of an external stimulus. These predictions could be tested using the doxycycline-inducible BCL10 system (Figure S2D), without establishing major new experimental avenues. We feel that such an experiment would strengthen the main conclusion. It might also help to shed light on whether being highly supersaturated enables a more decisive response than being just saturated.

This is a great idea. As the reviewer suggested, our Doxycycline-inducible BCL10 system enables us to induce and track the state of BCL10 over time. We have now performed the requested experiments (Fig. S9D, E) and incorporated the results into the relevant section of the text. In short, our new analyses show that BCL10 indeed has a concentration threshold for activation by stimulation, and that it can also nucleate spontaneously when overexpressed. Note that our original analyses in Fig. 4B and C also demonstrate spontaneous BCL10 activation at high concentrations. With this new evidence and the orthogonal approaches used in Fig. 5, we believe our data definitively support our conclusion that BCL10 is supersaturated.

2) Intuitively, readers might expect that if BCL10 is supersaturated then, once nucleated, it would rapidly assemble at the nucleation sites. In Figure 5B, CARD9CARD-miRFP670nano-Cry2 assemblies are optically induced throughout the cell. However, BCL10 appears to nucleate at just a few sites with a few minutes delay. More widespread nucleation and growth of BCL10 polymers seems to take longer (20-40 minutes, Figures 5B and 5C), after CARD9CARD-miRFP670nano-Cry2 has disassembled. Furthermore, in Figures 4D and 4E, very few BCL10 assemblies are visible/quantifiable after 70 minutes PMA exposure, but p65 has clearly entered the nucleus. It looks like BCL10 assembly slightly lags behind p65 nuclear entry. Can the authors provide a more detailed explanation of these kinetics?

We do note that the number of CARD9CARD clusters formed upon opto-stimulation exceeds the apparent number of BCL10 nucleation sites. We believe this is consistent with nucleation-limited kinetics, where the clustering of CARD9-CARD increases the local probability of nucleation. As nuclei form and grow, they lower the probability of subsequent nucleation elsewhere in the cell. Additionally, it is possible that our artificial seeds do not perfectly mimic the native CARD9 seeds that form upon natural stimulation (e.g. due to potential steric interference from the fluorophore and Cry2). We also acknowledge that there is a slight delay in the visible appearance of BCL10 polymers relative to p65 nuclear translocation. We expect that MALT1 activates already when the polymers are still too small to see (sub-resolution), whereas the polymers only become microscopically visible once they’ve grown quite a bit more.

3) Related to point 2 above, in Figure 5D, the leftmost cell in the field of view clearly contains CARD9CARD assemblies but there are no BCL10 assemblies and p65 is not imported into the nucleus (in contrast to the central cell in the field of view). How often does CARD9CARD optogenetic assembly lead to BCL10 assembly? In other words, can the authors quantify the cell-to-cell variability in this experiment?

Throughout our experiments, whether analyzing BCL10 puncta formation, NF-kB transcriptional activity, or p65 translocation, we observed a persistent nonresponsive fraction of cells even at saturating levels of stimulation. Specifically, approximately 30% of THP-1 cells failed to acquire T-Sapphire fluorescence or form BCL10-mEos3.2 puncta when stimulated with high levels of β-glucan (Fig 1B and E, respectively), and approximately 25% of 293T cells failed to acquire T-Sapphire fluorescence or exhibit p65 nuclear translocation when stimulated with high levels of PMA (Fig 1C and Fig 4E, respectively). Because these numbers did not depend on whether BCL10 was endogenously or exogenously expressed, we know that the underlying cell-to-cell heterogeneity involves factors upstream of BCL10. Indeed, the fraction of recalcitrant cells drops to 10% in our optogenetic experiments that bypass upstream factors (Fig S11E). Possible sources of heterogeneity include different physiological states of the cells or fluctuations in the expression levels of any upstream factor in the signaling pathway. We believe that this phenomenon is not unique to the CBM signalosome, as we (unpublished) and others (Fernandes-Alnemri T et al, 2009, Dick M et al, 2016) have similarly observed a fraction of non-responding cells upon activation of the inflammasome, which involves nucleation-limited polymerization of the adaptor protein ASC. While this phenomenon is interesting and may be important to our understanding of the full complexity of signalosomes in vivo, we believe that identifying the source of heterogeneity would be outside the scope of the present manuscript. We now describe this phenomenon in the final paragraph of the “Endogenous BCL10 is constitutively supersaturated” section.

Fernandes-Alnemri, T., Yu, JW., Datta, P. et al. AIM2 activates the inflammasome and cell death in response to cytoplasmic DNA. Nature 458, 509–513 (2009). https://doi.org/10.1038/nature07710

Dick, M., Sborgi, L., Rühl, S. et al. ASC filament formation serves as a signal amplification mechanism for inflammasomes. Nat Commun 7, 11929 (2016). https://doi.org/10.1038/ncomms11929

Minor comments

While the work is scientifically well done, the text reads as though it is meant for experts rather than a broad audience. This is a pity because it risks alienating readers. We suggest that some adjustments to the text (mainly additional explanations and not ruling out alternative interpretations of the data) would widen the audience and increase the impact of this important study. Below are some suggestions that might help.

1. In the first results section, the authors write: 'This suggests that Bcl10 but not CARD9 assembly occurs in a highly cooperative fashion that could, in principle (Koch, 2020), underlie the feed forward mechanism.' It isn't obvious how Figure 1 leads to this statement. Could the authors give a more detailed explanation?

We have now revised the text to elaborate on this interpretation.

One limitation of DAmFRET is that it can only detect a nucleation barrier where there is a difference in FRET between the monomer and the assembled form of the protein. However, it can't necessarily detect when there is not a nucleation barrier i.e. if there's no difference in FRET. The text seems to suggest that CARD9 and MALT1 don't have nucleation barriers to their assembly. While this might not be intentional, it would be helpful to explicitly state that CARD9 and MALT1 could also possess such barriers that are not detectable by this method. This wouldn't detract from the finding that BCL10 has a barrier that plays an important function.

The reviewer is correct that DAmFRET would not be able to detect a nucleation barrier if the assembled phase does not condense the fluorophore to a sufficiently high density for FRET to occur. In our experience, this is only a concern for very large proteins whose bulk “dilutes” the fluorophores within the assembly. Death domains, on the other hand, are only ~ 3 nm in diameter, and FRET occurs within a range of ~10 nm; hence we think it very unlikely that the death domains could be forming cryptic polymers that escape our detection. In any case, when assembly does produce a change in FRET, we can with confidence determine how strongly that form of assembly is governed by concentration. Hence, for CARD9, which does produce a FRET signal upon assembly, we can say that assembly has a smaller intrinsic nucleation barrier than that of BCL10. We further eliminated the possibility of multi-step nucleation (which would reduce the apparent nucleation barrier relative to the one-step ideal case) for CARD9 by showing that artificial condensates of the protein expressed in trans do not influence the concentration-dependence of FRET (Fig. 4 B). Finally, under all conditions where CARD9 lacked FRET, it also lacked signaling activity, suggesting there is not a cryptic functional assembly that evades our assay. Likewise MALT1, which lacked FRET at all concentrations, was entirely unable to activate NF-kB upon overexpression (Fig. S8 A and B), suggesting that it too is not forming a cryptic functional assembly that evades our assay. We therefore feel confident in our conclusion that CARD9 and MALT1 lack nucleation barriers of a magnitude comparable to that of BCL10. Note that our claim is not that they entirely lack a nucleation barrier (CARD9 after all does form a multi-dimensionally ordered polymer), but rather that we fail to observe a nucleation barrier and hence any barrier that may exist is insufficient to manifest at the cellular level.

In the final results section, the idea that MALT1 activation doesn't depend on BCL10 polymer structure doesn't necessarily follow from the data. An alternative interpretation is that optogenetic clustering of MALT1 causes it to recruit BCL10 and form BCL10-MALT1 filaments (structure solved by Schlauderer et al., 2018). Also, the optogenetic clustering of MALT1 may mimic some structure found in the BCL10 cluster. Therefore, we are neither convinced that the data unambiguously show that MALT1 activation strictly depends on multi-valency rather than an ordered structure of BCL10 polymers nor that this conclusion is truly necessary for the paper.

We agree that the reviewer’s alternative interpretation of this experiment is possible. However, we consider it unlikely because we performed the experiment with MALT1 lacking its Death Domain (residues 126-824), which mediates its interaction with BCL10 (Schlauderer et al., 2018). Our experiments then suggest that MALT1 clustering is sufficient for activation independent of any structuring mediated by BCL10. Nevertheless, we have now performed an additional control in which we treated these cells with PMA to induce BCL10 polymerization. As expected, the NF-kB transcriptional reporter utterly failed to activate in this condition, indicating that MALT1 does not interact with BCL10 polymers when it lacks its death domain. This aspect has been further elaborated in our response to reviewer 3 point 5.

What optical density do the yeast cells reach during the 16h induction in galactose? If they are in stationary phase, this could affect the assembly status of the proteins being expressed, as the cytoplasm becomes glassy when cells are starved, and this coincides with widespread protein aggregation/assembly (Joyner et al., 2016; Munder et al., 2016).

In our DAmFRET strategy, we first dilute an overnight culture and regrow the cells to log phase prior to resuspending them in galactose media. Our strain is engineered to undergo cell cycle arrest upon protein induction, hence exponential growth is prevented and the cells do not deplete galactose during the 16 hr induction. We have also performed many time courses of DAmFRET following induction and generally find no qualitative difference between early and late times (unpublished). Early time points simply have lower expression and correspondingly fewer cells in the high FRET state. Importantly, all comparisons between proteins are made with the same 16 hr induction.

Although these experiments show that thermodynamically lowering the BCL10 nucleation barrier (e.g. by post-translational modifications or protein expression levels) isn't required for a response, they don't rule it out. It would be good to state this in the discussion, as cells may have multiple mechanisms of switching on the signalosome.

We thank the reviewer for this suggestion and have now explicitly stated in the discussion that our experiments do not argue against possible thermodynamic tuning of the nucleation barrier.

The discussion compares signalosomes with condensates formed by liquid-liquid phase separation. This is an interesting comparison but it suggests that disordered assemblies would not be capable of performing signalosome-like functions. This needs to be explained more clearly. For example, non-amyloid prions seem to form gel-like assemblies with a high nucleation barrier that are capable of driving heritable traits, likely through self-templating (Chakravarty et al., 2020). Such examples could represent disordered assemblies with signalosome switch-like behaviour. Furthermore, there are examples of condensates that are induced by environmental changes e.g. Pab1 and Ded1 condensates (Riback et al., 2017; Iserman et al., 2020). This potentially allows the proteins to reach high concentrations and remain un-condensed until a change in heat or pH overcomes a nucleation barrier required for condensate formation. Although the condensates aren't self-templating, they seem to require energy for their disassembly. Combined, this also allows switch-like behaviour, where the switch is flipped back to the uncondensed off state once conditions return to normal. In general, crossing a phase boundary can represent a switch-like response. Finally, recent electron-tomography experiments show that ASC puncta comprise clusters of filaments (Liu et al., 2021, biorxiv). CARD9/BCL10 assemblies may have similar ultrastructures and liquid-liquid phase separation may well play a role in their assembly.

Indeed, we explicitly maintain that liquid phases cannot themselves perform signalosome-like functions. Chakravarty et al. 2020 did not observe amyloids associated with their phenomena, but the relevant experiments were not designed to exhaustively exclude an underlying ordered phase. To the extent that gelation is involved, their observations are fully consistent with ours. IUPAC defines a “gel” as a colloidal network involving a solid phase and a dispersed phase. The existence of a solid phase necessarily implies an underlying disorder-to-order transition, even if limited to small length scales. In the case of gelation associated with liquid-liquid phase separation, nucleation of the ordered phase simply occurs in two steps (first condensation, then ordering). Note also that a liquid phase could in principle give rise to a heritable phenotype if it activates a positive feedback in a molecular biological process involving the protein of interest (e.g. upregulation of its expression or a change in interacting factors). Chakravarty et al. did not exclude such phenomena (it would be very difficult to do so); hence it cannot be concluded that phase separation is responsible for the sustained phenotypic changes.

We do not fully follow the reviewer’s logic concerning the relevance of Pab1 and Ded1 condensates. These proteins only condense when their respective phase boundaries fall below the endogenous protein concentration, as upon thermal stress. The proteins are not supersaturated in the absence of such conditions (for example, they cannot be seeded), and it is incorrect to characterize the change in heat or pH as overcoming a pre-existing nucleation barrier. The concept of a nucleation barrier only applies under conditions where a phase is thermodynamically favored. It is also misleading to state that the Ded1 and Pab1 condensates require energy for disassembly. Rather, they require energy to disassemble rapidly. Unless the assemblies have accessed a more ordered phase as described above (two step nucleation), involving a lower phase boundary, they will inevitably dissolve after the conditions return to normal.

We have much prior experience with ASC. Although it has not been explicitly shown, that it forms ordered polymers and can behave as a prionoid in vivo suggests that it very likely operates the same way as BCL10 (i.e. is physiologically supersaturated). That full-length ASC forms clusters of filaments is not relevant (in our view) to the mechanism shown here, which only requires that filaments are indeed formed. Formally, the size of the relevant nucleus determines the minimum length scale at which ordering must manifest in our mechanism. Based on the structure of death domain filaments, this could be as small as tetramers or hexamers (a minimal but structurally complete “polymer”).

As stated above, and now elaborated in the discussion, our data do not exclude a role of thermodynamic regulation, as could lead to liquid-liquid phase separation, in tuning the nucleation barrier of Bcl10. What they do exclude is that such changes are required for Bcl10 to activate in the first place.

Can the authors comment on the loss of BCL10 in Echinodermata, Anthropoda, Nematoda? Is there another protein that plays a similar role? Could a CARD or PCASP protein possess self-templating properties? Could other methods of control be at play e.g. protein expression?

This is a very interesting question! We think the reviewer’s suggested explanations for the loss of BCL10 in those lineages are valid and worthy of future exploration. Nematodes such as C. elegans have lost multiple components of innate immunity. They have very few pathogen recognition receptors and also lack NF-kB! They do, however, have other adaptor proteins that the literature and our unpublished data suggest may have self-templating ability, such as TIR-1. Drosophila also encodes multiple TIR-containing proteins that are essential for innate immunity. In short, it is possible that other proteins have acquired the hypothetically essential role of supersaturation and nucleation-limited signaling in these organisms.

Figures 1B/1C: Can the authors comment on why the active cells plateau at about 70-75%? This is a striking feature of the plots, but the explanation may not be obvious to readers.

See our response to major point 3, above.

Figures 1D/1E: What was the concentration of B-glucan used in this experiment? This could be included in the figure legend. If greater than 1ug/ml this means that the % of active cells in Figure 1B matches the % of cells with BCL10 assemblies in Figures 1D/1E, which is potentially an important point.

We thank the reviewer for bringing this point to our attention. We have now indicated in the figure legend the concentration of B-glucan used in this experiment (10 μg/ml). That the percentage of active cells in Fig. 1B matches that of cells containing BCL10 polymers in Fig. 1D and E indeed strengthens the stated relationship between BCL10 assembly and NF-kB activation in THP-1 cells subjected to a relatively physiological stimulus. Additionally, we have performed experiments to measure the levels of p65 translocation in THP-1 cells treated with B-glucan that express BCL10-mEos3.2. This data is shown in Figs. S1D and E in response to reviewer 3.

Use of both 'BCL10' and 'Bcl10' when referring to the protein.

We have now replaced all instances where Bcl10 was used to follow guidelines for gene and protein name conventions.

Bruford EA, Braschi B, Denny P, Jones TEM, Seal RL, Tweedie S. Guidelines for human gene nomenclature. Nat Genet. 2020;52(8):754-758. doi:10.1038/s41588-020-0669-3

In the supplementary figures there are some formatting problems/missing words in the figure legends. In Figure S11 there is a black box covering the lower part of the figure.

We have now fixed these instances.

References used in this review

Chakravarty, A.K. et al. (2020) "A Non-amyloid Prion Particle that Activates a Heritable Gene Expression Program," Molecular Cell, 77(2), pp. 251-265.e9. doi:10.1016/j.molcel.2019.10.028.

Iserman, C. et al. (2020) "Condensation of Ded1p Promotes a Translational Switch from Housekeeping to Stress Protein Production," Cell, 181, pp. 818-831.e19. doi:10.1016/j.cell.2020.04.009.

Joyner, R.P. et al. (2016) "A glucose-starvation response regulates the diffusion of macromolecules," eLife, 5. doi:10.7554/eLife.09376.

Munder, M.C. et al. (2016) "A pH-driven transition of the cytoplasm from a fluid- to a solid-like state promotes entry into dormancy," eLife, 5(MARCH2016). doi:10.7554/ELIFE.09347.

Riback, J.A. et al. (2017) "Stress-Triggered Phase Separation Is an Adaptive, Evolutionarily Tuned Response," Cell, 168(6), pp. 1028-1040.e19. doi:10.1016/j.cell.2017.02.027.

Schlauderer, F. et al. (2018) "Molecular architecture and regulation of BCL10-MALT1 filaments," Nature Communications 2018 9:1, 9(1), pp. 1-12. doi:10.1038/s41467-018-06573-8.

Reviewer #2 (Significance (Required)):

The existence of a nucleation barrier had already been postulated, based on structural and other studies (referenced by the authors), it had lacked a rigorous demonstration. This work provides that demonstration, which is important for the signalosome field and more broadly applicable to researchers studying cellular decision making. The study further demonstrates that DaMFRET is an excellent to study protein assembly processes in their native environment, allowing the authors to tackle a question that would have been technically very difficult to address otherwise.

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

The study by Rodriguez Gama et al. addresses the molecular function of CBM complex-forming proteins CARD9, BCL10 and MALT1 in the activation of myeloid cells, using optogenetic tools, transcriptional reporters and biochemical approaches. It is known from previous studies that Bcl10 oligomerizes into filamentous oligomeric structures incorporating Malt1, and that these structures are nucleated by receptor-induced activation of CARD proteins such as CARD11 (in lymphocytes) or CARD9 (in myeloid cells), but the mechanism underlying the assembly of the resulting CBM complexes remain incompletely understood.

The authors develop beautiful optogenetic tools to address this question, and convincingly demonstrate that CARD9-mediated nucleation of BCL10 triggers a binary cellular NF-kB response in a spring-load-like fashion, and identify mutants of BCL10 and CARD9 that impact this capacity. Unfortunately, however, the authors do not do a good job to simplify this complex problem so it can be easily understood. In particular, the choices of mutants, models and experiments are not consistent between figures, and some data seem to be arbitrarily added or omitted. Complex hybrid constructs are also used, without assessing whether these are indeed functional in the corresponding ko cells. The paper would therefore benefit from a major overhaul. We also noticed that the literature is often not cited adequately and have included a (non-exhaustive) list of examples of wrong, incomplete, or erroneous citations below.

1. The initial observations of binary signaling are derived from a reporter system. Although there are controls to show that the reporter used does not function intrinsically cooperatively, it would be nice to see additional data to show that cooperativity occurs also at the level of endogenous response systems, for instance by qPCR-based assessment of a natural NF-kB target gene (induced for example by TNFa versus B-glucan in THP-1 cells, and by TNFa versus PMA in 293T cells).

As detailed in the introduction, NF-kB has been shown by multiple labs to activate in a binary fashion. Our manuscript shows that NF-kB activation occurs in a binary fashion both at the level of transcription and at the level of nuclear translocation (upstream of any transcriptional output). While we do agree that additional data could further illustrate the biological significance of our findings, we do not feel it is necessary for our conclusions. Note also that because NF-kB activation occurs in a binary fashion per cell, a simple qPCR experiment would not suffice to extend our findings to the broader Nf-kB regulon. Instead, one would have to use e.g. RNA-FISH or single cell RNA-seq, nontrivial experiments that would take months to complete.

The cell lines in Figures 1D-E (and also some of the BCL10 mutants used later on) would have been better run in the assays in the early parts of Figure 1. The final conclusion prior to the section The adaptor protein BCL10 is a nucleation-mediated switch is otherwise not justified. This is a central tenet of the paper, that is referred to again, with some other ancillary data to support it. These mutants reappear later in the paper, but it would have been better, and easier to make rescue lines of BCL10 KO in Figure 1, otherwise the logic is lost, and the models seem chosen arbitrarily.

The choice of experiments in different panels of Fig. 1 resulted from a chronological progression of reagent construction as the project evolved. We do appreciate that switching between the assays may lead readers to doubt one or the other. Therefore, we have now immunostained for endogenous p65 in the same experiment as for Fig. 1D and confirmed that p65 translocated to the nucleus only in THP-1 BCL10-KO cells that have been reconstituted with WT BCL10-mEos3.2, but not E53R. We think this additional evidence along with our orthogonal measurements in other reporter systems confirms our findings that BCL10 nucleation determines NF-kB activity.

Expression with microNS is not well controlled and gives little real evidence for what is occurring. It is unclear what the concentration of the protein expressed was, but certainly the relative expression of the CARD9(CARD) and the microNS version should be assessed.

We believe these concerns result from a misunderstanding. We assume the reviewer is referring to the experiment in Fig. 3B. Expression of muNS on its own has no effect on the DAmFRET of other proteins, and we have previously used it in exactly the same way as here (Holliday M et al. 2019 and Kandola T et al. 2021). Please note that muNS fusion proteins in our experiment have an orthogonal fluorescent protein whose spectra do not significantly overlap with those of mEos3.1. The experiment evaluates a protein’s ability, when condensed via its fusion to muNS, to nucleate an mEos3.1-fused protein that is expressed in trans. Fusion of proteins to muNS does not affect their expression levels, as we now show for CARD9CARD-muNS-mCardinal versus CARD9CARD-mCardinal (Fig. S6D).

Also, the AmFRET profile of CARD9CARD looks very weird, it cannot be compared to BCL10.

We are unsure in what way the AmFRET profile of CARD9CARD is “weird”. It is fully consistent with expectations and has been thoroughly explained in the text. We suspect the reviewer was bothered by the sharp acquisition of FRET at approximately 100 uM. As explained in the text, this represents the phase boundary, also known as the solubility line, for CARD9CARD polymers, which we previously showed in vitro (Holliday M et al. 2019). Above this concentration, the protein self-assembles without a nucleation barrier, hence the sharp but continuous change in FRET. BCL10 plots, in contrast, show a discontinuous acquisition of FRET, which indicates a nucleation barrier. In order to highlight that the CARD9CARD transition is understood and expected, we have also now added a line to the plot to demarcate the phase boundary.

We are not convinced of the usefulness of the introduction of a slew of disease-causing CARD9 mutations that may or may not be relevant to the authors' point. The fact that they do or do not function in a specific sub portion of an assay that may or may not be relevant to biological activity seems to be of interest but without biochemical understanding, little is clear.

While several reports have shown the clinical importance of these CARD9 mutations on susceptibility to fungal infections, little was known about the molecular mechanism underlying their effects. The inclusion of the disease-causing mutants to this paper is justified for the following reasons. First, they demonstrate the relevance of our work to disease. Second, they build off our findings to provide an otherwise unknown molecular mechanism of these mutants. We showed using independent methods that CARD9CARD mutations disrupt the ability to nucleate BCL10, via two different mechanisms. Finally, validating the disease-causing mutations allowed us to use them as controls for subsequent experiments demonstrating that BCL10 is supersaturated.

The Optogenetic experiments are interesting, but difficult to interpret without evidence that these MALT1 constructs are indeed still functional when expressed in MALT1-deficient THP-1 cells. We do not therefore think that this experiment shows a necessity for clustering to signal, just a sufficiency, and in a highly artificial construct.

We welcome the opportunity to elaborate on the optogenetic experiments. Since BCL10 and MALT1 are expressed ubiquitously across cell types, the validity of our findings should not depend on the cell type used. Indeed, much of what we already know about innate immunity signalosomes comes from work in HEK293T cells. Our optogenetic experiments using MALT1 were performed in 293T MALT1-KO cells in Figures 6E and F, and employed two distinct functional assays (p65 nuclear translocation and a transcriptional reporter). While our approach employs light to control clustering, similar approaches using (no less-artificial) chemically induced dimerization domains have been used to study caspase activation (Oberst A et al, 2010, Boucher D et al, 2018). Our use of light affords higher specificity, reversibility, and spatial and temporal control over MALT1 assembly than does chemically induced dimerization.

To demonstrate the necessity of clustering, we have now performed an experiment with MALT1(126-824)-miRFP670-Cry2 expressed in 293T MALT1 KO cells that contain a transcriptional reporter of NF-kB ,as in figures 6E and F. We added PMA to the cells and found that it failed to activate NF-kB (Fig. 6), confirming that the interaction of MALT1 (via its death domain) with polymerized BCL10 is required for activation. Note that MALT1 and BCL10 exist as a soluble heterodimer prior to BCL10 polymerization; hence it is polymerization, rather than the interaction itself, that activates MALT1. That artificial clustering rescues this defect strongly suggests that the effect of polymerization can be attributed to increased proximity rather than some allosteric effect communicated from BCL10 polymers through the MALT1 DD to its caspase-like domain.

Oberst, A., Pop, C., Tremblay, A.G., Blais, V., Denault, J.-B., Salvesen, G.S., and Green, D.R. (2010). Inducible dimerization and inducible cleavage reveal a requirement for both processes in caspase-8 activation. J. Biol. Chem. 285, 16632–16642.

Boucher, D., Monteleone, M., Coll, R.C., Chen, K.W., Ross, C.M., Teo, J.L., Gomez, G.A., Holley, C.L., Bierschenk, D., Stacey, K.J., et al. (2018). Caspase-1 self-cleavage is an intrinsic mechanism to terminate inflammasome activity. J. Exp. Med. 215, 827–840.

In the introduction and other parts of the paper, there are numerous instances where the previous literature in the field is not adequately cited. Examples include:

• In the introduction, it is weird to cite one original paper (a MALT1 ko study by Ruland et al., 2001; there are several other studies of ko papers for CBM components that would merit being citated along with this study) together with two reviews on that topic (Ruland and Hartjes 2019 and Gehring et al. 2018)

• In the introduction, the original study by Wang et al., 2002 should be cited together with Rebeaud et al., 2002; the two studies on the same topic were published back-to-back

• In the introduction, the statement "CARD10 and CARD14 are expressed in nonhematopoietic cells including intestinal and skin epithelia, respectively" should be supported by citations.

• Still in the introduction, the 2 references for the statement "... CARD14 gain of function mutations cause psoriasis (Howes et al., 2016; Jordan et al., 2012)" are not appropriate. There are several reports of patients with CARD14 mutations (the study by Jordan et al is only one of them) and several CARD14 mouse models that provoke a psoriasis-like phenotype, which would merit being cited.

• In the following sentence: "Point mutations and translocations involving BCL10 and MALT1 cause immunodeficiencies (Ruland and Hartjes, 2019), testicular cancer (Kuper-Hommel et al., 2013), and lymphomas (Zhang et al., 1999).", the citation style also seems completely random, combining the citation of a single original paper for lymphomas (Zhang et al. 1999) (there are several other important original studies on that topic or recent reviews that could be cited instead), together with a review on immunodeficiencies (Ruland and Hartjes, 2019) and then another single example for a role of BCL10 and MALT1 in carcinoma (the study by Kuper-Hommel et al. is one, but several other original publications exist on the latter topic, showing for example a role in breast carcinoma or glioblastoma).

• In the first section of the results, the reference cited for endogenous CARD10 expression in 293T cells (Ruland et al., 2001) is wrong, no endogenous CARD10 expression was assessed in that study

We have now revised the citations mentioned above and other instances to ensure adequate citations in each case.

Reviewer #3 (Significance (Required)):

The paper deals with a complex question, namely how the CBM signalosome assembles and functions to stimulate NF-kB signaling. This question is important to the understanding of pro-inflammatory immune responses and basic life sciences in general. As the focal point of the paper is complex, and tools to study such phenomena are at the limit of technical capabilities, this further increases the potential impact of the work.

Reviewer #4 (Evidence, reproducibility and clarity (Required)):

The characterization of open-ended signalosomes in a number of innate-immunity and cell-death pathways, in particular formed by domains from the death-fold family, has led to the suggestions that these complexes allow a switch-like signalling response suitable for these pathways. It appears that this has been widely accepted. However, these suggestions are based largely on indirect observations and speculation.

Rodriguez-Gama and coworkers have decided to test these suggestions more directly. Their results confirm the suggestions. Based on my own experience, papers that validate widely adopted suggestions are often not considered seriously by top journals, who are looking for hot topics/paradigm-changing/surprising type results. I would urge the editors to consider seriously work such as in this paper, which directly tests important suggestions and does so at a technically high standard. The authors use a range of ingenious approaches, both with recombinant proteins and in cells, and including proteins from organisms from different parts of the evolutionary tree, to support their interpretations, so it is an extensive and high-quality study. I am impressed that so many different fusion proteins with fluorescent tags continued to function as expected, but I guess the authors controlled for this as much as they could.

Having said all this, I do get the feeling the authors are "over-selling" the nucleation barrier aspect of these signalling mechanisms. It is clearly an important and critical aspect of signalling in many systems, but then it is not the only important aspect; a number of other regulatory inputs play a role in different systems. So the statement "Our findings introduce a novel structure-function paradigm" in my view is overstretching things somewhat. Further in the Discussion section, the authors state "Existing explanations for the preponderance of ordered polymers in immune cell signalosomes have centered on the functions of multivalency at steady state, such as scaffolding and sensitivity enhancement resulting from the cooperativity of homo-oligomerization". They cite a small (and non-exhaustive) number of papers discussing this topic; all these include "seeding" or "nucleation" as an important part of the proposed mechanism. So I suggest the authors provide a more balanced discussion of this aspect. Different pathways appear to display a different level of switch-like behaviour, and one thing that the current version of the manuscript is missing is more discussion of other death fold-based systems and how the results on the CBM signalosome apply to these, and also other systems such as TIR domain-based ones, which currently get no mention whatsoever. In the CBM system, there seems to be one main nucleation barrier; can there be more than one in others?

We appreciate the reviewer’s perspective and have now acknowledged in the introduction and discussion additional prior literature that has paved the way for our study. Nevertheless, we maintain -- as now stated in the abstract -- that “our results defy the usual protein structure/function paradigm, and demonstrate that protein structure can evolve via selection for energetic maxima in addition to minima”. We have elaborated in the introduction and discussion how immune signaling provides the functional context in which such a paradigm can evolve, and how our findings uniquely support the paradigm.

One other aspect I need to express some criticism about is attention to detail - especially with a paper focusing on the physics behind biological processes, I would expect a higher standard of getting the terminology and units correct - see specific examples below. This can obviously be fixed easily.

Specific points are listed below. No page or line numbers are provided so I have done my best to make it clear what the comments refer to.

1. Abstract line 6 and throughout: in "NF-kB", the "k" is supposed to be "kappa" (Greek letter) - it stands for "nuclear factor kappa-light-chain-enhancer of activated B cells", not fully defined in the manuscript as far as I can see. Occasionally, small k is also used instead of the small cap K or whatever the authors used most of the time, but I don't think any of them use the Greek letter.

We had indeed used a version of the small “kappa” κ. We have now fixed the cases where we mistakenly used k instead of κ.

Page 2 (Introduction) paragraph 2 line 9: period missing at the end of sentence. Same Page 4 (Results: Assembly) paragraph 4 line 3.

This is now fixed.

Page 2 (Introduction) paragraph 2 line 15 and throughout: in long sentences, more commas can help help readability, for example before "leading" here. Similar page 15 paragraph 2 line 3 after "Additionally", paragraph 4 line 2 before "which".

We have now included more commas and tried to improve readability throughout.

Page 4 (Results: Assembly) paragraph 2 line 2: is "positive feedback" different from "cooperativity"? Is it a broader term that includes cooperativity, nucleation and other mechanisms? It may be useful to introduce some of these terms to avoid confusion by the readers.

“Positive feedback” is the broadest term as it is agnostic to mechanism. “Nucleation” refers to the initiation of a first order phase transition, which is one mechanism of positive feedback. Nucleation involves “cooperativity”, in that a higher order species is more stable than smaller species. However, cooperativity can occur for oligomers of finite size, whereas nucleation is reserved for phase transitions to species of infinite size. We appreciate that the use of so many related terms may have created more confusion than necessary. Hence, we have now revised the text to omit the more general terms -- “positive feedback” and “cooperativity” where possible.

Page 4 (Results: Assembly) paragraph 2 line 3: please define "TNF".

We have now fixed this and other acronyms.

Page 4 (Results: Assembly) paragraph 3 line 2: the use of size-exclusion chromatography to follow the size of complexes would assume that they are irreversible or very stable. It appears this may be the case here, but some discussion may be warranted.

We have now explained that SEC is appropriate for this experiment because large nucleation barriers generally imply stable assemblies.

Page 4 (Results: Assembly) paragraph 3 line 4 and throughout: the symbol for "kilodalton" is "kDa".

We have now fixed this mistake.

Page 4 (Results: Assembly) paragraph 3: I am not sure how the results discussed in this paragraph demonstrate that assembly occurs in cooperative fashion - just that there is a change in oligomeric states upon stimulation.

Cooperativity is implied by the absence of oligomer sizes between monomer and the large assembly. Nevertheless, we realized this can only be concluded in the case of homotypic assembly, which we cannot yet assume at this point in the paper. Therefore, we have revised this paragraph to say that the distribution is “consistent with” an underlying phase transition (which we then go on to prove).

Page 4 (Results: Assembly) paragraph 4 line 2: "WT" is not defined. Wild-type what? I presume "protein"?

We refer here to the wild-type protein. We have now fixed this mistake.

Page 4 (Results: Assembly) paragraph 4: it may be worth pointing out here the wild-type and mutant proteins expressed at similar levels; clearly the outcomes will depend on protein concentration in the cell. I believe the supplementary figure shows this to a large extent.

Indeed, our supplementary figure shows that the WT and mutant protein express to comparable levels. We have now pointed this out in the text.

Page 4 (Results: The adaptor) paragraph 1 line 4: "CARD domain" would stand for "caspase activation and recruitment domain domain". Please check throughout (including Supplementary Material).

We have fixed this mistake.

Page 4 (Results: The adaptor) paragraph 1 line 9: "expressed over a range of concentrations in cells" - this would imply the authors controlled expression - please rephrase to explain what exactly was done.

We have now rephrased this sentence to indicate that the range of expression results from the use of a genetic construct with cell-to-cell variation in copy number.

Page 5 (Results: The adaptor) paragraph 2 line 3 and throughout (including Supplementary Material): please use the Greek letter rather that "u" for micro.

We have now fixed this mistake.

Page 5 (Results: The adaptor) paragraph 3: this analysis is rather simplistic, it is not just the RMSD value, it is the nature of conformational change that is important? Please elaborate, I would think the papers presenting structural work have already discussed this to some extent?

The reviewer is correct; it is the nature of the conformational change that is most important. We are unsure how to accurately estimate the energy barrier separating the two conformations for each protein. However, we have now undertaken a collaboration to attempt to do so via FAST molecular simulations (Zimmerman and Bowman 2015). In lieu of the results of these ongoing studies, we have modified the text to acknowledge that RMSD does not necessarily relate to nucleation barriers.

Maxwell I. Zimmerman and Gregory R. Bowman. Journal of Chemical Theory and Computation, 2015, 11 (12), 5747-5757 DOI: 10.1021/acs.jctc.5b00737

Page 5 (Results: The adaptor) paragraph 4 line 5 and further in this section: some symbol(s) do not show in the pdf - before "(delta)", next page line 3-5 after "higher" and "both".

We have fixed this issue that resulted from exporting to a PDF file from our text editor.

Page 6 (Results: The adaptor) paragraph 4: interface IIa and IIIb are not introduced, and there is not even any reference provided here.

We have now added a reference for these mutations and elaborated on the interfaces IIa and IIIb.

Page 6 (Results: Pathogenic) paragraph 1 line 12: "FL" is not introduced.

We have now fixed this mistake.

Page 8 (Results: Pathogenic) paragraph 7: the text "absent the pathogenic mutations" is missing something.

We have now reworded this section.

Page 10 (Results: BCL10) paragraph 3: why does CARD9 CARD clustering peak and then disassemble (I guess "clustering" doesn't disassemble, please rewrite as well).

We have now fixed this mistake.

Page 11 (Results: MALT1) paragraph 1: I presume dimerization doesn't achieve the same level of proximity as higher-order multimerization?

Our interpretation here is that for MALT1, activation requires close proximity of more than two molecules. Although our dimerization module did not activate the caspase-like domain of MALT1, we know that it achieves close enough proximity to activate the caspase domain of CASP8. Hence we believe the MALT1 mechanism has a stoichiometry requirement in addition to a proximity requirement. This is, of course, consistent with the fact that activation normally occurs in the context of polymers rather than dimers.

Page 11 (Results: Ancient) paragraph 1 line 4: is this AlphaFold2?

That is correct, we used AlphaFold2. We have added that detail.

Page 12 (Discussion) paragraph 4: not sure if "molecular examples of evolutionary spandrels" will be clear to most readers.

We have now explained what evolutionary spandrels are, and elaborated on the relationship to our findings.

Page 14 (Materials: Plasmid) line 2 and throughout: "Golden Gate" is usually capitalized. Similar for "Gibson" further in the paragraph. The English in this paragraph is not up to standard in general; for example "Then placing..." is not a complete sentence, and a number of sentences ending with "via gibson" need to be rewritten.

We have now rewritten this paragraph.

Page 16 (Materials: Cell) line 4 and throughout: "2" in "CO2" should be subscripted.

This is now fixed.

Page 16 (Materials: Transient) line 6 and throughout (including Supplementary Material): please use a space between number and unit ("35 mm").

This is now fixed.

Page 16 (Materials: Generation) line 4 and throughout: to distinguish from "gram", please italicize "g" and/or use "x g".

We have now fixed this.

Page 17 (Materials: Yeast) line 3: please specify which table is "table X".

We have now fixed this mistake.

Page 17 (Materials: Mammalian) line 1: please provide full reference. Same next paragraph line 2.

We have now fixed this.

Page 17 (Materials: DAmFRET) line 3: "SSC" and "FSC" are not defined.

We have now fixed this.

Page 18 (Materials: Fluorescence) line 10: "Coefficient" does not have to be capitalized. It does not have to be defined again in the next paragraph.

We have now fixed this.

Page 19 (Materials: Optogenetic) line 1: "performed" rather than "made"?

We have now fixed this.

Page 19 (Materials: Protein) line 12: the Compass software doesn't have a reference?

We have now added the reference to the software.

References: please make format consistent: articles titles in sentence or title case.

We have now formatted all references to be consistent.

Legend to Fig. 1: I suggest "Schematic diagram"; and "h" rather than "hrs"; please check throughout (including Supplementary Material).

We agree with this suggestion.

Legend to Fig. S1: is "TNF-a" supposed to be "TNF-alpha"?

We have fixed this.

Legend to Fig. S7: please capitalize "Figure 2H".

We have fixed this.

Legend to Fig. S10F: please move "Dox" behind the concentration.

We have fixed this.

Fig. S14B: the colours in the superposition make it difficult to see the differences.

We have used a different color now.

Legend to Fig. S14: I suggest "structure...predicted by AlphaFold" (2?) and include the reference.

We agree with this suggestion.

Reviewer #4 (Significance (Required)):

As argued above, the significance of this paper is that it tests directly important hypotheses proposed or assumed previously, and does so at a technically high standard. No published report has done so to a similar extent.

The paper should be of interest to a broad audience from cell biologists and immunologists to biochemists, biophysicists and structural biologists.

My expertise is in structural biology or systems similar to the one studied here.

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

Learn more at Review Commons

---

Referee #4

Evidence, reproducibility and clarity

The characterization of open-ended signalosomes in a number of innate-immunity and cell-death pathways, in particular formed by domains from the death-fold family, has led to the suggestions that these complexes allow a switch-like signalling response suitable for these pathways. It appears that this has been widely accepted. However, these suggestions are based largely on indirect observations and speculation.

Rodriguez-Gama and coworkers have decided to test these suggestions more directly. Their results confirm the suggestions. Based on my own experience, papers that validate widely adopted suggestions are often not considered seriously by top journals, who are looking for hot topics/paradigm-changing/surprising type results. I would urge the editors to consider seriously work such as in this paper, which directly tests important suggestions and does so at a technically high standard. The authors use a range of ingenious approaches, both with recombinant proteins and in cells, and including proteins from organisms from different parts of the evolutionary tree, to support their interpretations, so it is an extensive and high-quality study. I am impressed that so many different fusion proteins with fluorescent tags continued to function as expected, but I guess the authors controlled for this as much as they could.

Having said all this, I do get the feeling the authors are "over-selling" the nucleation barrier aspect of these signalling mechanisms. It is clearly an important and critical aspect of signalling in many systems, but then it is not the only important aspect; a number of other regulatory inputs play a role in different systems. So the statement "Our findings introduce a novel structure-function paradigm" in my view is overstretching things somewhat. Further in the Discussion section, the authors state "Existing explanations for the preponderance of ordered polymers in immune cell signalosomes have centered on the functions of multivalency at steady state, such as scaffolding and sensitivity enhancement resulting from the cooperativity of homo-oligomerization". They cite a small (and non-exhaustive) number of papers discussing this topic; all these include "seeding" or "nucleation" as an important part of the proposed mechanism. So I suggest the authors provide a more balanced discussion of this aspect. Different pathways appear to display a different level of switch-like behaviour, and one thing that the current version of the manuscript is missing is more discussion of other death fold-based systems and how the results on the CBM signalosome apply to these, and also other systems such as TIR domain-based ones, which currently get no mention whatsoever. In the CBM system, there seems to be one main nucleation barrier; can there be more than one in others?

One other aspect I need to express some criticism about is attention to detail - especially with a paper focusing on the physics behind biological processes, I would expect a higher standard of getting the terminology and units correct - see specific examples below. This can obviously be fixed easily.

Specific points are listed below. No page or line numbers are provided so I have done my best to make it clear what the comments refer to.

1. Abstract line 6 and throughout: in "NF-kB", the "k" is supposed to be "kappa" (Greek letter) - it stands for "nuclear factor kappa-light-chain-enhancer of activated B cells", not fully defined in the manuscript as far as I can see. Occasionally, small k is also used instead of the small cap K or whatever the authors used most of the time, but I don't think any of them use the Greek letter.
2. Page 2 (Introduction) paragraph 2 line 9: period missing at the end of sentence. Same Page 4 (Results: Assembly) paragraph 4 line 3.
3. Page 2 (Introduction) paragraph 2 line 15 and throughout: in long sentences, more commas can help help readability, for example before "leading" here. Similar page 15 paragraph 2 line 3 after "Additionally", paragraph 4 line 2 before "which".
4. Page 4 (Results: Assembly) paragraph 2 line 2: is "positive feedback" different from "cooperativity"? Is it a broader term that includes cooperativity, nucleation and other mechanisms? It may be useful to introduce some of these terms to avoid confusion by the readers.
5. Page 4 (Results: Assembly) paragraph 2 line 3: please define "TNF".
6. Page 4 (Results: Assembly) paragraph 3 line 2: the use of size-exclusion chromatography to follow the size of complexes would assume that they are irreversible or very stable. It appears this may be the case here, but some discussion may be warranted.
7. Page 4 (Results: Assembly) paragraph 3 line 4 and throughout: the symbol for "kilodalton" is "kDa".
8. Page 4 (Results: Assembly) paragraph 3: I am not sure how the results discussed in this paragraph demonstrate that assembly occurs in cooperative fashion - just that there is a change in oligomeric states upon stimulation.
9. Page 4 (Results: Assembly) paragraph 4 line 2: "WT" is not defined. Wild-type what? I presume "protein"?
10. Page 4 (Results: Assembly) paragraph 4: it may be worth pointing out here the wild-type and mutant proteins expressed at similar levels; clearly the outcomes will depend on protein concentration in the cell. I believe the supplementary figure shows this to a large extent.
11. Page 4 (Results: The adaptor) paragraph 1 line 4: "CARD domain" would stand for "caspase activation and recruitment domain domain". Please check throughout (including Supplementary Material).
12. Page 4 (Results: The adaptor) paragraph 1 line 9: "expressed over a range of concentrations in cells" - this would imply the authors controlled expression - please rephrase to explain what exactly was done.
13. Page 5 (Results: The adaptor) paragraph 2 line 3 and throughout (including Supplementary Material): please use the Greek letter rather that "u" for micro.
14. Page 5 (Results: The adaptor) paragraph 3: this analysis is rather simplistic, it is not just the RMSD value, it is the nature of conformational change that is important? Please elaborate, I would think the papers presenting structural work have already discussed this to some extent?
15. Page 5 (Results: The adaptor) paragraph 4 line 5 and further in this section: some symbol(s) do not show in the pdf - before "(delta)", next page line 3-5 after "higher" and "both".
16. Page 6 (Results: The adaptor) paragraph 4: interface IIa and IIIb are not introduced, and there is not even any reference provided here.
17. Page 6 (Results: Pathogenic) paragraph 1 line 12: "FL" is not introduced.
18. Page 8 (Results: Pathogenic) paragraph 7: the text "absent the pathogenic mutations" is missing something.
19. Page 10 (Results: BCL10) paragraph 3: why does CARD9 CARD clustering peak and then disassemble (I guess "clustering" doesn't disassemble, please rewrite as well).
20. Page 11 (Results: MALT1) paragraph 1: I presume dimerization doesn't achieve the same level of proximity as higher-order multimerization?
21. Page 11 (Results: Ancient) paragraph 1 line 4: is this AlphaFold2?
22. Page 12 (Discussion) paragraph 4: not sure if "molecular examples of evolutionary spandrels" will be clear to most readers.
23. Page 14 (Materials: Plasmid) line 2 and throughout: "Golden Gate" is usually capitalized. Similar for "Gibson" further in the paragraph. The English in this paragraph is not up to standard in general; for example "Then placing..." is not a complete sentence, and a number of sentences ending with "via gibson" need to be rewritten.
24. Page 16 (Materials: Cell) line 4 and throughout: "2" in "CO2" should be subscripted.
25. Page 16 (Materials: Transient) line 6 and throughout (including Supplementary Material): please use a space between number and unit ("35 mm").
26. Page 16 (Materials: Generation) line 4 and throughout: to distinguish from "gram", please italicize "g" and/or use "x g".
27. Page 17 (Materials: Yeast) line 3: please specify which table is "table X".
28. Page 17 (Materials: Mammalian) line 1: please provide full reference. Same next paragraph line 2.
29. Page 17 (Materials: DAmFRET) line 3: "SSC" and "FSC" are not defined.
30. Page 18 (Materials: Fluorescence) line 10: "Coefficient" does not have to be capitalized. It does not have to be defined again in the next paragraph.
31. Page 19 (Materials: Optogenetic) line 1: "performed" rather than "made"?
32. Page 19 (Materials: Protein) line 12: the Compass software doesn't have a reference?
33. References: please make format consistent: articles titles in sentence or title case.
34. Legend to Fig. 1: I suggest "Schematic diagram"; and "h" rather than "hrs"; please check throughout (including Supplementary Material).
35. Legend to Fig. S1: is "TNF-a" supposed to be "TNF-alpha"?
36. Legend to Fig. S7: please capitalize "Figure 2H".
37. Legend to Fig. S10F: please move "Dox" behind the concentration.
38. Fig. S14B: the colours in the superposition make it difficult to see the differences.
39. Legend to Fig. S14: I suggest "structure...predicted by AlphaFold" (2?) and include the reference.

Significance

As argued above, the significance of this paper is that it tests directly important hypotheses proposed or assumed previously, and does so at a technically high standard. No published report has done so to a similar extent.

The paper should be of interest to a broad audience from cell biologists and immunologists to biochemists, biophysicists and structural biologists.

My expertise is in structural biology or systems similar to the one studied here.

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

Learn more at Review Commons

---

Referee #3

Evidence, reproducibility and clarity

The study by Rodriguez Gama et al. addresses the molecular function of CBM complex-forming proteins CARD9, BCL10 and MALT1 in the activation of myeloid cells, using optogenetic tools, transcriptional reporters and biochemical approaches. It is known from previous studies that Bcl10 oligomerizes into filamentous oligomeric structures incorporating Malt1, and that these structures are nucleated by receptor-induced activation of CARD proteins such as CARD11 (in lymphocytes) or CARD9 (in myeloid cells), but the mechanism underlying the assembly of the resulting CBM complexes remain incompletely understood.

The authors develop beautiful optogenetic tools to address this question, and convincingly demonstrate that CARD9-mediated nucleation of BCL10 triggers a binary cellular NF-kB response in a spring-load-like fashion, and identify mutants of BCL10 and CARD9 that impact this capacity. Unfortunately, however, the authors do not do a good job to simplify this complex problem so it can be easily understood. In particular, the choices of mutants, models and experiments are not consistent between figures, and some data seem to be arbitrarily added or omitted. Complex hybrid constructs are also used, without assessing whether these are indeed functional in the corresponding ko cells. The paper would therefore benefit from a major overhaul. We also noticed that the literature is often not cited adequately and have included a (non-exhaustive) list of examples of wrong, incomplete, or erroneous citations below.

1) The initial observations of binary signaling are derived from a reporter system. Although there are controls to show that the reporter used does not function intrinsically cooperatively, it would be nice to see additional data to show that cooperativity occurs also at the level of endogenous response systems, for instance by qPCR-based assessment of a natural NF-kB target gene (induced for example by TNFa versus B-glucan in THP-1 cells, and by TNFa versus PMA in 293T cells).

2) The cell lines in Figures 1D-E (and also some of the BCL10 mutants used later on) would have been better run in the assays in the early parts of Figure 1. The final conclusion prior to the section The adaptor protein BCL10 is a nucleation-mediated switch is otherwise not justified. This is a central tenet of the paper, that is referred to again, with some other ancillary data to support it. These mutants reappear later in the paper, but it would have been better, and easier to make rescue lines of BCL10 KO in Figure 1, otherwise the logic is lost, and the models seem chosen arbitrarily.

3) Expression with microNS is not well controlled and gives little real evidence for what is occurring. It is unclear what the concentration of the protein expressed was, but certainly the relative expression of the CARD9(CARD) and the microNS version should be assessed. Also, the AmFRET profile of CARD9CARD looks very weird, it cannot be compared to BCL10.

4) We are not convinced of the usefulness of the introduction of a slew of disease-causing CARD9 mutations that may or may not be relevant to the authors' point. The fact that they do or do not function in a specific sub portion of an assay that may or may not be relevant to biological activity seems to be of interest but without biochemical understanding, little is clear.

5) The Optogenetic experiments are interesting, but difficult to interpret without evidence that these MALT1 constructs are indeed still functional when expressed in MALT1-deficient THP-1 cells. We do not therefore think that this experiment shows a necessity for clustering to signal, just a sufficiency, and in a highly artificial construct.

6) In the introduction and other parts of the paper, there are numerous instances where the previous literature in the field is not adequately cited. Examples include:

• In the introduction, it is weird to cite one original paper (a MALT1 ko study by Ruland et al., 2001; there are several other studies of ko papers for CBM components that would merit being citated along with this study) together with two reviews on that topic (Ruland and Hartjes 2019 and Gehring et al. 2018)
• In the introduction, the original study by Wang et al., 2002 should be cited together with Rebeaud et al., 2002; the two studies on the same topic were published back-to-back
• In the introduction, the statement "CARD10 and CARD14 are expressed in nonhematopoietic cells including intestinal and skin epithelia, respectively" should be supported by citations.
• Still in the introduction, the 2 references for the statement "... CARD14 gain of function mutations cause psoriasis (Howes et al., 2016; Jordan et al., 2012)" are not appropriate. There are several reports of patients with CARD14 mutations (the study by Jordan et al is only one of them) and several CARD14 mouse models that provoke a psoriasis-like phenotype, which would merit being cited.
• In the following sentence: "Point mutations and translocations involving BCL10 and MALT1 cause immunodeficiencies (Ruland and Hartjes, 2019), testicular cancer (Kuper-Hommel et al., 2013), and lymphomas (Zhang et al., 1999).", the citation style also seems completely random, combining the citation of a single original paper for lymphomas (Zhang et al. 1999) (there are several other important original studies on that topic or recent reviews that could be cited instead), together with a review on immunodeficiencies (Ruland and Hartjes, 2019) and then another single example for a role of BCL10 and MALT1 in carcinoma (the study by Kuper-Hommel et al. is one, but several other original publications exist on the latter topic, showing for example a role in breast carcinoma or glioblastoma).
• In the first section of the results, the reference cited for endogenous CARD10 expression in 293T cells (Ruland et al., 2001) is wrong, no endogenous CARD10 expression was assessed in that study

Significance

The paper deals with a complex question, namely how the CBM signalosome assembles and functions to stimulate NF-kB signaling. This question is important to the understanding of pro-inflammatory immune responses and basic life sciences in general. As the focal point of the paper is complex, and tools to study such phenomena are at the limit of technical capabilities, this further increases the potential impact of the work.

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 their manuscript entitled "A nucleation barrier springloads..." Rodriguez-Gama et al. dissect the assembly mechanism of the signalosome, composed of the proteins CARD9, BCL10 and MALT1, using a novel in-cell biophysical approach (DAmFRET). They first overexpressed fluorescently tagged versions of the proteins to promote their assembly in yeast and mammalian cells, finding that CARD9 forms higher order assemblies across a wide range of concentrations with no discontinuity in the DAmFRET profile. In contrast, the DAmFRET profile of BCL10 showed a clear separation between monomers and higher order assemblies, which started to form spontaneously only at higher BCL10 concentrations. Furthermore, at the two states of the proteins co-exist at all concentrations. These observations imply that there is a nucleation barrier to forming BCL10 assemblies. MALT1 showed no change in FRET regardless of its expression level. These observations, alongside fluorescence microscopy of the assemblies, and previous structural studies, suggest that BCL10 forms self-templating polymers that act as a switch for an all-or-nothing immune response, assayed in this case by monitoring the nuclear translocation of the NF-kB subunit p65. The authors also assessed the effects of known disease-causing mutations on the nucleation barrier, showing that changes in the strength of the nucleation barrier can have major effects on signalosome function. Finally, they used optogenetic methods to trigger assembly of individual signalosome components, providing insight into the minimal components/conditions required for signalosomes to work.

Major comments:

Overall, the experiments by Rodriguez-Gama et al. offer convincing evidence that there is a nucleation barrier to BCL10 polymerisation, and that a CARD9 template is sufficient to overcome the barrier. Although the existence of a nucleation barrier had already been postulated, based on structural and other studies (referenced by the authors), it had lacked a rigorous demonstration. This work provides that demonstration, which is important for the signalosome field and more broadly applicable to researchers studying cellular decision making. The study further demonstrates that DaMFRET is an excellent to study protein assembly processes in their native environment, allowing the authors to tackle a question that would have been technically very difficult to address otherwise. The optogenetic experiments are a nice sufficiency test for their ideas.

We feel there are a few key points to address before publication.

1) One of the main conclusions is that spring-loading the nucleation barrier with high super-saturating BCL10 concentrations allows a decisive response. Although much of the data strongly imply this conclusion, the dependence of the immune response on BCL10 concentration was not tested directly. A key prediction of the nucleation barrier is that at concentrations below saturation, BCL10 should not be able to induce an all-or-nothing response when stimulated. At saturated/super-saturated concentrations BCL10 should be able to induce a response. At deeply super-saturated concentrations the response should start to be activated spontaneously in the absence of an external stimulus. These predictions could be tested using the doxycycline-inducible BCL10 system (Figure S2D), without establishing major new experimental avenues. We feel that such an experiment would strengthen the main conclusion. It might also help to shed light on whether being highly supersaturated enables a more decisive response than being just saturated.

2) Intuitively, readers might expect that if BCL10 is supersaturated then, once nucleated, it would rapidly assemble at the nucleation sites. In Figure 5B, CARD9CARD-miRFP670nano-Cry2 assemblies are optically induced throughout the cell. However, BCL10 appears to nucleate at just a few sites with a few minutes delay. More widespread nucleation and growth of BCL10 polymers seems to take longer (20-40 minutes, Figures 5B and 5C), after CARD9CARD-miRFP670nano-Cry2 has disassembled. Furthermore, in Figures 4D and 4E, very few BCL10 assemblies are visible/quantifiable after 70 minutes PMA exposure, but p65 has clearly entered the nucleus. It looks like BCL10 assembly slightly lags behind p65 nuclear entry. Can the authors provide a more detailed explanation of these kinetics?

3) Related to point 2 above, in Figure 5D, the leftmost cell in the field of view clearly contains CARD9CARD assemblies but there are no BCL10 assemblies and p65 is not imported into the nucleus (in contrast to the central cell in the field of view). How often does CARD9CARD optogenetic assembly lead to BCL10 assembly? In other words, can the authors quantify the cell-to-cell variability in this experiment?

Minor comments:

While the work is scientifically well done, the text reads as though it is meant for experts rather than a broad audience. This is a pity because it risks alienating readers. We suggest that some adjustments to the text (mainly additional explanations and not ruling out alternative interpretations of the data) would widen the audience and increase the impact of this important study. Below are some suggestions that might help.

1) In the first results section, the authors write: 'This suggests that Bcl10 but not CARD9 assembly occurs in a highly cooperative fashion that could, in principle (Koch, 2020), underlie the feed forward mechanism.' It isn't obvious how Figure 1 leads to this statement. Could the authors give a more detailed explanation?

2) One limitation of DAmFRET is that it can only detect a nucleation barrier where there is a difference in FRET between the monomer and the assembled form of the protein. However, it can't necessarily detect when there is not a nucleation barrier i.e. if there's no difference in FRET. The text seems to suggest that CARD9 and MALT1 don't have nucleation barriers to their assembly. While this might not be intentional, it would be helpful to explicitly state that CARD9 and MALT1 could also possess such barriers that are not detectable by this method. This wouldn't detract from the finding that BCL10 has a barrier that plays an important function.

3) In the final results section, the idea that MALT1 activation doesn't depend on BCL10 polymer structure doesn't necessarily follow from the data. An alternative interpretation is that optogenetic clustering of MALT1 causes it to recruit BCL10 and form BCL10-MALT1 filaments (structure solved by Schlauderer et al., 2018). Also, the optogenetic clustering of MALT1 may mimic some structure found in the BCL10 cluster. Therefore, we are neither convinced that the data unambiguously show that MALT1 activation strictly depends on multi-valency rather than an ordered structure of BCL10 polymers nor that this conclusion is truly necessary for the paper.

4) What optical density do the yeast cells reach during the 16h induction in galactose? If they are in stationary phase, this could affect the assembly status of the proteins being expressed, as the cytoplasm becomes glassy when cells are starved, and this coincides with widespread protein aggregation/assembly (Joyner et al., 2016; Munder et al., 2016).

5) Although these experiments show that thermodynamically lowering the BCL10 nucleation barrier (e.g. by post-translational modifications or protein expression levels) isn't required for a response, they don't rule it out. It would be good to state this in the discussion, as cells may have multiple mechanisms of switching on the signalosome.

6) The discussion compares signalosomes with condensates formed by liquid-liquid phase separation. This is an interesting comparison but it suggests that disordered assemblies would not be capable of performing signalosome-like functions. This needs to be explained more clearly. For example, non-amyloid prions seem to form gel-like assemblies with a high nucleation barrier that are capable of driving heritable traits, likely through self-templating (Chakravarty et al., 2020). Such examples could represent disordered assemblies with signalosome switch-like behaviour. Furthermore, there are examples of condensates that are induced by environmental changes e.g. Pab1 and Ded1 condensates (Riback et al., 2017; Iserman et al., 2020). This potentially allows the proteins to reach high concentrations and remain un-condensed until a change in heat or pH overcomes a nucleation barrier required for condensate formation. Although the condensates aren't self-templating, they seem to require energy for their disassembly. Combined, this also allows switch-like behaviour, where the switch is flipped back to the uncondensed off state once conditions return to normal. In general, crossing a phase boundary can represent a switch-like response. Finally, recent electron-tomography experiments show that ASC puncta comprise clusters of filaments (Liu et al., 2021, biorxiv). CARD9/BCL10 assemblies may have similar ultrastructures and liquid-liquid phase separation may well play a role in their assembly.

7) Can the authors comment on the loss of BCL10 in Echinodermata, Anthropoda, Nematoda? Is there another protein that plays a similar role? Could a CARD or PCASP protein possess self-templating properties? Could other methods of control be at play e.g. protein expression?

8) Figures 1B/1C: Can the authors comment on why the active cells plateau at about 70-75%? This is a striking feature of the plots, but the explanation may not be obvious to readers.

9) Figures 1D/1E: What was the concentration of B-glucan used in this experiment? This could be included in the figure legend. If greater than 1ug/ml this means that the % of active cells in Figure 1B matches the % of cells with BCL10 assemblies in Figures 1D/1E, which is potentially an important point.

10) Use of both 'BCL10' and 'Bcl10' when referring to the protein.

11) In the supplementary figures there are some formatting problems/missing words in the figure legends. In Figure S11 there is a black box covering the lower part of the figure.

References used in this review

Chakravarty, A.K. et al. (2020) "A Non-amyloid Prion Particle that Activates a Heritable Gene Expression Program," Molecular Cell, 77(2), pp. 251-265.e9. doi:10.1016/j.molcel.2019.10.028.

Iserman, C. et al. (2020) "Condensation of Ded1p Promotes a Translational Switch from Housekeeping to Stress Protein Production," Cell, 181, pp. 818-831.e19. doi:10.1016/j.cell.2020.04.009.

Joyner, R.P. et al. (2016) "A glucose-starvation response regulates the diffusion of macromolecules," eLife, 5. doi:10.7554/eLife.09376.

Munder, M.C. et al. (2016) "A pH-driven transition of the cytoplasm from a fluid- to a solid-like state promotes entry into dormancy," eLife, 5(MARCH2016). doi:10.7554/ELIFE.09347.

Riback, J.A. et al. (2017) "Stress-Triggered Phase Separation Is an Adaptive, Evolutionarily Tuned Response," Cell, 168(6), pp. 1028-1040.e19. doi:10.1016/j.cell.2017.02.027.

Schlauderer, F. et al. (2018) "Molecular architecture and regulation of BCL10-MALT1 filaments," Nature Communications 2018 9:1, 9(1), pp. 1-12. doi:10.1038/s41467-018-06573-8.

Significance

The existence of a nucleation barrier had already been postulated, based on structural and other studies (referenced by the authors), it had lacked a rigorous demonstration. This work provides that demonstration, which is important for the signalosome field and more broadly applicable to researchers studying cellular decision making. The study further demonstrates that DaMFRET is an excellent to study protein assembly processes in their native environment, allowing the authors to tackle a question that would have been technically very difficult to address otherwise.

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

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

In this manuscript, Gama et al. use a biophysical assay DAmFRET, structural analysis, and optogenetic tools to uncover the nucleation mechanism of CBM signalosome. They performed experiments first in yeast cells that lack death folds or related signaling networks, then confirmed their discoveries in human cells. The results presented here are clear and convincing. The paper is very well presented and clearly written.

They found it is the CARD domain of BCL10 that acts as a molecular switch that drives all-or-none activation of NF-kB. Monomeric BCL10 possesses an unfavorable conformation and serves as a nucleation barrier, keeping BCL10 in a supersaturated inactive state that allows for binary activation upon stimulation.

They also characterized CARD9 CARD domain and a coiled-coil region. They reasoned that CARD9CARD functions as a polymer seed to nucleate BCL10, and that the coiled-coil region has multimerization ability to facilitate nucleation. Furthermore, they characterized that MALT1 activation doesn't depend on BCL10 polymers but its own proximity. And MALT1 induces graded NF-kB activation, thus further demonstrating the binary activation is conferred by BCL10.

Major comments:

1) Fig S1D and E, the authors used TNF-a to activate NF-kB independent of CBM signalosome and found the activation in each cell increased with dose. In contrast, CBM activation led to bimodal cell activation. The authors claim that this is evidence that positive feedback upstream of NF-kB. We do not believe this claim can be made from this comparative experiment alone. We agree that positive feedback is important for activating an NF-kB response, but the comparison between CBM and TNFa is inaccurate and glosses over published data. Specifically, there is published data that TNF-a does activate a 'switch-like' or digital response, as defined by the translocation of p65 (see (Tay et al. 2010) among other studies that have examined p65 translocation at the single-cell level). The difference in T-sapphire expression between CBM and TNF activation is most likely due to TNFa induced oscillations of p65 translocation (although this is speculation on our part). Therefore we suggest to the authors that the TNF-a data (Fig S1D and E) should be omitted, as the claim of switch or not-switch as pertains to TNF signaling is more complex and nuanced than presented here. We believe omitting this data will strengthen the manuscript and avoid confusion in the field. The bimodal expression of the T-sapphire NF-kB reporter driven by the CBM signalosome activation is sufficient to claim an all-or-none response.

2) Fig 3B, the authors introduced CARD9CARD-µNS as a stable condensed seed for BLC10. However, considering CARD9CARD can form polymers at high concentration (Fig 3B and S3D), are these high expression levels of CARD9CARD able to induce BCL10-mEos3.1 assembly (as measured by DamFRET in yeast cells)? Can the authors examine BCL10 FRET at these high expression level of CARD9CARD? We assume that BCL10 will be assembled in these cells. This would provide a valuable control experiment and support the author's conclusions.

3) Fig 3C, the text said "Whereas WT CARD9CARD assembled into polymers at high concentration, the pathogenic mutants R18W, R35Q, R57H, and G72S failed to do so (Fig 3C and S7B,C), explaining why they cannot nucleate BCL10". This claim that these mutants can not nucleate BCL10 does not have a figure call out or a reference. The authors then show the results in Fig 3E which supports this claim. Even though they were done in the context of full length CARD, all proteins contain the I107E mutation that releases autoinhibition. For clarity, the authors should consider rearranging the text to avoid explaining a phenomenon and making conclusions before showing the results.

4) Fig 4D, E and Video 1, the authors showed the nucleation of BCL10 into puncta within live cells is followed by p65 translocation to the nucleus. The authors claim that 'this result suggests that BCL10 is indeed supersaturated prior to stimulation' (paragraph 2 section titled BCL10 is endogenously supersaturated'). We fail to understand how this live-cell experiment leads to the conclusion BCL10 is supersaturated before stimulation. We think this text should be deleted from the text, or put into context with the DAmFRET data that lead the authors to make this claim. It would be interesting for the authors to define in discussion what are the golden criteria to claim a protein exists in a supersaturated state with live cells (by microscopy or other methods)? Adaptor protein assembly into puncta and the subsequent nuclear translocation of transcription factors is a common phenomenon across signalling pathways. Not all these pathways rely on signalling adaptors existing in a supersaturated state. The field of cell signaling (and cell biology in general) would benefit from a detailed definition of how these physical-chemical definitions of proteins are supported by experimental data. We believe that this paper will become a seminal paper in the field, and future work will benefit from a clear definition of how a claim of supersaturation is derived from the data.

5) Regarding the supersaturated state of BCL10, the authors convincingly use optogenetics to show how transient assemblies of CARD-Cry2 can template BCL10 assembly. This is a convincing experiment that shows templated nucleation of BCL10. To strengthen the claim that BCL10 is supersaturated endogenously we suggest the author quantify the expression of BCL10-mScarlet and CARD-Cry2 and ideally show that this phenomenon can be observed at expression levels equivalent to endogenous.

Minor comments:

1) Special character "delta" is not displayed in the text (instead only a space).

2) Several cell lines including mouse, human, and yeast lines were used across this manuscript. It would be clearer and more helpful if the exact cell type of the line could be indicated. Such as, "BCL10-mEos3.1 yeast cells" instead of "BCL10-mEos3.1 cells", "BCL10-mScarlet HEK293T cells" instead of "BCL10-mScarlet cells".

3) Fig 5B, the authors indicated that BCL10 colocalized with CARD9CARD, then please show the merged image as well.

4) Fig 6E, authors claimed that cells were stimulated with blue light for the indicated durations. The longest duration is 12 hours. Please specify if it was continuous exposure or several rounds of exposure in the indicated durations.

Significance

This work used a combination of FRET and optogenetic tools to engineer CBM signaling and visualize the effects. They incorporated knowledge from structure biology, together with their results from mutations and truncations, dissected the significance of each protein in CBM signalosome, and demonstrated in detail how higher-order assemblies make all-or-none cellular decisions. We believe this paper will be a seminal paper in the field of cell signalling and cytoplasmic organization. It defines a new paradigm of macromolecules assembly of signalling complexes as being dependent on protein existing in a supersaturated state. Importantly this paper opens up new questions regarding macromolecular signaling complexes (found in many innate immune signaling pathways): How is protein supersaturation maintained and used throughout evolution to construct biochemical signalling switches?

This paper will be of particular interest to scientists working on immunity and cell signalling, especially in the field of higher-order assemblies. However, we feel the impact of this paper goes beyond these fields, and we believe this manuscript will be of broad interest to the cell biology and biophysics communities. For reference, our expertise is in innate immunity and cell biology.

Referees cross-commenting

In general, I agree with reviewer 4. However, I'm afraid I have to disagree with reviewer 3 that the paper requires 'a major overhaul'. I also believe that reviewer 3 suggestion #1 to use qPCR to assess NF-kB target genes is not a 'constructive and realistic suggestion'. Or, to put it another way, not within the guidelines of the RC for reviewers. This type of suggestion is too open-ended to be of use to the authors. Which should the authors analyze of the tens to hundreds of genes activated by NF-kB? A rigorous and robust editor should ignore this comment.

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

Two reviewers commented on the smeared appearance of Tae1 bands in our Western blot analyses (Figure 4F and 5B) and asked us to improve their technical quality.

-We agree and will repeat these experiments with more careful attention to lysate preparation, using a higher percentage SDS gel for better separation of low molecular weight proteins as suggested.

Reviewer 2 requested that we assess how Tae1 variants impact interbacterial competition outcomes.

-We agree that this would be interesting to take a look at. While this will not be feasible for every variant we examine in the paper, we can conduct comparative interbacterial assays between P. aeruginosa and E. coli using P. aeruginosa strains with a tae1 point mutation for c110s. Given that our biochemical experiments show that this hyperactive variant evades inhibition by the cognate immunity protein, we expect that this may decrease P. aeruginosa fitness, even in the context of competition.

More generally, we think that examining Tae1 variants in the context of interbacterial competitions would be a critical orthogonal approach in order to validate that the DMS results have any bearing on competition outcomes. However, we feel that major focus of this paper is on the more molecular and biophysical insights that our approach can offer. Our study tests our assumptions about the kinds of features and surfaces that are important for proteins that engage with non-canonical complex substrates. It is, of course, interesting to think about the implications of this for physiological phenotypes and the drivers of toxin evolution. It is also exciting to imagine how this kind of information could be used to one day engineer certain interbacterial outcomes. We hope that others in the field will push our efforts into these directions, but we do not feel that these directions are essential for our conclusions. However, our conclusions on the molecular and biophysical aspects have helped generate interesting hypotheses in microbial ecology that could be largely followed up on by others.

In order to conduct well-controlled P. aeruginosa:E. coli competition assays for more Tae1 variants, we would need to generate a significant number of new P. aeruginosa strains encoding point mutations for each of our variants across several genetic backgrounds. The competitions themselves also require a considerable amount of work to optimize and quantify. We are able to do this for one of the variants as previously mentioned (C110S). It’s important to note that the first author of this paper, who was the primary driver of this work, is no longer in my lab or in academia. As for myself, I am also in the middle of a transition out of academia and am actively ramping down my lab at UCSF. I no longer have the space or appropriate set-up to support this longer-term effort.

Reviewer 2 asked that we examine Tae1 (WT and C110S) expression levels in vivo to more precisely examine whether increased self-intoxication by Tae1C110S in P. aeruginosa was due to differences in toxin activity or toxin levels.

We agree with this suggestion and will look at toxin protein levels by Western blot analysis in the context of P. aeruginosa cells grown 1) alone on solid media and 2) together with E. coli on solid media during interbacterial competition using conditions that match our other competition assays.

All 3 reviewers asked us to provide more experimental evidence addressing the hypothesis that differential peptidoglycan (PG) affinity across Tae1 variants could explain variation in toxic activity.

-We agree that this is an interesting point to follow up on further. To be clear, we also do not know whether this hypothesis is true at this stage, and the answer is not necessarily critical for our central advance, but we would like to give it a try! We have devised an approach to ask the question experimentally across a subset of our deep mutational scanning (DMS) variants.

Reviewer 1 suggested that we quantify in vitro binding affinities for PG using isothermal titration calorimetry (ITC). However, given that ITC requires high concentrations of well-defined homogeneous substrates, which we are not able to generate for more complex higher order structures of cell wall PG, we propose a pull-down based approach.

Briefly, we plan to conduct pull-downs using insoluble, purified cell wall sacculi from our two E. coli grown under the two conditions as bait for recombinant Tae1 proteins. Given that intact sacculi or inherently insoluble, we can simply collect bound Tae1 through centrifugation of sacculi pellets and examine the amount of Tae1 associated by Western blot analysis. These analyses will need to be conducted across a titration of Tae1 concentrations and also with catalytic activity inhibited to avoid solubilization of sacculi. We will block Tae1 hydrolysis by carrying out pull-downs in the presence of a general commercially-available cysteine hydrolase inhibitor, E64. If there is indeed differential affinity for PG underlying lytic differences across Tae1 variants, we would expect to see greater relative association of Tae1 variants with the type of cell wall sacculi that they more effectively lyse in our DMS screen. We would expect the reverse trend to also be true (lower affinity for less active variants).

Reviewer 1 would like to know if we have done lysis experiments with any E. coli mutants that only impact PG density but not PG polymer structure? If they haven’t tested any E. coli mutants, have we done lysis experiments using drugs that have a similar impact on PG? Even if we don’t include these data in the paper, the reviewer would like us to comment on the trends we have observed.

We have not done experiments in any mutants or chemical backgrounds known to only impact PG density but not polymer structure. We think this would be a very interesting angle! But unfortunately this is outside the scope of this study. It would require that we first experimentally confirm that the restrictive effect on only density is clearly demonstrated using a variety of techniques, including microscopy, chemical analyses, and biophysical probing of sacculi.

Reviewer 1 asked for additional DMS screens in more conditions

We love this idea! In fact, we hope that others are motivated to adopt our workflow to run many more DMS screens for T6S toxins, as we believe these screens provide a lot of useful and sometimes surprising insights that could be of great interest to others. However, we believe that the primary goal of this paper is to establish this methodology as a compelling approach for studying toxins and, more generally, proteins with complex cellular substrates. It does not necessarily fall within the scope of this paper to fully assess the mechanistic implications of cell wall diversity across a wide range of conditions.

In our experience, rigorously conducting DMS screens requires a significant amount of effort and resources to establish consistent experimental conditions. Also, a non-trivial number of costly sequencing-based experiments are required across control and variables for the results to be statistically sound and meaningful. Furthermore, experimental validation of results are ultimately important for our ability to confidently generate hypotheses stemming from these datasets. As stated above, the first author of this paper, who was the primary driver of this work, is no longer in my lab or in academia. As for myself, I am in the middle of a transition out of academia and am actively ramping down my lab at UCSF. I no longer have the space or appropriate set-up to support this longer-term effort.

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

Learn more at Review Commons

---

Referee #3

Evidence, reproducibility and clarity

This paper by Radkov et al. represents an extensive structure-function-evolution treatment of the Type VI secretion effector protein Tae1. Using mutational scanning, the authors identify multiple residues that either enhance or reduce Tae1 function in an E. coli model, and validate these residues through direct functional assays. The main conclusion is that Tae1 contains a surprising number of non-intuitive residues important for its activity, particularly several surface-exposed residues far from the active site. The authors then suggest that these residues mediate binding to specific PG architectures and supply some evidence that the functional mutation landscape changes when the DMS assays is repeated in E. coli with altered cell wall architecture. Lastly, natural variants of Tsae1 are identified and discussed in the context of the trade-off between optimal toxicity and maintenance of self-immunity.

I have no major comments. The study is beautifully-done, with all controls in place. It might be worth following up on their putative PG binding residue mutants with an additional binding assay (MST, or just a crude cell wall pulldown assay), but that is not critical to support the main conclusions.

Minor comments

• The Western Blot of the vector control in Fig. 4F has the same impurities as the one in Fig. 5 B. Was the control blot re-used? If so, please indicate in the figure legend. Also, please show full Western Blots in supplemental material.
• Small typo in Fig. 5 legend ("does is not")
• The citation in line 108 seems a little off - that does not seem to support a physiologically relevant context for Tae.
• Line 122/123 - something seems to be missing in this sentence.
• Line 148 - this is not clear to me. Did they sequence plasmid barcodes (are those in the plasmid backbone?), or the mutated orfs?

Significance

This paper makes an impactful contribution to the open question of substrate- and species-specificity of PG hydrolases, particularly those weaponized by Type VI secretion systems. The major advance here is that PG binding by the hydrolase, and PG architecture of the substrate, are important determinants of Tae function and that this has important evolutionary consequences. The study will be of interest to the Type VI secretion community, but also to the PG turnover field.

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

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

Summary:

This manuscript aims to investigate the molecular basis underlying the differential toxicity of the bacterial T6SS amidase effector (Tae) by using Pseudomonas aeruginosa Tae1 as a model. The rationale is that while Tae is a conserved T6SS toxin degrading bacterial peptidoglycan (PG) by specific cleavage activity, different Tae toxins of the same family exhibit distinct lysis/antibacterial activity. Thus, the authors used combination of a deep mutagenesis scanning (DMS) coupled with fitness assay, NMR, and PG-binding/amidase activity to address this question by expressing Tae1 variants in E. coli. Besides finding the residues at/near the catalytic site critical for amidase activity, the authors further discovered many surface-exposed resides distant from the catalytic cleft also contributed to Tae activity likely by affecting binding or hydrolysis of PG. The authors further explored whether the residues contributing to loss or gain of Tae1 activity could be different against different PG structure by performing the same suite of DMS analyses from E. coli grown in the presence of D-Met, which resulted in reduced PG density and crosslinks. They discovered the fitness landscape of Tae1 variants shift dramatically, suggesting that Tae1 toxicity is highly context-dependent and optimizable for specific PG forms. A hyperactive Tae1 C110S variant is also naturally encoded in a subset of Proteobacteria outside of P. aeruginosa. This further led to a prediction that Tae1 C110S variant may evade binding and inhibition by cognate immunity, which was confirmed by the higher binding affinity of WT Tae1 than C110S Tae1 determined by ITC analysis. Together, the authors concluded that substrate-specificity and toxin-immunity interactions are the two distinct selective pressures for shaping diversity across the Tae1 toxin superfamily.

Major comments:

This is a well thought, carefully designed and executed research article reporting important and interesting findings. The conclusions made are mostly supported by the provided data. However, the toxicity assay for Tae1 variants except C110S was only validated by ectopic expression in E. coli or in vitro activity assay. Considering Tae1 is a bacterial toxin involved in interbacterial antagonism, the mutants with newly discovered key residues contributing to loss or gain of function shall be also evaluated for their role in the context of interbacterial competition, not simply by the cell lysis assay of expressed Tae1 variants in E. coli. Below are the specific comments that shall be addressed in order to claim the findings of this work .

1. It is an exciting finding that several surface residues distal from catalytic core mediate PG hydrolysis or binding. While the validation of their cell lysis activity by expressing each Tae1 variants fused with LepB signal peptide is informative, the role of these surface residues in toxicity shall be also tested by interbacterial competition assay either using E. coli or susceptible Pseudomonas aeruginosa strain as a prey. Tai may be expressed in E. coli prey to determine its neutralization activity during interbacterial competition context.
2. Based on the results that fitness landscape of Tae1 variants grown in the presence or absence of D-Met, the authors stated in line 334 "Condition-specific phenotypes suggests that Tae1 toxicity in vivo is highly context-dependent and optimizable for specific PG forms." However, there could be other physiological changes due to D-methionine. To claim this, the authors may test the surface residues with altered impacts on fitness between two growth conditions for their PG-binding activity using PG isolated from culture in the presence or absence of D-Met .
3. Quality of western blotting for Tae1 variants in Fig. 4F, 5B should be improved as the signals from WT is not clearly detected for comparison. The authors may use higher percentage of SDS-PAGE for better resolution of small Tae1 proteins. Relevant protein marker should be indicated. In addition, why there is no western blot analysis of C30A variant?
4. It is exciting that a hyperactive Tae1 C110S variant is also naturally encoded in a subset of Proteobacteria outside of P. aeruginosa. The authors showed higher binding affinity of WT Tae1 than C110S Tae1, which correlated with lower fitness of C110S variant in a competition setup (Fig. 6C, 6E). The authors suggest that "Tae1 of C110S variant lyses kin cells at a faster rate than Tae1 WT can bind and inhibit killing, leading to a fitness cost for this strain" (Line 460-463). To claim this, expression levels of endogenous Tae1 of both WT and C110S should be shown as well as their secretion levels to rule out the effect of protein abundance and secretion levels may affect the fitness. It would be also recommended to set up a real interbacterial competition assay by selecting the survival cfu of prey cells.

Minor comments:

1. Is Tae1 previously named as Tse1? Please clarify and indicate the previous name and accession number. As stated in Line 61" Although many T6S bacteria deploy similar toxins, interbacterial outcomes can vary considerably depending on the bacterial species engaged in T6S-mediated competition", the authors should also cite other relevant references showing differential Tae toxicity from different organisms (such as Serratia marcescens. Ssp1 and Ssp2 from English et al., 2012, Enterobacter cloacae Tae4 from Zhang et al., 2012, and Agrobacterium tumefaciens Tae from Yu et al., 2021). The manuscript shall gain more insights by discussing the biological significance of their conservation yet distinct toxicity and potential condition-specific activity of Tae toxins studied in different bacterial lineage besides those in P. aeruginosa.
2. Line 111-113 "the Tae1 protein from P. aeruginosa, which is injected into E. coli and leads to cell lysis": citations are needed here.
3. The heatmap in Fig. 4E also include those with mixed phenotypes. Are the averaged fitness score meaningful since some residues are likely derived from the mixed phenotypes, which make the data less reliable. I suggest the authors to only include those true GOF or indicate which one is true GOF and which one is from mixed phenotypes.

Significance

This manuscript used innovative approaches to investigate the mechanism and biological significance underlying the differential toxicity of the bacterial T6SS amidase effector (Tae). Tae superfamily can be classified into four families (Tae1-4), which are universally encoded in T6SS of diverse Proteobacteria. It is intriguing that Tae toxins classified in the same family produced by different bacterial lineage/strains exhibit distinct lysis/antibacterial activity but the underlying mechanism is unknown. This manuscript provided evidence suggesting the existing natural diversity of Tae1 in substrate-specificity and toxin-immunity interactions, which are the key selective pressures for shaping diversity across the Tae1 toxin superfamily. The findings provide an explanation how bacterial toxin effectors evolve in the context of interbacterial antagonism, which have not been answered from previous literatures (Russell et al., 2012; English et al., 2012; Chou et al., 2012; Zhang et al., 2013; Yu et al., 2021). The methods combining deep mutagenesis scan, biochemical, and structural analysis provide a comprehensive and unbiased view to understand the diversity of Tae1 family and their corresponding phenotypes and biochemical features. As a molecular microbiologist working on bacterial secretion systems and their effectors not familiar with structural studies, I am better qualified in evaluating the biological and biochemical data but not the structural studies in NMR and structural modeling. However, I highly appreciate the authors who nicely presented the story by explaining the concept of each method which allows the reviewers/readers to understand the data even though not within their expertise.

Russell et al., Cell Host Microbe 11:538-549. https://doi.org/10.1016/j.chom.2012.04.007 English et al., Mol Microbiol 86:921-936. https://doi.org/10.1111/mmi.12028. Chou, S. et al. Cell Reports 1, 656-664. DOI: 10.1016/j.celrep.2012.05.016 Zhang et al., J Biol Chem 288:5928-5939. https://doi.org/10.1074/jbc.M112.434357 Yu et al., J Bacteriol 203:e00490-20. https://doi.org/10.1128/JB.00490-20.

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

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

In this paper the authors characterize a member of the bacterial T6SS amidase effector (Tae) superfamily of toxins that are delivered by the Type 6 Secretion system of Pseudomonas aeruginosa into target prey bacterial cells. The authors focused on why this toxic effector because it shows different potency when delivered to different target species despite the fact that all target species have peptidoglycan, the substrate that Tae attacks. The authors use powerful approaches such as deep mutational scanning (DMS), to define critical residues near the Tae active site and other sites that affect its enzymatic activity and interaction with its cognate immunity protein. The discovery of the C110S mutation which increases Tae activity is a fine example of the power of this approach. When combined with structural biological analysis, the results of the study and discussion in the manuscript is of broad interest to the community of scientists interested in toxic bacterial effectors that digest the cell wall and also others that are interested in the remodeling of peptidoglycan during cell growth and shape determination. I would recommend acceptance of this paper for publication after the authors address a few minor comments:

1. The authors observe PG changes caused by D-met (Fig 4 and relevant text). I'm curious as to whether changes in lysis are caused by differences in PG crosslinking or PG density. They point out that the sugar binding surface of WT could localize the PG digestion (paragraph at line 521) which would no longer be required at lower density PG. However, my concern is that they propose that variation in Tae activity in different target organisms could be explained by differences in PG affinity without testing this DMS screen in any other strain, let alone species.
2. They also don't screen any Tae1 homologs, though they address one residue in their phylogeny. I'm not sure if there are species with such a low-density PG layer, so their repeated connection to Tae's variable lytic capacity between species in the text and discussion seems tenuous until they do a DMS screen with their plasmid library in another species (or at least another strain).
3. If WT Tae1 has some checking mechanism to ensure it's in the PG layer, I can imagine it might be slower to fire in less dense PG. That would also make sense given the chemical perturbations in Fig 3 where residues on the opposite face from the catalytic site are involved in binding PG. I would be interested to see if WT Tae1 can bind multiple PG chains or binds at higher affinity. A calorimetry approach like the one they use later may answer those questions, but that might be outside the range of this paper.
4. It's also worth looking around for E. coli PG synthesis mutants that don't change the PG polymer structure, only the density. That might also happen if the bugs are grown under osmotic stress, which should at the very least stretch out the sacculus. That may help differentiate differences in PG composition from differences in chain density. Perhaps subinhibitory amounts of drugs that affect PG synthesis my have the same of effect of increasing or even decreasing Tae potency by modulating PG density. Of course, this may have to be done under protective osmotic conditions. Have the authors tried these sorts of experiments and if so, please comment on the trends even if the data will not be presented in this report.

Significance

When combined with structural biological analysis, the results of the study and discussion in the manuscript is of broad interest to the community of scientists interested in toxic bacterial effectors that digest the cell wall and also others that are interested in the remodeling of peptidoglycan during cell growth and shape determination.

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

Learn more at Review Commons

---

Reply to the reviewers

1. General Statements [optional]

We would like to thank the reviewers for their prompt and thoughtful input on our manuscript, and their willingness to participate in more portable review through ReviewCommons.

2. Description of the planned revisions

Reviewer #1, major comments:

• A major concern is that the data are reported and analyzed on a per tomogram basis when many tomograms contain multiple mitochondria. Given that the mitochondria appear mostly well separated in Sup. Fig 1 with only a few connections visible, and the high degree of pleomorphism noted by the authors, I would strongly suggest that the authors use each mitochondrion as the basis for reporting their metrics rather than the FOV/tomogram as this would avoid mixing metrics from different mitochondria that may be in different states (e.g., fusion/fission). This would apply to data shown in Figures 3, 4, 5, and 6.

We appreciate the reviewer suggestion to separate on a per mitochondrion vs per tomogram basis for our analysis. While we do not anticipate that this will significantly change the overall findings, we agree that splitting per mitochondrion will account for any possible variability between mitochondria within the given field of view. Furthermore, we anticipate that this will actually improve our analysis and statistical power by effectively increasing the total sample size per experimental group. For our next revision, we will divide surfaces on a per mitochondrion basis within a given tomogram, and re-run the full analysis pipeline. Additionally, per reviewer request, we will include an output histogram for each measurement per mitochondrion surface in a supplemental figure.

• In Figure 3C the authors show the combined distribution of OMM-IMM distances within each condition. This may obscure some variability within populations. Individual histograms for all mitochondria should be included as supplementary material. Currently, it is difficult to judge if the peak of the combined distribution is appropriate and impossible to judge the variability between tomograms (preferably mitochondria, see above comment). Additionally, the shape of the distributions appears significantly different between conditions, suggesting that selecting a single peak value as representative and the basis for the statistical tests (Fig 3D) might not be appropriate. Please comment.

We will include individual histograms for each measurement per mitochondrion surface in a supplemental figure.

We agree that peak-based statistical tests limit our ability to quantify more complex differences, and this is why we chose to output histograms in addition to violin plots, so that shape differences can be observed qualitatively. A major challenge of shape-based statistical quantification is the assessment of independent samples. By using peak-based quantification, we could assume that each tomogram (and in the planned revision, each mitochondrion) is an independent sample, but for shape distribution this is inappropriate since there is more than one value represented per tomogram. Running a KS test with N equal to the number of tomograms yields no significance even in the visible cases where the shape appears very different.

However, the number of triangles also poorly represents the number of independent samples, since 1) the number of triangles used to represent a surface is somewhat arbitrary and remeshing can change it dramatically and 2) Our chosen triangle size is considerably smaller than the visually observed feature size in order to allow effective vector voting in the pycurv AVV algorithm. The result of this is that when we use a KS test on the distribution of values per triangle, even visually identical distributions yield p-values below 10^-200.

We do estimate the approximate smallest feature size during our calculations, since that is used to generate the radius used by pycurv in vector voting, to be 12 nm (the radius hit parameter in pycurv). During a public presentation of this work an audience member suggested that we might use the area implied by this feature size (~450 nm^2) as the size of an independent sample. This would yield around 1000 independent samples per tomogram. Because the choice of feature size is heuristic and manual, this is not as statistically sound as the peak-based metric, which is why we believe that the more conservative peak-based statistical testing is the gold standard for proving differences, but we believe this will be the most reliable way to quantify differences in shape of distributions. We plan to implement this quantification in our revision, and will evaluate whether it gives “expected” statistical results by a bootstrapping approach using subsampling of triangles from the same vs different mitochondria.

We would welcome reviewer suggestions for additional shape-based metrics and will explore other potential metrics to capture shape as part of our revision. While our peak-based metrics demonstrate our ability to statistically capture small changes in ultrastructure with this method, shape-based quantification will significantly enhance the capability to capture finer changes in structure that may be critical to understand physiologically.

Once this additional testing is complete, we will add a section to the results section describing choice of statistical framework. We also plan to generate a supplementary table showing the results of the peak-based quantification alongside all shape-based quantifications.

• In Figure 4C-F, again combined distributions are shown. Authors should include individual histograms for all mitochondria as supplementary material. The diversity of distributions in the metrics are more pronounced than the distances in reported in Fig 3, again making assessment of variability difficult and raising doubt about using the single peak value.

We will include individual histograms for each measurement per mitochondrion surface in a supplemental figure.

As we describe above, we will make test several options for distribution-based statistical quantifications and incorporate the results in the manuscript. We expect them to be useful for every measurement we make.

• It would be helpful to include the curvature or curvedness of the OMM for each mitochondrion in the supplementary material. The data to correlate OMM curvature with elongated/fragmented mitochondria should be available and might be of interest to some readers.

We will calculate curvedness of the OMM for each mitochondrion and include these data in the supplemental material. The inverse of the curvedness of the OMM gives a reasonable approximation of the radius of the mitochondrial “tube”, a feature which can be challenging to quantify fully automatically, and we agree that this may be of particular interest to some of our readers – particularly if morphology changes or stress-driven changes alter that radius in a statistically significant way!

Reviewer #1, minor comments:

• For all data, exact n per condition should be given (in text and captions as appropriate), not a range for the whole set.

We will report the exact n per condition in text and in captions after we separate our data on a per mitochondrion basis and update the analysis.

• Fig 5E middle, legend obscures some of the data.

We will reformat the graph such that the legend does not obscure the data after we separate our data on a per mitochondrion basis and update the analysis.

Reviewer #2, major comments:

Barad, Medina et al. presents a new toolkit for the analysis of membrane ultrastructure in cryo-tomograms. More specifically, the toolkit is designed to compare curvature, angles and spacing between different membrane types in mitochondria. These analyses allow for the quantitative comparison of membrane features e.g. for different growth conditions. To demonstrate the utility of the toolkit tomogram datasets of mitochondria in the presence and absence of ER stress were analyzed. The authors conclude that ER stress affects mitochondria morphology through remodeling of the membrane structure. The presented biological results and statistics are convincing and show active mitochondrial membrane remodeling in the cell when exposed to ER stress. It is also clear that there is a need for more quantitative evaluation based on the wealth of tomographic image features and mitochondrial membranes are certainly a well-chosen application. For this purpose, the authors developed a new workflow even though most of the discussed analyses are very specific to mitochondrial structures. Therefore, broader applications of these tools to other organelles are not easily envisaged without significant adaption. In that context, the title and abstract overpromise a much more powerful utility that can be applied to any other membrane analysis. Rather it seems that the proposed workflow is more of a specific tool or a pipeline for mitochondrial inner and outer membrane analysis instead of a toolkit for general morphological analysis. Hence, the manuscript cannot be accepted in its current form. In particular, the structure needs a significant rework of editing to become more comprehensible.

We appreciate the criticism that our workflow as implemented at the time of preprint is seemingly too focused on mitochondrial membranes and is not general. We’ve overhauled our workflow into a configurable (through a project YML file) scripted workflow that can take a folder with arbitrary segmentations and convert them into high quality meshes, followed by per-triangle quantification of the four primary metrics we describe in the manuscript: inter-membrane distance, intra-membrane through-space distance, curvature, and orientation. Generating fully automated visualization tools is more challenging, because which quantities are measured and how they are sub-classified (e.g., as we did for cristae, junctions, and IBM) is very project-specific; however, we did convert our visualization script into a library of utilities to combine tomograms into experiment objects, with methods to serialize for rapid access and functions for generating statistics and plots. Our converted visualizations script has been reorganized to act as an example of how similar questions could be asked for arbitrary membranes.

We propose to further demonstrate the generality of this updated approach by segmenting several examples of another organelle, the autophagosome, found in our dataset and applying the workflow to them in a supplementary figure.

The focussing to a method paper will also require more in-depth descriptions of the methodology in the main text. Although the code is deposited at github, there is no script-based workflow and description presented in the manuscript. Although Figure 1 puts the work into context of tomography, it remains very superficial on the image analysis. What are the input and output formats required for each step to follow the sequence of the workflow and at which steps critical interactive input is needed? What are the hardware requirements (CPU, GPU) or performance characteristics (CPU hours for certain operations)?

In addition to the changes mentioned above, we also added a “Supplemental Table 1” detailing computational requirements and time for each step.

We expanded on the description of this approach in the first paragraph of the results section:

“With this strategy, we were able to segment 32 tomograms containing mitochondria, divided between the elongated and fragmented bulk morphology populations and the two treatment groups (Figure 2, Supplementary Figure 1). The segmentation output was fed into the fully automated surface morphometrics pipeline (Figure 2B, Supplementary Figure 2, Supplementary Table 1). The voxel segmentation was converted to high quality membrane surfaces using the screened poisson algorithm32. Next, these surfaces were converted into triangle graphs and curvedness was estimated using pycurv15, and the distances within and between surfaces as well as the relative orientations of different surfaces were estimated using the resulting graph. Finally, the quantifications for each tomogram were combined into experiments to allow aggregate statistics and visualizations. This 3D surface morphometrics pipeline is configurable for any segmented membrane and is available at https://github.com/grotjahnlab/surface_morphometrics.”

We added a description of the up to date workflow in the methods section:

“Software workflow

The surface morphometrics pipeline is a python 3 scripted workflow with requirements that can be installed as a conda environment contained in an environment.yml file. The workflow is fully scripted and configurable with a config.yml file, and is run in 3 steps, with statistical analysis and visualization as an optional fourth step. First, a segmentation MRC file is converted automatically to a series of surface meshes formatted in the VTP file format. Second, for each mesh, the surface is converted to a graph (tg format) and curvature is estimated using pycurv. Third, orientations and distances between and within surfaces are calculated using the resulting graphs, and a CSV with quantifications as well as a final VTP surface file is output with all quantifications built in. Fourth, the outputs from multiple tomograms are combined for visualization and statistical analysis. Times and computational requirements are shown in supplementary table 1.”

Figures 3-7 contain colorful 3D renderings of the measured quantities. In addition, they are filled with histograms of every possible quantitative parameter, which often are not very significant or different between. The authors should focus the main results and the figures to show the most relevant and significant findings and put the remaining panels and results into the supplement.

Figures 3-7 were organized around the different methodologies (inter and intra-membrane spacing, curvature, orientation) but we agree that focusing to the main results of each methodology is sufficient to show the value of these results. We propose to address this criticism by moving figure 4D,F (inter-crista and junction spacing), figure 6 E,G (the junction measurements) and Figure 7 to supplemental figures. These supplemental figures will also be joined by the previously requested OMM curvature analysis and our proposed analysis of autophagosomes.

One of the key steps is the generation of a smooth surface from a segmented membrane, there is a question whether true membrane disruptions will be smoothed and may be overlooked in this approach. When these disruptions present true membrane ruptures, they may be of particular biological importance. The authors should support the choice and selection of the smoothing parameters in order to illustrate this potential pitfall.

The smoothing and hole-filling parameters are now configurable using the point_weight and extrapolation_voxels parameters in the config.yml file. Notably, the surfaces used for quantification used minimal smoothing, and any triangles more than a single voxel away from the point cloud were deleted, in order to ensure that the quantifications were minimally impacted by “hallucinated” surfaces. Additionally, the following text was added to the methods section discussion surface reconstruction:

“A surface mesh was calculated from the oriented point cloud using the screened Poisson algorithm32, with a reconstruction depth of 9, an interpolation weight of 0.7, and a minimum number of samples of 1.5. These settings were chosen to maximize correspondence to the data, rather than smoothness. The resulting surface extended beyond the segmented region, so triangles more than 1 voxel away from the point cloud were deleted. Interpolation weight (point_weight) and the mask distance (extrapolation_voxel) are both configurable in the surface morphometrics pipeline if more aggressive smoothing and hole filling are desirable.”

Throughout the manuscript, the authors mention statistical significance several times and one of the main aims of the study is perform statistical hypothesis testing. It is important to specify the significance test (not only in the methods) and the p-value in order to support this claim. In the manuscript, the authors use exclusively the Mann-Whitney test. What is the rationale for choosing this test? Have the authors considered comparing the total distributions and not just the peaks with e.g. a Kolmogorov-Smirnov test? For a statistical methods paper, there are also no discussion on error analysis.

This was a common concern raised by both reviewers, and we agree that a test based on total distribution would be more powerful than only looking at peaks. We address the use of the Kolmogorov-Smirnov test and the limitations we have run into thus far in our response to reviewer 1 in detail. In brief, KS tests tend to vastly overestimate statistical significance because the number of samples (the number of triangles) is vastly larger than the true number of independent features sampled in the data, so that even very similar looking distributions such as those in figure 5C yield p values in the range of 10^-200. We propose several approaches to better estimate the number of independent variables. We will also use a random subsampling approach within individual mitochondria to ensure sampling from the same distribution does not yield statistically significant results.

In addition to testing additional approaches to incorporate KS testing (based on estimation of number of independent features in each tomogram), we propose to improve our peak-based statistics by estimating a standard error for the peak of each tomogram using a bootstrap approach, getting the peaks from different random subsamples of triangles.

Reviewer #2, minor comments:

1. https://github.com/grotjahnlab/surface_morphometricsshould include an example data set or tutorial for dissemination.

We are in the process of uploading all frame-averaged tilt series, tomograms, segmentations, and reconstructed surfaces to EMPIAR. Additionally, we propose to implement a complete tutorial including a single tomogram for readier workflow testing, separate from the complete data upload.

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

Reviewer #1, major comments:

• As the work reported here is heavily computational, additional details about the computer hardware used and the time it took for the calculations to complete would be helpful for readers considering applying the code to their own data.

We appreciate the suggestion and included Supplementary Table 1 in the supplemental material outlining the computation time per step in our analysis pipeline:

“Supplemental Table 1. Approximate time and for each step of the surface morphometrics workflow.

Representative times and computational resources used for each step of the surface morphometrics workflow for each tomogram (unless otherwise noted) by the authors. Most time-intensive calculations were run in parallel on a compute cluster for each tomogram.

Step

Human Time (HH:MM)

Computational Wall Clock time (HH:MM)

CPU Cores Used

RAM Used

Automated initial segmentation (TomoSegMemTV)

00:10*

00:10*

8

64GB

Manual segmentation cleanup and classification

03:00

N/A

8

64GB

Point cloud conversion and mesh generation

00:01

00:03

4

16GB

Graph generation and curvature estimation (pycurv)

00:01

01:40

16

128GB

Distance and orientation measurement

00:01

00:10

16

128GB

Assembly of outputs from multiple tomograms into dataframes and serialization

00:01

00:10

1

16GB

Visualizations and statistical tests

00:01

00:10

1

16GB

* Tomosegmemtv is sometimes run iteratively with different settings to improve output. 10 minutes is approximately the time taken for a run without iteration, in the case of good output.”

Reviewer #1, minor comments:

• Pink and purple very close, consider alternative pair of colors or different shades to distinguish OMM and IMM

We kept OMM as purple but changed IMM to orange for Figure 3-7, and will make the associated changes to Figure 2 and Supplementary Movie 1 on final submission.

• Orientation of scaleboxes/scalebars should be consistent per figure panel. If knowledge of the axes is important to the reader, these should be included as well.

We followed the reviewer’s suggestion and updated the scale cubes to be standardized per panel.

• In the last sentence of the introduction, the term "organellar architectures" is used, instead of the previously defined "membrane ultrastructure." Consider changing for clarity.

We changed “organellar architectures” to “membrane ultrastructure” in the last sentence of the abstract.

• Inconsistent use of the phrase "cryo-electron tomography" after defining and using "cryo-ET"

We changed all instances of “cryo-electron tomography” to “cryo-ET” after defining in the first instance in the introduction.

• Authors argue that the distinction between curvedness and curvature is important and that curvature is less appropriate in this context, but then use curvature in the abstract, throughout introduction and in the results section. Usage can be improved for readability.

We changed all instances of “curvature” to “curvedness” throughout the text and figure legends.

• In section "Development of a framework to automate quantification of ultrastructural features of cellular membranes" the second last sentence should read "... higher quality membrane surfaces as compared..."

We changed “surface” to “surfaces” in text.

• In section "IMM curvedness is differentially sensitive to Tg treatment in elongated and fragmented mitochondrial networks" the fourth sentence should perhaps read "... despite apparent visual differences, no significant..."

We changed “difference” to “differences” in text.

• The term "cell's growth plane" is not clear from the text nor from Fig 6A. Do the authors mean surface of the substrate the cell is growing on?

We clarified and further defined the “cell’s growth plane” in the text by adding the following phrase:

“… the cell’s growth plane (i.e. the plane of electron microscopy grid substrate to which the cell is adhered) (Figure 6A).”

• In Materials and Methods:

• The authors report that manual back-blotting was used in a Vitrobot. This is non-standard usage and more details should be provided.

We added the following description to clarify our manual back-blotting procedure on the Vitrobot:

“After 8 hours of incubation, samples were plunge-frozen in a liquid ethane/propane mixture using a Vitrobot Mark 4 (Thermo Fisher Scientific). The Vitrobot was set to 37° C and 100% relative humidity and blotting was performed manually from the back side of grids using Whatman #1 filter paper strips through the Vitrobot humidity/temperature chamber side port. The Vitrobot settings used to disable automated blotting apparatus were as follows: Blot total: 0, 2; Blot force: 0, 3; Blot time: 0 seconds.”

• In section "Fluorescence Guided Milling" in the third sentence, the word "based" is repeated, second can be removed.

We deleted the second instance of “based” in this sentence.

• Symbol for degree (or the word degree) should be added to angular increment and tilt range for clarity.

Added degree symbols to the following sentence in the “Tilt Series Data Collection” portion of the materials and methods:

“Tilt series were acquired using SerialEM software (Mastronarde, 2005) with 2° steps between -60° and +60°.”

• Capitalization of TomoSegMemTV is inconsistent.

We changed all mentions to TomoSegMemTV.

• Fig 3 title - consider replacing "Inter-mitochondrial membrane..." with "Intra-mitochondrial membrane..." for clarity.

We clarified this point by changing “Inter-mitochondrial membrane distance” to “Distance between inner and outer mitochondrial membranes” in the figure legend:

“Figure 3. Distance between inner and outer mitochondrial membranes is dependent on mitochondrial network morphology and presence or absence of ER stress.”

• Fig 3C caption - should explicitly state it is a combined histogram and that the dashed lines correspond to the peak of the pooled data.

We changed “Quantification of” to “Combined histogram of” and added the sentence ” to each of the relevant figure captions (Fig. 3c, 4c-f, 5b-e, 6d-g, 7c):

“Dashed vertical lines correspond to peak histogram values of pooled data”

• Fig 6B and 6C caption - upper and lower parts not explicitly described.

We modified Fig 6B&C caption to more clearly describe the figure panel:

“(B) Two representative membrane surface reconstructions of lamellar Tg-treated elongated mitochondria, colored by angle of IMM relative to OMM.

(C) Two representative membrane surface reconstructions of a less rigidly oriented Tg-treated elongated mitochondria, colored by angle of IMM relative to the growth plane of the cell.”

Reviewer #2, major comments:

1. Title and abstract need to be toned down not to overpromise a very general toolkit. The presented method may be a tool or a collection of scripts - a toolkit can be used to address other types of (membrane) analysis problems. In the end, the analysis builds to a large extent on the previous developments and implementation of PyCurve. Perhaps, the most interesting contribution here is the application of the mesh generation by the Poisson reconstruction method to the segmented membranes, which is, however, well implemented in the used pymeshlab framework. The computation of distances and angles is straightforward.

We appreciate this critique and do not want to overpromise with our work, although we believe the overhaul to a fully configurable workflow addresses the primary concern. We are quite clear in the text that we build on top of pycurv, and recommend citation of the original tool as well as our pipeline in the github repository as a result. With that said,

We have changed the title as follows:

“Quantifying mitochondrial ultrastructure in cryo-electron tomography using a surface morphometrics pipeline”

We have also renamed our method to the surface morphometrics pipeline to reduce over-implication of generality, and made other small changes to increase degree of detail about what our method is resolving.

When reading the manuscript, the reader is left in the open whether this is a method paper or a biological results paper. The title/abstract suggests that this is a method paper and the manuscript is more of a mitochondrial membrane report in ER stress. Therefore, the title/abstract does not reflect the manuscript very well.

We aim to use this manuscript to describe the development of a workflow that enabled novel and interesting biological results. We adjusted the title to better match the combined development of a new pipeline and application to an interesting biological system as proof of concept:

“Quantifying mitochondrial ultrastructure in cryo-electron tomography using a surface morphometrics pipeline”

The manuscript also requires substantial structural editing. Several references to Figures are not appearing in the text in the order that the Figure panels are built. Excessive cross-referencing of figures also make the manuscript hard to read.

We simplified our referencing of figures and made sure the text matched the order of the figure panels.

The exact morphological discrimination between fragmented and elongated mitochondria is not easily understood from the results section. What is really meant by blinded manual classification? It only became clear when reading the methods. The results section should stand on its own. How is the overall population between fragmented and elongated cells is affected after Tg application?

To clarify our methodology for blinded classification of mitochondrial network morphologies we included the following text:

“We categorized cells for mitochondrial network morphology by blinded manual classification in which five researchers were given fluorescence microscopy images of exemplar network morphologies (elongated and fragmented) as references to assign morphologies to the experimental fluorescence micrographs.”

We targeted similar ratios of elongated and fragmented cells in both vehicle and Tg treated conditions for tomography, but qualitatively saw the expected increase in the elongated population to what has been previously described during Tg treatment. Because of our single cell targeting approach we did not quantify the population shift.”

Similarly, what is meant by manual classification of IMM, OMM and ER? Is there any clustering involved?

Our automated segmentation approach labels all membranes, and the separation of the IMM, OMM, and ER membranes is done by an expert user selecting and relabeling each membrane based on cellular context (e.g. IMM is inside of OMM and contains cristae). We have added the following text to clarify our methodology for manual classification of IMM, OMM, and ER:

“This was followed by manual labeling of membranes into mitochondrial IMM and OMM and ER membrane based on cellular context, as well as manual cleanup of individual membrane segmentations using AMIRA software (Thermo Fisher Scientific).”

Reviewer #2, minor comments:

What is meant by growth plane? This term is not defined in the manuscript.

We clarified and further defined the “cell’s growth plane” in the text by adding the following phrase:

“… the cell’s growth plane (the plane of electron microscopy grid substrate on which the cell is grown) (Figure 6A).”

What is meant by vehicle treatment? There is no explanation in the main text of the manuscript.

We clarified and further defined vehicle treatment in the main text by adding the following:

“We applied our correlative approach to identify and target specific Tg-treated and vehicle (media with DMSO) treated MEFmtGFP cells with either elongated or fragmented mitochondrial network morphologies for cryo-FIB milling and cryo-ET data acquisition and reconstruction.”

Have the authors noticed/calculated any differences in the width of the cristae?

We measure this difference in figure 4C (Intra-crista distance). We found significant changes in width/intra-crista distance in response to Tg treatment in both elongated and fragmented morphologies.

Methods: Automated surface reconstruction: "In cases where the resulting surface was very complex, the surface was simplified..." How was the complexity determined?

With the updated state of the software, we simplify all surfaces to generate a maximum of 150,000 triangles. This has minimal effect on very small surfaces, but greatly speeds computation on very large surfaces. We corrected the language to match this:

“The resulting mesh was simplified with quadric edge collapse decimation to produce a surface that represented the membrane with 150,000 triangles or fewer.”

Methods: Calculation of distances between individual surfaces: "For surfaces with small numbers of triangles, this was accomplished using a distance matrix...". What is the threshold for a small number of triangles?

As part of our software overhaul we have changed to always using a more memory-efficient KD tree based quantification, since the additional speed for the distance matrix approach is minimal when there are few enough triangles for it to be appropriate, and the hardwired cutoff was not as flexible for different hardware configurations. The updated text is below, but to satisfy any potential reviewer curiosity, the decision was made when the required distance matrix would use more than 128GB of memory. In the case of two identically sized surfaces, this crossover happens when there are approximately 45,000 triangles in each surface.

“For calculations of distances between respective surface meshes, the minimum distance from each triangle on one surface to the nearest triangle on the other surface was calculated using a KD-tree.”

Reviewer #2 (Significance (Required)):

The aim of the paper is well motivated. Cryo-ET is a growth field and there is a need for quantitative parameterization of cryo-ET data. Recently a toolkit for the analysis of filaments from cryo-ET has been published (Dimchev et al. 2021 DOI: 10.1016/j.jsb.2021.107808). Given the specific nature of the implementation, i.e. the membrane structures of mitochondria, I cannot easily see that this implementation will be useful beyond the analysis of mitochondrial membrane structure.

We hope that we have addressed this concern with generality has been addressed by our previously described updates to the software implementation.

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

Review 2, minor comments:

Angle between OMM and cristae: Maybe use the average angle of each cristae for comparison or fit a plane for each cristae because you are interested in the angle between the cristae and the OMM and the membrane of the cristae has a lot of uneven surfaces

We believe that the advantage of our approach is the ability to incorporate more complex geometric information from uneven surfaces such as those seen in cristae. With that said, the ability to quantify metrics for individual cristae in an automated manner would be very appealing, since in many ways cristae are functionally independent compartments. Accomplishing this would require either subdividing the larger surface into individual cristae, which will require development of additional sub-graph processing strategies. Additionally, pairing surfaces to represent opposite sides of a crista will require additional development. While we agree that this will be an excellent extension of the surface morphometrics approach, we feel that the additional development required is out of the scope of this initial manuscript focused on the general workflow. New methods leveraging sub-graph analysis will be explored in future manuscripts.

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

Barad, Medina et al. presents a new toolkit for the analysis of membrane ultrastructure in cryo-tomograms. More specifically, the toolkit is designed to compare curvature, angles and spacing between different membrane types in mitochondria. These analyses allow for the quantitative comparison of membrane features e.g. for different growth conditions. To demonstrate the utility of the toolkit tomogram datasets of mitochondria in the presence and absence of ER stress were analyzed. The authors conclude that ER stress affects mitochondria morphology through remodeling of the membrane structure. The presented biological results and statistics are convincing and show active mitochondrial membrane remodeling in the cell when exposed to ER stress. It is also clear that there is a need for more quantitative evaluation based on the wealth of tomographic image features and mitochondrial membranes are certainly a well-chosen application. For this purpose, the authors developed a new workflow even though most of the discussed analyses are very specific to mitochondrial structures. Therefore, broader applications of these tools to other organelles are not easily envisaged without significant adaption. In that context, the title and abstract overpromise a much more powerful utility that can be applied to any other membrane analysis. Rather it seems that the proposed workflow is more of a specific tool or a pipeline for mitochondrial inner and outer membrane analysis instead of a toolkit for general morphological analysis. Hence, the manuscript cannot be accepted in its current form. In particular, the structure needs a significant rework of editing to become more comprehensible.

Major comments:

1. Title and abstract need to be toned down not to overpromise a very general toolkit. The presented method may be a tool or a collection of scripts - a toolkit can be used to address other types of (membrane) analysis problems. In the end, the analysis builds to a large extent on the previous developments and implementation of PyCurve. Perhaps, the most interesting contribution here is the application of the mesh generation by the Poisson reconstruction method to the segmented membranes, which is, however, well implemented in the used pymeshlab framework. The computation of distances and angles is straightforward.
2. When reading the manuscript, the reader is left in the open whether this is a method paper or a biological results paper. The title/abstract suggests that this is a method paper and the manuscript is more of a mitochondrial membrane report in ER stress. Therefore, the title/abstract does not reflect the manuscript very well.
3. The manuscript also requires substantial structural editing. Several references to Figures are not appearing in the text in the order that the Figure panels are built. Excessive cross-referencing of figures also make the manuscript hard to read.
4. The focussing to a method paper will also require more in-depth descriptions of the methodology in the main text. Although the code is deposited at github, there is no script-based workflow and description presented in the manuscript. Although Figure 1 puts the work into context of tomography, it remains very superficial on the image analysis. What are the input and output formats required for each step to follow the sequence of the workflow and at which steps critical interactive input is needed? What are the hardware requirements (CPU, GPU) or performance characteristics (CPU hours for certain operations)?
5. Figures 3-7 contain colorful 3D renderings of the measured quantities. In addition, they are filled with histograms of every possible quantitative parameter, which often are not very significant or different between. The authors should focus the main results and the figures to show the most relevant and significant findings and put the remaining panels and results into the supplement.
6. The exact morphological discrimination between fragmented and elongated mitochondria is not easily understood from the results section. What is really meant by blinded manual classification? It only became clear when reading the methods. The results section should stand on its own. How is the overall population between fragmented and elongated cells is affected after Tg application?
7. Similarly, what is meant by manual classification of IMM, OMM and ER? Is there any clustering involved?
8. One of the key steps is the generation of a smooth surface from a segmented membrane, there is a question whether true membrane disruptions will be smoothed and may be overlooked in this approach. When these disruptions present true membrane ruptures, they may be of particular biological importance. The authors should support the choice and selection of the smoothing parameters in order to illustrate this potential pitfall.
9. Throughout the manuscript, the authors mention statistical significance several times and one of the main aims of the study is perform statistical hypothesis testing. It is important to specify the significance test (not only in the methods) and the p-value in order to support this claim. In the manuscript, the authors use exclusively the Mann-Whitney test. What is the rationale for choosing this test? Have the authors considered comparing the total distributions and not just the peaks with e.g. a Kolmogorov-Smirnov test? For a statistical methods paper, there are also no discussion on error analysis.

Minor comments:

1. https://github.com/grotjahnlab/surface_morphometrics should include an example data set or tutorial for dissemination.
2. What is meant by growth plane? This term is not defined in the manuscript.
3. What is meant by vehicle treatment? There is no explanation in the main text of the manuscript.
4. Angle between OMM and cristae: Maybe use the average angle of each cristae for comparison or fit a plane for each cristae because you are interested in the angle between the cristae and the OMM and the membrane of the cristae has a lot of uneven surfaces
5. Have the authors noticed/calculated any differences in the width of the cristae?
6. Methods: Automated surface reconstruction: "In cases where the resulting surface was very complex, the surface was simplified..." How was the complexity determined?
7. Methods: Calculation of distances between individual surfaces: "For surfaces with small numbers of triangles, this was accomplished using a distance matrix...". What is the threshold for a small number of triangles?

Significance

The aim of the paper is well motivated. Cryo-ET is a growth field and there is a need for quantitative parameterization of cryo-ET data. Recently a toolkit for the analysis of filaments from cryo-ET has been published (Dimchev et al. 2021 DOI: 10.1016/j.jsb.2021.107808). Given the specific nature of the implementation, i.e. the membrane structures of mitochondria, I cannot easily see that this implementation will be useful beyond the analysis of mitochondrial membrane structure.

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:

Barad and Medina, along with their co-authors, report on the development of a new software toolkit to quantitatively assess membrane structures that are observed in cryo-ET. This new toolkit builds upon existing methodologies by successfully incorporating additional methods and applying this to cryo-ET data to allow for more automated and reliable segmentations. This work addresses a long-standing difficulty in generating membrane segmentations, which are either done manually with huge labor investments or with automated methods that are known to be error prone. The authors demonstrate that their toolkit can generate high quality segmentations across multiple tomograms with limited manual intervention. They use correlative light and electron microscopy in combination with these segmentations to gain insight into the ultrastructural morphology of mitochondria within embryonic fibroblasts, both under control conditions and under endoplasmic reticulum stress induced by treatment with the drug Thapsigarin. Unlike changes to the ER which are more dramatic under stressed conditions, the changes to the mitochondria are more subtle and impossible to quantify without high quality segmentations. The authors show that inner and outer membrane distances change under stress, and that the distances between cristae, their junctions, and the angle of the cristae with respect to the margin of the mitochondria change. While they characterize the curvedness under the same set of conditions, they report no significant differences.

Major comments:

• A major concern is that the data are reported and analyzed on a per tomogram basis when many tomograms contain multiple mitochondria. Given that the mitochondria appear mostly well separated in Sup. Fig 1 with only a few connections visible, and the high degree of pleomorphism noted by the authors, I would strongly suggest that the authors use each mitochondrion as the basis for reporting their metrics rather than the FOV/tomogram as this would avoid mixing metrics from different mitochondria that may be in different states (e.g., fusion/fission). This would apply to data shown in Figures 3, 4, 5, and 6.
• In Figure 3C the authors show the combined distribution of OMM-IMM distances within each condition. This may obscure some variability within populations. Individual histograms for all mitochondria should be included as supplementary material. Currently, it is difficult to judge if the peak of the combined distribution is appropriate and impossible to judge the variability between tomograms (preferably mitochondria, see above comment). Additionally, the shape of the distributions appears significantly different between conditions, suggesting that selecting a single peak value as representative and the basis for the statistical tests (Fig 3D) might not be appropriate. Please comment.
• In Figure 4C-F, again combined distributions are shown. Authors should include individual histograms for all mitochondria as supplementary material. The diversity of distributions in the metrics are more pronounced than the distances in reported in Fig 3, again making assessment of variability difficult and raising doubt about using the single peak value.
• It would be helpful to include the curvature or curvedness of the OMM for each mitochondrion in the supplementary material. The data to correlate OMM curvature with elongated/fragmented mitochondria should be available and might be of interest to some readers.
• As the work reported here is heavily computational, additional details about the computer hardware used and the time it took for the calculations to complete would be helpful for readers considering applying the code to their own data.
• Discussion should be expanded to include a comparison of semi-automated segmentations generated here versus manual results from Navarro (Ref 35) & Burt (Ref 54 / doi: 10.1371/journal.pbio.3001319) and how one might estimate the error.

Minor comments:

• In the fourth sentence of the third paragraph of the introduction, Hoppe 1992 is cited as evidence of the limitations of work published in 2020, which is confusing. Perhaps the sentence can be re-phrased?
• Pink and purple very close, consider alternative pair of colors or different shades to distinguish OMM and IMM
• For all data, exact n per condition should be given (in text and captions as appropriate), not a range for the whole set.
• Orientation of scaleboxes/scalebars should be consistent per figure panel. If knowledge of the axes is important to the reader, these should be included as well.
• In the last sentence of the introduction, the term "organellar architectures" is used, instead of the previously defined "membrane ultrastructure." Consider changing for clarity.
• Inconsistent use of the phrase "cryo-electron tomography" after defining and using "cryo-ET"
• Authors argue that the distinction between curvedness and curvature is important and that curvature is less appropriate in this context, but then use curvature in the abstract, throughout introduction and in the results section. Usage can be improved for readability.
• In section "Development of a framework to automate quantification of ultrastructural features of cellular membranes" the second last sentence should read "... higher quality membrane surfaces as compared..."
• In section "IMM curvedness is differentially sensitive to Tg treatment in elongated and fragmented mitochondrial networks" the fourth sentence should perhaps read "... despite apparent visual differences, no significant..."
• The term "cell's growth plane" is not clear from the text nor from Fig 6A. Do the authors mean surface of the substrate the cell is growing on?
• In Materials and Methods:
  • The authors report that manual back-blotting was used in a Vitrobot. This is non-standard usage and more details should be provided.
  • The description of the Leica microscope is insufficient. The objective lens and camera used should be included.
  • In section "Fluorescence Guided Milling" in the third sentence, the word "based" is repeated, second can be removed. A second Pt coat on top of the GIS would also be unusual, please check writing for accuracy.
  • Symbol for degree (or the word degree) should be added to angular increment and tilt range for clarity.
  • Capitalization of TomoSegMemTV is inconsistent.
• Fig 1B: showing computational steps twice does not provide additional information. Consider just one example. Also, labels for elongated and fragmented would be more useful than the duplicated labels for each computational step.
• Fig 2A caption - should report actual thickness range measured (as given in Materials and Methods section) instead of estimated range.
• Fig 3 title - consider replacing "Inter-mitochondrial membrane..." with "Intra-mitochondrial membrane..." for clarity.
• Fig 3C caption - should explicitly state it is a combined histogram and that the dashed lines correspond to the peak of the pooled data.
• Fig 5E middle, legend obscures some of the data.
• Fig 6B and 6C caption - upper and lower parts not explicitly described.

Significance

This work primarily describes a technical advancement in methods to analyze cryo-ET data. The novelty arises from the combination of methods and their application rather than completely new ideas or approaches. Demonstrations of the utility of this toolkit based on the authors' analyses are convincing and will likely help a number of researchers in the field who are engaged in explorations of cellular ultrastructure and organelle responses to stimuli. Importantly, this work will help the field move past qualitative descriptions, historically accepted only because quantitative measurements at this level have not been feasible. Overall, in this reviewer's opinion, while the biological findings are modest, the utility of the toolkit for the field is indisputable and the work is of sufficient quality for publication.

My expertise is in cryo-EM, both single particle analysis and tomography, as well as CLEM workflows, applied mostly to cytoskeletal research and some ER stress. I do not have or strong background in mitochondrial biology nor sufficient computer science expertise to evaluate the numerical methods employed, but based on inspection of the github contents, the screened Poisson reconstruction algorithm is not reimplemented here.

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

Responses to reviewers’ comments are in blue text, original reviewers’ comments in black text.

Response to Reviewer 1.

Reviewer #1 (Evidence, reproducibility and clarity (Required)): In this manuscript Neiro et al. aim to expand our knowledge on the regulation of gene expression in stem cells of the planarian model organism. As a first step the authors used published available data to expand the repertoire of the planaria transcriptome. By combining 183 RNAseq datasets the authors were able to identify thousands of new coding and non-coding transcripts. They then screened for TF motifs in the new annotations, identifying 551 putative TFs, of which 248 were already described in the planarian literature. The most substantial contribution of this work to the field of stem cells and planaria biology is the characterization of new putative enhancers that were identified by performing H3K27ac ChIP-seq and ATAC-seq and combining these data with previously published H3K4me1 ChIPseq dataset.

We thank the reviewer for their careful assessment of our work, we agree that the identification of likely enhancers genome wide is a substantial contribution. Equally the improved annotation of all genes, including transcription factors we choose to focus on here, is a substantial step forward for the planarian research community.

By overlapping H3K27ac and H3K4me1the authors find 5,529 new enhancers, for which they report a higher chromatin accessibility than random points in the genome as assessed by ATAC-seq. By using ATAC-footprints Neiro et al. refined the subset of TFs that have binding motifs in the predicted enhancer-like regions and present a list of 22,489 such factors. The manuscript is well written and organized and overall, the reported data will provide an important resource to study gene expression regulation in planaria's stem cells. However, this manuscript would greatly benefit from some functional validation to support the predicted gene regulatory networks. One option would be to use a CRISPR-dCas9-KRAB system to silence the putative enhancers identified in the manuscript and check by qPCR the expression of nearby genes.

Currently mis-expression technologies, in order too directly test enhancer elements in driving expression, are still not available in planarians. This also preempts us using the suggested silencing system used in mammals and other animals with robust mis-expression tools.

If this type of experiment is not feasible in planaria (I am not an expert in this model organism) another simple but key experiment would be to perform a knockdown of one (or more) putative enhancer-bound TFs identified in this study followed by RNA-seq. This would allow the authors to verify what are the target genes of the putative enhancer-bound TFs and if they correspond to the predicted gene networks they identified. Simultaneously, this experiment would allow the authors to verify if there are any changes in the expression of differentiation/pluripotency markers as a result of the knockdown of the putative enhancer-bound TF.

These experiments are possible, but this would be the work of many labs in the future expert in studying those TFs and their roles in planarian stem cells and regeneration. However, what we can do is analyze existing RNA-seq data further. There are a number of studies where TF have been studied and RNA-seq performed after RNAi. Although these studies are performed in specific experimental regenerative contexts, and not specifically in stem cells, it will be possible to look at expression changes of genes with predicted enhancers bound by these TFs. We propose to execute this analysis and add it to the manuscript, rather than perform further TF RNAi experiments. This analysis is feasible within a 3-month revision time. We would add that currently their no genes are implicated in controlling pluripotency in the same way we might consider, for example, OSKM in mammals. Our identification of the TFs enriched in stem cell expression and implicated in binding predicted enhancers suggests future candidates.

Minor revision: • The authors have mostly focused on the identification of enhancer-bound TFs. However, it would be interesting to look at differential enrichment of TFs in promoters versus enhancers and identify if there are specific factors that are enriched specifically at the planarian newly identified enhancer regions.

We have not looked at potential TF binding sites near promoters/transcriptional start sites. We will try to add an analysis that considers this in our revision.

• All tornado plots are missing a colorbar (Fig3 and FigS2)

We will fix this error

• There is a typo in the discussion: "the combined use of chip-seq data, RNAi of a histone methyltransferase combines with chip-seq" should be changed to "combined".

We will fix this and other typographical errors.

Reviewer #1 (Significance (Required)):

The manuscript is well written and organized and overall the reported data will provide an important resource to study gene expression regulation in planaria's stem cells.

We thank the reviewer for their appreciation of our work

**Referees cross-commenting**

I agree with the other reviewers that additional functional data should be added to support the author's claims (such as knock down of potential TFs that are identified by computational analyses and assessing the impact on gene expression).

See response above, with regard to adding further analysis for testing this possibility.

In addition, as noticed by the third reviewer, all data should be made publicly available to the scientific community.

We have made all data publicly available and will submit all relevant data to public database repositories in advance of final publication after final peer review.

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

Summary:

This manuscript aims at identifying enhancers in the planarian Schmidtea mediterranea. The authors start with the integration of transcriptome with genome sequencing data to more precisely annotate the genome of the planarian Schmidtea mediterranea. The second part of the manuscript actually then deals with the identification of potentially active enhancer elements in adult stem cells of this regenerating organism using genomic techniques like ATAC-seq and ChIP-seq of histone marks combined with motif searches and in silico footprint analysis. Using these data, the authors predict regulatory interactions potentially critical for pluripotency and regeneration in planarian adult stem cells.

MAJOR COMMENTS:

• Are the key conclusions convincing? 1) The authors claim (already in the abstract) that their study identifies enhancers regulating adult stem cells and regenerative mechanisms. This is an over-statement found throughout the manuscript, as none of these enhancers are functionally tested nor is it shown that target gene expression changes when transcription factors predicted to interact with such enhancers are knocked down.

We agree and it was not our intention to overstate our results, this is why we have tried to refer to putative enhancers, enhancer-like elements etc in manuscript from the title onwards. Only once we have demonstrated a set of elements with key conserved and widely supported characteristics do we suggest we have a set of higher confidence enhancers to study. However, we will adjust the manuscript to reflect that our claims await direct testing as is the case for all enhancers implicated with the approaches used here.

Another example is at the end of paragraph 1 of section 2.4. Here the authors claim that identifying many fate-specific transcription factor genes in the vicinity of potential enhancers is a further proof that the identified regions represent "real enhancers". It strongly supports this hypothesis, but no evidence for real enhancer activity.

We agree the total body of evidence strongly supports that we have identified enhancer elements, but as above will adjust the language to suggest further directed functional work will follow from many groups.

Thus, although the authors state that the regulatory interactions and networks they predict from their data can be studied now in future, they should be more careful with their wording and correct these over-statements. Therefore, the key conclusion is that they identified by various techniques potential enhancers, which are close to genes controlling adult stem cells and potentially controlling these genes, which has to be shown by further analyses.

We agree

Thus, also the title needs to be changed.

We propose changing ‘enhancer-like’ to “predicted enhancers” in the title, and "defines" to "predicts" as well as broadly adjusting the text to caveat that further work will clarify their functions and roles.

The authors have no proof that the networks are active in planarian adult stem cells, as they do not show that the predicted networks are active in the presented way.

We agree, see comments above. It was not our attention to claim we are showing pathways that were definitely active, rather predicted by our experiments and analyses of the data from these experiments.

2) Similarly, the identification of TF motifs within these potential motifs strongly suggests but not shows that these factors are binding, even when these sites were found to be bound by a protein using the ATAC-seq footprinting analysis. Thus, the authors need to be careful with their wording. One example is in the second paragraph of section 2.5, where the authors write that "We found that numerous FSTFs were binding to putative intronic enhancers ... ". The motif suggests that these factors bind, however, they have no experimental confirmation that these sequences are indeed bound by the planarian TFs.

We agree. We will clarify that ATAC foot printing is the only data suggestive of these motifs being bound and that further experiments will be required for more evidence. We will state this in the section of results and add this explicitly to the discussion

In sum, this manuscript uses existing genomic tools to define potential enhancer regions in the planarian Schmidtea mediterranea. The manuscript is informative yet descriptive, as tit presents no functional evidence for any of the predictions. If further toned down, the key conclusions are valid.

Future functional experiments to test the roles of all TFs and enhancers is now possible due to our work.The combination of data and analyses provides strong support of enhancer elements activity in stem cells across the genome.

• Should the authors qualify some of their claims as preliminary or speculative, or remove them altogether? The experiments performed are well designed and in line with what is known in the field about enhancer architecture. However, as this model system is not very well characterized on that level and the authors do not provide real experimental evidence that any of the identified regions has really enhancer activity and that any of the identified motifs binds indeed the predicted TF, the authors need to be very careful with their statements. The authors should maybe emphasize even stronger that all the GRNs predicted under section 2.6 are really preliminary and need to be validated.

Yes, we are happy to be even clearer about this as the reviewer suggests

• 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. One experiment that could provide more evidence for their predicted regulatory interactions is to knock-down one of the FSTFs for which motifs have been identified in potential enhancer regions and to study expression of associated genes (to confirm that the enhancers potentilla bound by these TFs control the expression of associated genes) or by analyzing the chromatin status of selected chromatin regions (by Q-PCR). These experiments would strongly support the claims of the authors. However, it also depends strongly on the journal whether I would consider these experiments essential or "nice to have".

This suggestion of possible extra experiments is very similar to that of Reviewer 1. We are copying our earlier comment as this also addresses this point.

“These experiments are possible, but this would be the work of many labs in the future expert in studying those TFs and their roles in planarian stem cells and regeneration. However, what we can do is analyze existing RNA-seq data further. There are a number of studies where TF have been studied and RNA-seq performed after RNAi. Although these studies are performed in specific experimental regenerative contexts, and not specifically stem cells, it will be possible to look at expression changes of genes with predicted enhancers bound by these TFs. We propose to execute this analysis and add it to the manuscript, rather than perform further TF RNAi experiments. This analysis is feasible within a 3-month revision time. We would add that currently their no genes implicated in controlling pluripotency in the same way we might consider OSKM in mammals. Our identification of the TFs enriched in stem cell expression and implicated in binding predicted enhancers suggests future candidates.”

• 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. This reviewer is not an expert in Schmidtea mediterranea, thus it is hard to judge how time consuming these experiments would be. Cost-wise they should be feasible, as it would include primarily Q-PCR experiments. And some functional back-up of their claims would be very helpful.

See previous comment regarding additional analysis.

• Are the data and the methods presented in such a way that they can be reproduced? For the parts I can judge, yes.

• Are the experiments adequately replicated and statistical analysis adequate? It is not clear from the manuscript how many replicates of the ChIP-seq experiments were done.

Chip-Seq replicate data description will be explicitly added to the methods

MINOR COMMENTS:

• Specific experimental issues that are easily addressable.

• Are prior studies referenced appropriately? For the literature I can judge, yes.

• Are the text and figures clear and accurate? The figures are clear, the text (besides over-statements) is clear. However, the writing can be improved. A few examples: section 2.2 paragraph 1: "... we found 248 to be described in the planarian literature in some way." In which way described?; same paragraph: "... but significantly we could identify new homologs of ..." what does significantly mean? Which test etc? section 2.2, last paragraph: "Most TFs assigned to the X1 and Xins compartments and the least to the X2 compartment", "Very few TFs had expression in X1s and Xins to the exclusion of X2 expression as would be expected by overall lineage relationships"; what do these sentences mean?

We thank the reviewer for paying careful attention to the language in our manuscript throughout. We will provide clearer explanation of the sentences indicated. We will better explain terms specific to the planarian model system that are obviously not intuitive

. - Do you have suggestions that would help the authors improve the presentation of their data and conclusions? No over-statements.

See previous comments agreeing with the need to carefully adjust our language to avoid this

Reviewer #2 (Significance (Required)):

• Describe the nature and significance of the advance (e.g. conceptual, technical, clinical) for the field. This manuscript identifies genome-wide potential enhancers in adult planarian stem cells, and thus represents a very valuable resource for the community to study these enhancers and the gene regulatory networks they control in the future.

• Place the work in the context of the existing literature (provide references, where appropriate). As I am not a planarian scientist, it is hard to judge this part.

• State what audience might be interested in and influenced by the reported findings. In my opinion, this work will be primarily interesting for people working with planarian. When functional data exist, this might be also interesting for researchers working generally on regeneration.

Given the nature of our data we also think all groups working on animal stem cells would be interested in our data and analyses

• 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. My field of expertise is transcriptional regulation using genomic techniques, however I am not familiar with the model Schmidtea mediterranea.

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

Neiro et al. capitalize on existing genomic data for the planarian Schmidtea mediterranea and new ChIP-seq and ATAC-seq data to use computational approaches to identify putative enhancers in the planarian genome. They integrate analysis of enhancers with transcription factor binding sites to generate testable hypotheses for the regulatory function of transcription factors active in stem cells or control of cell lineage trajectories. Their work creates an excellent resource for future work to resolve the regulatory logic underpinning stem cell biology and tissue regeneration in planarians.

We are glad the reviewer likes our research.

Major: Overall, the work in this manuscript and methodology are well executed and presented. However, the authors should consider the following comments to improve the clarity and accessibility of the data and interpretations.

1) The new transcriptome does not appear to be publically accessible. The links to Github resources are broken, and there is nothing on Neiro's Github page. Will the new transcriptome be integrated with Planmine?

The new annotation has been available for over a year as we wished the community to have access to it ASAP (see Garcia Castro, 2021, Genome Biology https://doi.org/10.1186/s13059-021-02302-5). We tested the links in the paper before depositing our preprint and after review and they seemed to work for us both within and outside our institutional network. We can only apologize if they were broken or have not worked for the reviewer. We are unclear if this new annotation will be included in Planmine, but we will ask the colleagues maintaining this database to consider including it.

2) Figure 1: Ternary plot in 1F. The legend is not clear or could be explained better. What is the metric? It could be my misunderstanding, but I didn't consider the ternary plots as insightful or unnecessary. Perhaps the authors can expand on what they are showing.

These plots are important in demonstrating the distribution of mRNA expression of all genes across cell sorted compartments. Given the broad lineage relationship between sorted cell compartments This analysis allows us to identify genes expressed predominantly in one cell compartment or another, or across a specific transition. For example, genes enriched in X2 cells and Xins, but not X1 are likely to be enriched in post-mitotic differentiating progeny and differentiated cells. In contrast to single cell data where expression data can be sparse this analysis with bulk data allows identification and assignation of low expressed genes, like transcription factors. We will provide further explanation of this in the revised text.

1I is a map of exons, not alternative splicing. So, it isn't clear what the authors intend t show. Are the specific exons that are more likely to be spliced? Is the figure necessary?

We wish to demonstrate the power of annotation approach and the richness of the annotation for looking at alternate splicing. We propose to a more informative figure that indicates the variety of splice forms. We apologize for this oversight.

3) Figure 2: 2A labels Xins as irradiation responsive. Is this the case (just making sure)?

The reviewer is correct, this is wrong! This should read “irresponsive” or “irradiation resistant” In Figure 1A. We thank the reviewer for spotting this error. We will fix this.

2F-G: Ternary plot in F seems redundant with G, but that could be my lack of understanding. In 2G, what is represented on the plots on the right of the hierarchical clusters?

The ternary plot (2F) and heatmap of hierarchical clustering (2G) are complementary ways to visualize the proportional expression values of transcription factors. The ternary plot (2F) allows an overview of all the proportional expression values, while the heatmap (2G) shows how the proportional values may be grouped into clusters of similar expression profiles and displays the relative size of these clusters. For example, the heatmap shows that the clusters of X1 and Xins are more prominent than X2, suggesting that there are realtivey a few X2-specific transcription factors. We will add text to better to explain this difference.

4) Figure 3: The heat maps need a legend (i.e., please define the colors). In addition, labeling the figures could help the reader. For example, in G-J, a header about the different experiments above each map, such as "enhancers" and "random," etc., would make the figure more accessible.

We agree we label the figures to be more easily interpretable and provide an independent scale and legend for the heatmaps.

5) Figure 5: Although it is in the figure legend, the authors could label the 6th track as "RNA-seq in X1."

We will add this to the figure.

6) Section 2.6 second page last sentence of the first paragraph "GRN of asexual reproduction is not active in neoblasts" data in the supplement? Is it not shown?

We apologize for this poorly written sentence. In line with Reviewer 2s comments this statement needs to be toned down and clarified. The raw information is included in the general table of enhancers (Supplementary Table 2), but the genomic tracks visually highlighting the motifs at the promoters of lox5b and post2b were not included. We will add these to the Supplementary information and clarify Supplementary Table 2.

7) Discussion: The discussion about pluripotency factors in planarians could be expanded. The authors could contrast the study's findings with Önal et al. 2012.

We agree we will expand our discussion to compare with previous studies and also summarize what is available from other animals with pluripotent adult stem cells

Minor: The manuscript has no page numbers or line numbers, so I'll provide a general location of the potential issues.

1) Section 2 - newly identified isoforms are shorter (1656 vs. 1618). Is the order of the median length reversed?

Yes, we will correct this.

2) No mention of Figure S1B in the text.

It is mentioned in the paragraph regarding splicing, but perhaps not in a useful context. We will add a correct reference to this figure in the presentation of transcript diversity.

3) Figure 1H should be 1I in the text?

Yes, we will correct this

4) The discussion contains some minor typos and grammatical errors.

We will address with careful rereading.

We thank the reviewer for spotting these errors and we will fix them in revision.

Reviewer #3 (Significance (Required)):

Neiro et al. provide an excellent resource for the planarian community. The paper is generally very well written and easy to read. The new transcriptome described, which improves the annotation of the planarian genome, should be made readily available. It would be excellent if the transcriptome could be incorporated in Planmine.

We will ask Planmine and the Rink lab to consider this. The annotation (without broad analysis) has been available since the pre-print for Garcia Castro, 2021, Genome Biology was deposited in BioRxiv.

Furthermore, the authors provide a comprehensive list of transcription factors in the planarian Schmidtea mediterranea. Their work provides insight into which factors are highly expressed in the stem cell compartment. Their computational identification of transcription factors and putative enhancers will be helpful to the growing community of researchers studying stem cell and regenerative biology using planarians. In addition, the large dataset generated in this study could inform studies in the evolution of regulatory sequences and transcription factor function.

**Referees cross-commenting**

The data presented are well supported by previous studies. As noted by the authors, it is not possible to make transgenic planarians, and thus the field needs to rely on indirect methods. The authors focus on using the stem cell population, which can be isolated from the animals. Overall, I don't think additional experiments are necessary. Additional RNAi experiments combined with RNA-seq (using the stem cells) could take 6-12 months to complete. I believe this is a solid contribution that should be framed as a resource paper. The authors should pay close attention to Reviewer #2's suggestions and edit the paper accordingly.

I have 20 years of experience in the field. It would be unreasonable to ask the authors to do more experiments, especially in this post-pandemic environment. I hope this helps.

We thank the reviewer for the comments.

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

Learn more at Review Commons

---

Referee #3

Evidence, reproducibility and clarity

Neiro et al. capitalize on existing genomic data for the planarian Schmidtea mediterranea and new ChIP-seq and ATAC-seq data to use computational approaches to identify putative enhancers in the planarian genome. They integrate analysis of enhancers with transcription factor binding sites to generate testable hypotheses for the regulatory function of transcription factors active in stem cells or control of cell lineage trajectories. Their work creates an excellent resource for future work to resolve the regulatory logic underpinning stem cell biology and tissue regeneration in planarians.

Major:

Overall, the work in this manuscript and methodology are well executed and presented. However, the authors should consider the following comments to improve the clarity and accessibility of the data and interpretations.

1. The new transcriptome does not appear to be publically accessible. The links to Github resources are broken, and there is nothing on Neiro's Github page. Will the new transcriptome be integrated with Planmine?
2. Figure 1: Ternary plot in 1F. The legend is not clear or could be explained better. What is the metric? It could be my misunderstanding, but I didn't consider the ternary plots as insightful or unnecessary. Perhaps the authors can expand on what they are showing.

1I is a map of exons, not alternative splicing. So, it isn't clear what the authors intend t show. Are the specific exons that are more likely to be spliced? Is the figure necessary? 3. Figure 2: 2A labels Xins as irradiation responsive. Is this the case (just making sure)?

2F-G: Ternary plot in F seems redundant with G, but that could be my lack of understanding. In 2G, what is represented on the plots on the right of the hierarchical clusters? 4. Figure 3: The heat maps need a legend (i.e., please define the colors). In addition, labeling the figures could help the reader. For example, in G-J, a header about the different experiments above each map, such as "enhancers" and "random," etc., would make the figure more accessible. 5. Figure 5: Although it is in the figure legend, the authors could label the 6th track as "RNA-seq in X1." 6. Section 2.6 second page last sentence of the first paragraph "GRN of asexual reproduction is not active in neoblasts" data in the supplement? Is it not shown? 7. Discussion: The discussion about pluripotency factors in planarians could be expanded. The authors could contrast the study's findings with Önal et al. 2012.

Minor:

The manuscript has no page numbers or line numbers, so I'll provide a general location of the potential issues.

1. Section 2 - newly identified isoforms are shorter (1656 vs. 1618). Is the order of the median length reversed?
2. No mention of Figure S1B in the text.
3. Figure 1H should be 1I in the text?
4. The discussion contains some minor typos and grammatical errors.

Significance

Neiro et al. provide an excellent resource for the planarian community. The paper is generally very well written and easy to read. The new transcriptome described, which improves the annotation of the planarian genome, should be made readily available. It would be excellent if the transcriptome could be incorporated in Planmine.

Furthermore, the authors provide a comprehensive list of transcription factors in the planarian Schmidtea mediterranea. Their work provides insight into which factors are highly expressed in the stem cell compartment. Their computational identification of transcription factors and putative enhancers will be helpful to the growing community of researchers studying stem cell and regenerative biology using planarians. In addition, the large dataset generated in this study could inform studies in the evolution of regulatory sequences and transcription factor function.

Referees cross-commenting

The data presented are well supported by previous studies. As noted by the authors, it is not possible to make transgenic planarians, and thus the field needs to rely on indirect methods. The authors focus on using the stem cell population, which can be isolated from the animals. Overall, I don't think additional experiments are necessary. Additional RNAi experiments combined with RNA-seq (using the stem cells) could take 6-12 months to complete. I believe this is a solid contribution that should be framed as a resource paper. The authors should pay close attention to Reviewer #2's suggestions and edit the paper accordingly.

I have 20 years of experience in the field. It would be unreasonable to ask the authors to do more experiments, especially in this post-pandemic environment. I hope this helps.

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

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

Summary:

This manuscript aims at identifying enhancers in the planarian Schmidtea mediterranea. The authors start with the integration of transcriptome with genome sequencing data to more precisely annotate the genome of the planarian Schmidtea mediterranea. The second part of the manuscript actually then deals with the identification of potentially active enhancer elements in adult stem cells of this regenerating organism using genomic techniques like ATAC-seq and ChIP-seq of histone marks combined with motif searches and in silico footprint analysis. Using these data, the authors predict regulatory interactions potentially critical for pluripotency and regeneration in planarian adult stem cells.

Major comments:

• Are the key conclusions convincing?
  1. The authors claim (already in the abstract) that their study identifies enhancers regulating adult stem cells and regenerative mechanisms. This is an over-statement found throughout the manuscript, as none of these enhancers are functionally tested nor is it shown that target gene expression changes when transcription factors predicted to interact with such enhancers are knocked down. Another example is at the end of paragraph 1 of section 2.4. Here the authors claim that identifying many fate-specific transcription factor genes in the vicinity of potential enhancers is a further proof that the identified regions represent "real enhancers". It strongly supports this hypothesis, but no evidence for real enhancer activity. Thus, although the authors state that the regulatory interactions and networks they predict from their data can be studied now in future, they should be more careful with their wording and correct these over-statements. Therefore, the key conclusion is that they identified by various techniques potential enhancers, which are close to genes controlling adult stem cells and potentially controlling these genes, which has to be shown by further analyses. Thus, also the title needs to be changed. The authors have no proof that the networks are active in planarian adult stem cells, as they do not show that the predicted networks are active in the presented way.
  2. Similarly, the identification of TF motifs within these potential motifs strongly suggests but not shows that these factors are binding, even when these sites were found to be bound by a protein using the ATAC-seq footprinting analysis. Thus, the authors need to be careful with their wording. One example is in the second paragraph of section 2.5, where the authors write that "We found that numerous FSTFs were binding to putative intronic enhancers ... ". The motif suggests that these factors bind, however, they have no experimental confirmation that these sequences are indeed bound by the planarian TFs.

In sum, this manuscript uses existing genomic tools to define potential enhancer regions in the planarian Schmidtea mediterranea. The manuscript is informative yet descriptive, as tit presents no functional evidence for any of the predictions. If further toned down, the key conclusions are valid. - Should the authors qualify some of their claims as preliminary or speculative, or remove them altogether?

The experiments performed are well designed and in line with what is known in the field about enhancer architecture. However, as this model system is not very well characterized on that level and the authors do not provide real experimental evidence that any of the identified regions has really enhancer activity and that any of the identified motifs binds indeed the predicted TF, the authors need to be very careful with their statements. The authors should maybe emphasize even stronger that all the GRNs predicted under section 2.6 are really preliminary and need to be validated. - 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.

One experiment that could provide more evidence for their predicted regulatory interactions is to knock-down one of the FSTFs for which motifs have been identified in potential enhancer regions and to study expression of associated genes (to confirm that the enhancers potentilla bound by these TFs control the expression of associated genes) or by analyzing the chromatin status of selected chromatin regions (by Q-PCR). These experiments would strongly support the claims of the authors. However, it also depends strongly on the journal whether I would consider these experiments essential or "nice to have". - 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.

This reviewer is not an expert in Schmidtea mediterranea, thus it is hard to judge how time consuming these experiments would be. Cost-wise they should be feasible, as it would include primarily Q-PCR experiments. And some functional back-up of their claims would be very helpful. - Are the data and the methods presented in such a way that they can be reproduced?

For the parts I can judge, yes. - Are the experiments adequately replicated and statistical analysis adequate?

It is not clear from the manuscript how many replicates of the ChIP-seq experiments were done.

Minor comments:

• Specific experimental issues that are easily addressable.
• Are prior studies referenced appropriately?

For the literature I can judge, yes. - Are the text and figures clear and accurate?

The figures are clear, the text (besides over-statements) is clear. However, the writing can be improved. A few examples: section 2.2 paragraph 1: "... we found 248 to be described in the planarian literature in some way." In which way described?; same paragraph: "... but significantly we could identify new homologs of ..." what does significantly mean? Which test etc? section 2.2, last paragraph: "Most TFs assigned to the X1 and Xins compartments and the least to the X2 compartment", "Very few TFs had expression in X1s and Xins to the exclusion of X2 expression as would be expected by overall lineage relationships"; what do these sentences mean? - Do you have suggestions that would help the authors improve the presentation of their data and conclusions?

No over-statements.

Significance

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

This manuscript identifies genome-wide potential enhancers in adult planarian stem cells, and thus represents a very valuable resource for the community to study these enhancers and the gene regulatory networks they control in the future. - Place the work in the context of the existing literature (provide references, where appropriate).

As I am not a planarian scientist, it is hard to judge this part. - State what audience might be interested in and influenced by the reported findings.

In my opinion, this work will be primarily interesting for people working with planarian. When functional data exist, this might be also interesting for researchers working generally on regeneration. - 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.

My field of expertise is transcriptional regulation using genomic techniques, however I am not familiar with the model Schmidtea mediterranea.

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

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

In this manuscript Neiro et al. aim to expand our knowledge on the regulation of gene expression in stem cells of the planarian model organism.

As a first step the authors used published available data to expand the repertoire of the planaria transcriptome. By combining 183 RNAseq datasets the authors were able to identify thousands of new coding and non-coding transcripts. They then screened for TF motifs in the new annotations, identifying 551 putative TFs, of which 248 were already described in the planarian literature. The most substantial contribution of this work to the field of stem cells and planaria biology is the characterization of new putative enhancers that were identified by performing H3K27ac ChIP-seq and ATAC-seq and combining these data with previously published H3K4me1 ChIPseq dataset. By overlapping H3K27ac and H3K4me1the authors find 5,529 new enhancers, for which they report a higher chromatin accessibility than random points in the genome as assessed by ATAC-seq. By using ATAC-footprints Neiro et al. refined the subset of TFs that have binding motifs in the predicted enhancer-like regions and present a list of 22,489 such factors.

The manuscript is well written and organized and overall, the reported data will provide an important resource to study gene expression regulation in planaria's stem cells. However, this manuscript would greatly benefit from some functional validation to support the predicted gene regulatory networks. One option would be to use a CRISPR-dCas9-KRAB system to silence the putative enhancers identified in the manuscript and check by qPCR the expression of nearby genes.

If this type of experiment is not feasible in planaria (I am not an expert in this model organism) another simple but key experiment would be to perform a knockdown of one (or more) putative enhancer-bound TFs identified in this study followed by RNA-seq. This would allow the authors to verify what are the target genes of the putative enhancer-bound TFs and if they correspond to the predicted gene networks they identified. Simultaneously, this experiment would allow the authors to verify if there are any changes in the expression of differentiation/pluripotency markers as a result of the knockdown of the putative enhancer-bound TF.

Minor revision:

• The authors have mostly focused on the identification of enhancer-bound TFs. However, it would be interesting to look at differential enrichment of TFs in promoters versus enhancers and identify if there are specific factors that are enriched specifically at the planarian newly identified enhancer regions.
• All tornado plots are missing a colorbar (Fig3 and FigS2)
• There is a typo in the discussion: "the combined use of chip-seq data, RNAi of a histone methyltransferase combines with chip-seq" should be changed to "combined".

Significance

The manuscript is well written and organized and overall the reported data will provide an important resource to study gene expression regulation in planaria's stem cells.

Referees cross-commenting

I agree with the other reviewers that additional functional data should be added to support the author's claims (such as knock down of potential TFs that are identified by computational analyses and assessing the impact on gene expression). In addition, as noticed by the third reviewer, all data should be made publicly available to the scientific community.

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

Learn more at Review Commons

---

Reply to the reviewers

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

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

Learn more at Review Commons

---

Referee #3

Evidence, reproducibility and clarity

In this study the authors present the discovery of two new splicing factors NRDE2 and CCDC174 that interact with the U1 snRNA and with 5' splice sites and modulate usage of 5' splice sites with generally weaker pairing potential to the U1 snRNA. They develop a new cross-linking technique BCLIP for monitoring RNAs interacting with a particular protein by deep sequencing, which modifies the classical CLIP protocol and appears to allow them to detect interactions with proteins of low abundance such as NRDE2. They propose that NRDE2 and CCDC174 may form alternative U1 snRNP complexes distinct from the canonical U1 snRNP, which may be partly responsible for selecting alternative, weaker 5'SS.

The study provides a plethora of experiments that provide strong experimental support for a model in which NRDE2 interacts with the U1 snRNA, recruits CCDC174, and together they tend to promote correct usage of weak 5' splice sites that are often flanked by several weak, cryptic 5' splice sites. The RNA-Seq supports a genome-wide role for NRDE2 in promoting splicing of weaker 5'-splice sites, while the in vivo reporter assays are elegant experiments showing a role for NRDE2 in enforcing correct usage of the most upstream weak 5'splice site.

While the authors provide strong evidence in support of the main model proposed in their discussion, there are a few significant matters that are not addressed. Firstly, the fact that only a small proportion of 5'SS bound by NRDE2 appears to be sensitive to NRDE2 KO, even when translation-linked surveillance is blocked by cycloheximide, raises the possibility that the RNA-Seq technique may miss a significant proportion of transcripts that are unspliced and are rapidly degraded, potentially co-transcriptionally, by the nuclear exosome in a manner that may not necessarily depend on MTREX. Given that a significant proportion of unspliced transcripts may follow such a pathway (reviewed in Gordon et al., Curr. Op. Gen. and Dev. 2021), the authors should at least consider this possibility in their presentation of results and discussion. Ideally one could try to combine rapid depletion of NRDE2 with depletion or partial inactivation of one of the nuclear exosome components RRP6 or RRP44, although this reviewer recognizes that this may be technically challenging and lead to indirect effects on cell growth that might confound the analysis. Sequencing of specifically nascent RNAs associated with Pol II from a chromatin fraction, might offer a way to uncover additional NRDE2-sensitive transcripts.

Secondly, the fact that NRDE2D200 shows a massive increase in U1 snRNP reads by the BCLIP procedure potentially suggests that NRDE2 may actually be part of a surveillance pathway to enforce usage of specific 5'SS and minimise cryptic 5'SS use. In this model, NRDE2 might bind all 5'SS but needs to be dissociated from the U1 snRNP either before or during 5'SS transfer by a helicase (e.g. MTREX or DDX5) within a certain time frame to prevent transfer of cryptic 5'SS. This model would be reminiscent of the initial binding and subsequent dissociation of Mud2 by Sub2 during E complex formation in yeast or of U2AF by DEK during proofreading of initial 3'SS recognition in humans. The fact that targets sensitive to NRDE2, as judged by RNA-Seq and expression profiling, mostly do not overlap with those MTREX, does not exclude the possibility that NRDE2 may act in cooperation with MTREX to prevent usage of cryptic 5'SS, which might result in production of rapidly degraded unspliced transcripts that are not detectable by the RNA-Seq methodology used here. Minimally, this reviewer thinks this possibility needs to be considered and briefly discussed by the authors.

Finally, although provocative, the idea that NRDE2 binds an alternative U1 snRNP is not necessarily implied by the data. The fact that U1A, U1C, and U1-70k are not detected by MS in a tandem IP set-up that uses disrupts RNA structure and potentially protein-RNA interactions cannot be considered clear evidence for an alternative snRNP. Structures of the U1 snRNP suggest that association of such auxiliary proteins may depend on the structure of the U1 snRNA. The authors need to either modulate and clarify their claim, or provide stronger evidence, e.g. from more gentle IPs with U1A and U1-70K that U1 snRNPs that associate with these factors are not also associated with NRDE2. Related to this, this reviewer thinks it is tenuous to claim that the harsher xTAP-MS analysis involving formaldehyde is more indicative of "native" interactions because it uncovers binding of one core spliceosomal component and of TFIP11 and DHX15.In fact, the opposite seems more likely, that such reported interactions are not indicative of direct proximity, but rather result from perturbations to the native RNP structure induced by benzonase. In this sense, the claim that these observations suggest an "alternative" spliceosome assembly pathway seems particularly problematic, especially in view of the fact that in vitro studies suggest that DHX15 can associate with the U2 snRNP and can disassemble complexes at all stages of spliceosome assembly and catalysis, including during the pre-catalytic stage. The authors should be more careful with their wording and interpretation here.

Depending on how the authors address the issues raised above, and how they modify their claims in the text, additional experiments may be deemed beyond the scope of the present study and should not be strictly necessary for publication, with the likely exception of the issue of alternative U1 snRNPs, where additional IPs might clarify, and potentially strengthen, the authors' claims.

Significance

This reviewer is an expert in the biochemical and structural study of pre-mRNA splicing and considers the present study an important contribution to understanding 5' splice site usage in higher eukaryotes. While the observation that spliceosomes from higher eukaryotes use additional protein factors to modulate 5' splice site selection is not new, the discovery of specific factors bound to U1 snRNA that may directly affect its binding to stronger or weaker 5'SS is certainly novel and of potentially broader significance. Although the author's claim that this may reflect alternative U1 snRNPs is not fully supported by the evidence presented, the proposal itself is an important potential advance, if it holds up to more stringent testing. The potential for NRDE2 to be part of a more complex surveillance mechanism to enforce use of specific 5'SS, which the data may also point to, would be an equally important advance. Finally, the observed interactions with U4 and U6 snRNAs, on which the authors do not comment much, provide further support for the idea that transfer of the 5'SS from U1 to U6 may be a particularly crucial step in 5'SS selection that is modulated in higher eukaryotes by non-canonical factors. Indeed, this latter point is also a significant contribution of the present study.

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

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

In this manuscript, Flemr and colleagues describe novel functions of the splicing-associated proteins NRDE2 and CCDC174. These two proteins have previously been implicated in splicing (Jiao et al., RNA, 2018) and interact with the helicase Mtr4 and the exon-junction complex (Richard et al., RNA, 2018), which targets RNAs for degradation by the exosome. Here, the authors use a combination of genomics, including a modified CLIP protocol, genetics, and mass spectrometry to establish two key findings: (1) NRDE2 and CCDC174 act in concert to promote pre-mRNA splicing from non-consensus 5' splice sites (5'SSs) and (2) together with U1 snRNA they form a novel non-canonical U1 snRNP. The data themselves are clearly presented and of very high quality, but I do not agree with the authors interpretations and the two key claims.

On claim (1), instead of NRDE2/CCDC174 specifically and actively promoting the usage of correct weak 5'SSs, they could alternatively promote correct 5'SS choice indirectly, by suppressing nearby cryptic 5'SSs. This is consistent with known functions of the EJC (e.g. Boehm et al., Mol Cell, 2018), with which NRDE2 and CCDC174 interact (Jiao et al., RNA, 2018) and their known association with Mtr4, which is also re-produced in the manuscript. This is further supported by the authors data, which shows that NRDE2 and CCDC174 CLIP signal peaks at the same sites as EJCs, upstream of 5'SSs. Prior to publication the authors should experimentally distinguish between active (direct) and passive (indirect) 5'SS selection mechanisms by NRDE2 and CCDC174.

On claim (2), a new U1 snRNP would be a major discovery, yet, given the presented data, this conclusion should either be removed completely from the manuscript or needs to be rigorously tested. See comments below.

Major comments

1. The authors show an enrichment by MS-IP of NRDE2 in Fig 1A, 1C (the improvied xTAP-MS protocol) for late-stage spliceosome components, such as TFIP11 that is required for spliceosome disassembly (ILS complex), consistent with earlier data in C. elegans (Jiao et al., RNA, 2018). Given the consistency of the late stage-spliceosome interaction and the EJC with published results, how do the authors reconcile the proposed functions in 5'SS selection with known interactions of NRDE2 and CCDC174 with the EJC and disassembling spliceosomes? If NRDE2 and CCDC174-U1 formed, they would dissociate from the spliceosome with U1 snRNA during the Prp28-dependent pre-mRNA handover from U1 to U6 snRNA. How would NRDE2 and CCDC174 re-associate after the subsequent Pre-B to B to Bact to C transitions in C, when the EJC binds the spliceosome, or after the subsequent C* to P to ILS transitions in the ILS, when e.g. TFIP11 binds. In a more likely model, early and late splicing factors co-IP in the authors MS experiments because splicing factors are enriched generally with eachother and in e.g. nuclear speckles. Perhaps more stringent washes in the xTAP-MS experiment could home in on more direct interactions of NRDE2 or CCDC174 to the spliceosome?
2. The following comments relate to the claim of a non-canonical U1 snRNP.
   • Fig. 6B: to assess the predictive power of the CLIP signal to reveal protein-snRNA interactions, can the authors comment on the expected crosslink efficiency and specificity of a bona fide U1 snRNP protein to U1 or a U2/U4/U5/U6 snRNP protein to its respective snRNA as well as all other snRNAs? How would these efficiencies and specificities compare to NRDE2-U1?
   • Fig. 6C: Can the relative differences in snRNA abundance, U1 being the most abundant, explain the CLIP crosslink efficiencies without the requirement of a bona fide NRDE2-U1 complex?
   • In Fig. 6C, have the authors looked at other spliceosomal snRNAs and their enrichments in the northern?
   • Fig. 6G, did the authors measure cellular snRNA levels after SmE dTAG depletion? The prediction would be that all snRNAs are reduced in steady-state abundance, due to improper biogenesis, which could explain why the U1 snRNA CLIP-seq signal is reduced. This would be independent of an NRDE2-U1 interaction.
   • As it would be surprising and exciting, if U1A, U1C, and U1-70k were absent from a functional U1 snRNP, this requires additional proof. Can they authors use an anti-U1 snRNA oligo in tandem with the NRDE2 IP or CCDC174 IP to show that the Sm-ring proteins and U1 snRNA are highly enriched but not U1A, U1C, and U1-70k proteins or any other snRNA?
   • U1C provides a ZnF domain that stabilizes the pre-mRNA 5'SS in its binding to U1 snRNA (Kondo et al., Elife, 2015). U1-70k stabilizes the U1 snRNP (Kondo, Elife, 2015) and can couple to RNA polymerase II (Zhang et al., Science, 2021), and is important for U1 snRNP biogenesis (Byung Ran So et al., NSMB, 2016). How would NRDE2 or CCDC174 compensate for these essential activities? Given the various crucial functions known U1 proteins perform, the claim that NRDE2 or CCDC174 can substitute them, should be supported by proof of their functional substitution.
   • Since NRDE2 or CCDC174 and U1 snRNA would be conserved and presumably for a high-affinity complex, ideally the authors would provide biochemical proof of their interactions, though this is may be beyond the scope of the current manuscript.

Minor comments:

• The conditions of the dTAG experiment are insufficiently described, what was the efficiency of depletion (Western blots or mass spec?) and over which time-scale was this applied?
• Introduction, in the sentence '[...] encode much shorter U1 snRNA [...]' the authors imply that longer U1 snRNAs are correlated with a lack of splice site degeneracy. Yet, structural and mechanistic data show that the expanded U1 snRNA segments in e.g. S. cerevisiae (or U2 snRNA, which contains a 1000 nt) insertion are distant from the U1 snRNA 5'-end that recognizes the 5'SS or the U2 snRNA BSL, which binds the branch site, and thus are unlikely to influence splice site selection. Please rephrase.

Significance

NRDE2 and CCDC174 are enigmatic proteins that are likely to function during mRNA biogenesis and both have been linked to splicing and RNA decay. It is thus interesting to understand their precise modes of action. While the authors provide excellent data, the conclusions are not substantiated. I have expertise in the mechanistic study of pre-mRNA splicing, based on which several of the authors claims such as a new U1 snNRP complex, are challenging to reconcile with the past decades of splicing research. Given the sizable impact a new U1 snRNP would have on the field, these data must be unimpeachable.

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

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

In this manuscript, Flemr et al. characterize the roles that proteins Nrde2 and CCDC174 play in mammalian splicing regulation. The authors perform native (nTAP-MS) and cross-linking (xTAP-MS) based IP-MS methods to identify the NRDE2 and NRDE2 mutant interactomes, and demonstrate NRDE2's enriched interactions with splicing factors. The authors further develop a RNA footprinting method (BCLIPseq) in order to capture NRDE2 and CCDC174 binding patterns on RNA, revealing a preference for binding unspliced introns close to the 5' SS. Furthermore, upon NRDE2 knockdown, the authors note a significant increase in alternative 5'SS splicing events, making NRDE2 a putative regulator of cryptic 5'SS. Following up on this observation through luciferase reporter assays, the authors demonstrate how NRDE2 and CCDC174 work together to inhibit cryptic splicing at weak 5'SSs. Finally, the authors demonstrate that NRDE2 and CCDC174 interact with U1 snRNA (but not core U1 snRNP protein components), providing a basis for their interactions with 5'SS. Overall, the authors thoroughly characterize protein and RNA interactions with NRDE2, demonstrating its role in mammalian pre-mRNA splicing, and its concerted role with CCDC174 in regulating splicing at some weak 5' splice sites.

The study would be greatly improved through additional controls and more careful analyses. For instance, many controls are missing for the cell lines used throughout the study, making interpretation of the data more difficult. Other issues include that techniques developed within the study are presented without validation experiments, and key analyses, including microscopy experiments and rMATs analysis of RNAseq data, are performed without proper quantification, weakening the authors' conclusions from these experiments. Finally, major conclusions of the paper, such as the potential role of NRDE2 in a non-canonical U1 snRNP complex, would be greatly strengthened by additional experiments. However, overall, it appears that many of the major concerns should be readily addressable.

Major comments

1. Data are not present to demonstrate that cell lines were validated and compared properly:
   • a. The authors say "We assessed potential consequences of the tagging approach on protein function by comparing RNA-seq gene expression profiles with that of untagged cells, which remained unchanged for the proteins we report hereinafter" (p.6). This analysis should be made available in the supplement. They should additionally show western blots of tagged/untagged protein, in order to demonstrate similar expression levels for endogenous and engineered proteins.
   • b. Knockdown (KD) levels of all proteins from engineered lines should be shown over the KD timeline used in the study. For instance, no westerns in the paper show the degradation efficiency of the dTAG KD lines.
   • c. The authors should address why knockdown lines were made in different ways for different proteins (ex. only one cell line is a dTAG degradation line) and why knockdowns were performed for different amounts of time for every protein.
   • d. The authors should consider that, since knockdowns are performed for different amounts of time, results between protein knockdowns may not be directly comparable. For instance, in figure 3D, Ccdc174 dTAG lines have less misspliced target introns than the other knockdown lines. However, this may simply be because the knockdown period is shorter for the dTAG line than the other knockdown lines, and the length of treatment affects the overall number of introns affected.
2. Nuclear localization experiments would benefit from further controls and quantification:
   • a. The authors conclude that "NRDE2 localisation to nuclear speckles depends on active pre-mRNA splicing" (p.7), which seems to contradict their result that "Chemical inhibition of splicing with Thailanstatin A (Liu et al., 2013) resulted in...wild-type NRDE2 remaining concentrated in enlarged NSs (Figure 1H and S1J)" (p.8).
   • b. Since the splicing inhibitor Thailanstatin A also changes the localization patterns of U2AF2 (Fig. 1G-H), it is unclear if U2AF2 is still a reliable nuclear speckle marker in the presence of the drug. Additional controls (such as staining for other nuclear speckle markers) are necessary to make this assertion.
   • c. To make the conclusion that "NRDE2-D174R accumulated in nucleoli" (p. 8), the authors should also include a nucleoli marker in their microscopy experiments.
   • d. Signal quantification of NRDE2 distribution/overlap with U2AF2 signal would strengthen the conclusions in Fig. 1G-H.
   • e. Quantification would again be helpful in Fig. 5C to demonstrate changes in NS localization. In addition, it looks like Nrde2-KO does not just lead to lack of CCDC174 accumulation, but to a decrease in its overall expression. The authors should comment on this observation, or quantify CCDC174 signal in both images to demonstrate that the overall levels remain the same.
3. Since BCLIPseq is a technique developed by the authors, a more in-depth discussion of the technique development and quality control of the resulting data is warranted.
   • a. The authors mention that BCLIPseq offers a "streamlined and sensitive alternative to existing CLIP techniques" (p.9), but they don't provide any specifics into the ways they improve existing CLIP techniques in the main text. In what ways is it more streamlined and sensitive? This should be discussed in the main text rather than just the discussion, in order for the assertions made to be backed up with (supplementary) figures. A comparison of the coverage provided by a BCLIPseq library for NRDE2 to a CLIP library, for instance, would help to support these assertions.
   • b. The authors should address or provide evidence for why on-bead polyadenylation is preferable/more efficient than adapter ligation, especially as polyadenylation may be variable across transcripts. For instance, the authors could show more controls demonstrating the efficiency of on-bead polyadenylation or cite papers that have already extensively tested on-bead polyadenylation.
   • c. Many other RNA footprinting techniques (eCLIP, RIPseq) have noted significant nonspecific background in the resulting libraries, and usually use input controls to filter for this nonspecific background. The authors should clearly state if their BCLIPseq libraries also suffer from the same nonspecific background, and if so, what quality control steps exist in their analysis pipeline to minimize this background.
   • d. Related to the previous point, there is a high amount of rRNA reads in all the BCLIP libraries except EIF4A3. The authors suggest it is likely background, but if they are using a FLAG antibody for all of these, I'm not sure why there would be so much more background for some and not the others. If it's because EIF4A3 pulls down much more RNA with it because it binds most exon-exon junctions, whereas binding of the others is more rare, then isn't it possible that the mRNA reads are also partially background? This could explain why there is a very small overlap between the BCLIP bound loci and the affected 5'SS. An input control would help to determine what is indeed background.
4. The conclusion that "This [U1 snRNA binding] leads to the provocative idea that NRDE2 could potentially mediate the formation of a non-canonical U1 snRNP" (p.20) is a very intriguing conclusion that would largely benefit from additional experiments to strengthen the claim.
   • a. Depletion of a U1-specific protein (U1-70k, U1C) and analysis of the effect (or lack thereof) on Nrde-U1 snRNA interactions would strengthen the assertion that Nrde-U1 snRNA interactions are independent of core U1 snRNP components.
   • b. Depletion of a U1-specific protein (U1-70k, U1C) and analysis of the effect (or lack therefore) on Nrde2-KO sensitive introns would also strengthen the assertion that Nrde2 regulates introns as part of a non-canonical U1 snRNP.
   • c. Overall, a schematic in the last figure that depicts the splicing model presented in the discussion would be helpful for describing the Nrde2/Ccdc-174 model proposed.
   • d. The authors show that the majority of ms-snRNA reads map to U1 snRNA. However, U1 snRNA is generally more abundant than other snRNAs (Dvinge et al. 2019), so the authors should show how the distribution shown in Fig. 6B compares to the input distribution of snRNA levels in the cell line used. Also, relative levels of U1 snRNA detected by IP-Northern Blot (Fig. 6C) don't seem to match the results shown in Fig. 6A and B, as U1 snRNA seems most abundant in the NRDE2 IP by Northern Blot and most abundant in NRDE-Δ200 by BCLIP-seq.
   • e. In Figs. 6D, E and F, the authors suggest that NRDE2 and CCDC174 contact U1 snRNA at multiple positions based on observing highest enrichment over SL2 and SL3. However, snRNAs are highly structured and modified, which may interfere with reverse transcription during library preparation and lead to uneven signal throughout the gene body. To show that the proteins of interest are really enriched at these positions, the authors could perform the same experiment for a protein that is known to bind at a different location on U1 snRNA.
5. The rMATs analysis performed is very lenient; notably, there is no reported filtering for splicing events with some minimum coverage across replicates, and the inclusion level difference threshold of >0 (rather than >0.1 etc) is extremely low. As the rMATs analysis is key to the authors conclusion that there is "frequent cryptic 5'SS upon Nrde2 knockout" (p. 13), it seems important that this analysis is performed with more stringency in order to capture robust and meaningful splicing changes.

Minor comments

1. Some parts of the paper are organized in a confusing manner:
   • a. It is unclear why development of the xTAP-MS protocol is under the section "NRDE2 Localization to Nuclear Speckles Depends on Active pre-mRNA Splicing"
   • b. The section "Nrde2 and Mtrex Knockouts Induce a 2C-like State", while interesting, seems to be outside the scope of the paper
2. It would be interesting for the authors to look into the BCLIPseq data to see if there are any enriched RBP binding motifs for the proteins studied.
3. Western blots for IPs (ex. Fig 1F) should show the input for both the IP bait and prey proteins, not just the prey. In addition, input and IP'ed protein should be displayed in the same western blot image (without cropping in-between).
4. Previous studies (Boehm et al 2018) have found that other EJC-associated proteins also are important for regulation of 5' cryptic splice site usage. It would be interesting for the authors to compare the 5' cryptic splice sites identified in these earlier studies to look for overlap between the 5' cryptic splice sites regulated by these proteins vs. NRDE2.
5. The luciferase reporter assays are an especially strong portion of the paper, and are a nice orthogonal validation of the link between Nrde2, splicing regulation, and SS strength.
6. It would be interesting for the authors to investigate the effect of the mutations in the luciferase reporter constructs on the binding patterns of Nrde2 on construct transcripts. This may help provide a mechanistic basis for the chosen cryptic 5' SSs.
7. "Thus, NRDE2 promotes splicing from the most upstream of a series of 5'SSs" (p. 16) is an interesting conclusion, but this statement is far too general given the low number of genes surveyed using the luciferase assay. The statement should be rewritten to reflect that this statement has only been shown to be true for the few genes tested.
8. The statement "NRDE2 and CCDC174 promote splicing at many of the same weak 5'SSs" (p. 16) would be stronger if it was not just based on the genes studied through the luciferase assay, but based on splicing changes analyzed genome-wide through rMATs analysis. Do NRDE2 and CCDC174 promote splicing of the same weak splice sites globally?
9. R squared values should be added to the correlations presented in Fig. S2B to support the claim that replicates "correlated strongly", since that is the basis for merging replicates in subsequent analyses.
10. In Fig. 3A, the four clusters should be annotated on the figure to increase clarity.

Significance

Overall, the study is thorough in its approaches to studying NRDE2 biology and makes a strong case for the exciting role of NRDE2 and CCDC174 in 5'SS selection. Combined IP-MS and RNA footprinting approaches compellingly demonstrate NRDE2's associations with splicing factors and splice sites in vivo. In addition, the combination of genome-wide approaches (RNAseq) with targeted analyses (luciferase reporter assays) allow for detailed analyses of cryptic splice site choice in the absence of NRDE2, NRDE2 mutants, MTREX, or CCDC174. These experiments support the novel role of NRDE2 and its associated proteins in splice site choice.

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

Learn more at Review Commons

---

Reply to the reviewers

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

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

Learn more at Review Commons

---

Referee #3

Evidence, reproducibility and clarity

Summary:

This paper seeks to identify genomic safe harbor loci in the human genome for the integration of transgenes. The authors use computational analysis to identify a set of potentially useful sites for transgene integration; they subsequently test a small subset of these identified locations in human embryonic stem cells to determine the impact of transgene integration on the transcriptome and the ability of these cells to differentiate into numerous cell types. They determine that the subset of sites they identify and test all seem promising as no major changes in transcription or differentiation were observed following integration.

Major Comments:

Overall, the conclusions of this paper are reasonably convincing, and the authors do a nice job of laying out the criteria for designation of genomic safe harbor loci and characterizing three of these loci. However, there are several places where the data and rationale for the experimentation could be clarified to make the conclusions more convincing. One major question is whether or not these safe genomic loci identified are actually better than traditionally used loci such as Rosa26 or CCR5. While the authors note in the discussion section that these traditional loci do not meet their criteria for a safe genomic integration site, they do no direct comparison of their new loci vs these more traditional ones. A side-by-side comparison would make the data more convincing that these loci are better suited for genomic integration (such as noting fewer changes in the transcriptome etc).

The second major issue is that it is unclear how the authors picked seven loci for more extensive targeting out of the 25 initially identified. Without knowing the criteria used for these selections, it is challenging to know if there was a bias in selection of sites for further analysis that could alter results, or if the other identified sites are truly acceptable targets. In addition, only three of those GSH sites were successfully targeted. As a result, it is hard to determine the validity of the authors claim that they identified 25 unique GSH loci when they only fully characterize three of them. While it is not necessary to test all 25, it might be beneficial to test more than three before making these conclusions.

The data and methods in the paper are generally presented in an understandable fashion, and the use of three biological replicates for characterization of the hESC lines seems reasonable.

Minor Comments:

There are several minor issues that addressing could help strengthen the claims in this manuscript. First, it is unclear how the authors used BLAT to narrow down their initial list of 49 safe loci down to 25. A more detailed explanation in the text (vs methods) would aid in reader understanding of methodology. In addition, a deeper explanation of how differentially expressed genes were identified would be helpful. The authors state that many of the DE genes in their GSH targeted loci were identical to those found in both control and untargeted cells. It is unclear what the comparator was in these experiments that was used to identify those DE genes; clarification of this in the text would be helpful for the reader. In figure 2, the labeling of the panels is quite confusing, as panel F appears between panels E and D. Finally, in figure 3, while the three new cell lines are shown to be differentiated into various cell types, no control images are shown for comparison. This would be helpful to add in.

Significance

Overall, the major advancement of this paper is the identification of numerous putative genomic safe harbor loci (GSH) for the integration of transgenes in the human genome. With the rapid development of novel gene therapy techniques, the characterization of locations in the genome that are acceptable for transgene integration with the lowest likelihood of unintended off-target or downstream consequences is important. As a result, these sites have the potential to be quite valuable for the gene therapy field and of great interest to many scientists. Thus far, few widely accepted safe locations for genomic integration have been identified, making these sites of interest to numerous labs. As someone who has generated numerous transgenic mouse lines using random integration, the ability to selectively target transgenes and know there will be minimal issues with silencing or off-target impacts is appealing. However, my knowledge of genomic and bioinformatics techniques is minimal, making it challenging for me to adequately assess that piece of this manuscript, though I believe it contains valuable information for the gene editing and gene therapy community.

Referees cross-commenting

I agree with comments from the other reviewers and think they are all very reasonable suggestions.

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

Autio MI and colleagues report their study aiming to identify novel genomic safe harbor (GSH) loci in the human genome. First, they conducted a computational analysis of publicly available data with a list of criteria previously suggested for GSH loci. Since expression units placed at GSH loci should stably active, they also examined candidate loci using GTex data and against chromatin regions in the active (A) compartment reported by Schmitt et al. They found 25 candidate loci after these analyses. Then, they successfully placed landing pad constructs on three loci in hESCs by use of the CRISPR technology. They have demonstrated that expression of Clover by the CAG promoter is homogeneous and stable even after differentiation to neuronal, live and cardiac cell lineages.

Major points:

1. They only examined expression from the CAG promoter unit. However, this does not guarantee stable expressions from other promoters. Since the CAG promoter is very strong, it may be resistant against cellular silencing activity. For research purpose, tetracycline-regulatable promoters are often used, and it has been reported that although CAG promoter is not silenced, the TRE promoter is silenced when an expression unit is placed at AAVS1 locus (Ordovas L et al. Stem Cell Rep, 5: 918-931, 2015). Therefore, before concluding that these loci are GSH, expression from the TRE promoter should be tested.
2. They examined off-target integration by PCR and Sanger sequencing of the top 5 predicted off target sites. However, Southern blot analyses are needed to rule out off-target integrations. (This reviewer cannot evaluate data of copy number analysis using Digital PCR).

Significance

Identification of GSH loci will advance basic research and clinical applications.

This reviewer is not good at bioinformatics and cannot evaluate the first half of this study.

Referees cross-commenting

I have found that comments by other reviewers are important. As suggested, functionality of differentiated cells should be tested and demonstrated. Again, examine other promoters beside CAG should be tested in those differentiated cells.

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 Authors have taken a bioinformatic approach to identification of safe harbour loci in human genome, and then validated three of these in the H1 hES cell line. Overall, the rationale and data presented in clear and and the experiments appear to be reproducible.

Minor concerns:

1. please expand on the rationale for selection of the seven sites that were selected for initial targeting (i.e. what differentiated these from the other 18 sites as being suitable), and on the results for why no successful edits were identified for 4 of these loci.
2. please add data that quantifies the number of cells expressing Clover, in the differentiated cell types. Ideally, multiple markers for each lineage should be used.
3. Functional studies of the differentiated cell types would add substantial value to this paper. in the absence of such data, additional marker proteins that reflect functional properties or the maturity of the derived cell types could be added.

Significance

Identification and characterization of new safe harbour sites offers potential for generation of research tools and potentially for clinical applications. Those working in the fields of iPSC-based disease modelling and pre-clinical gene therapy are likely to be interested in this work and the cell line resources developed.

Reviewer expertise: iPSC, CRISPR/Cas, neuronal differentiation.

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

Learn more at Review Commons

---

Reply to the reviewers

Manuscript number: RC-xx-xx

Corresponding author(s): Yang Hong

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

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

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

1. General Statements [optional]

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

We would like to thank all three reviewers for their comprehensive and constructive comments. We are in particular grateful to_ reivewer#3_ for suggestions on improving the manuscript text and figures. We have already incorporated most of these suggestions into the revised manuscript. Based on reviewers’ comments, it appears no additional significant experiments are required for the revision. Nonetheless, for the final revision we will do the following:

2. Description of the planned revisions

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

---

REVIEWER #1

(Evidence, reproducibility and clarity (Required)): Summary: This manuscript takes a closer look at how hypoxia affects the accumulation of PI4P and PI4,5P2 (PIP2) in the plasma membrane of Drosophila ovarian follicular epithelial cells and how ATP depletion similarly affects the localization of the same phospholipids in HEK293 cells. They demonstrate that hypoxia results in the reversible loss of plasma membrane (PM) association of both lipids, with PIP2 disappearing ahead of PI4P, and recovering more slowly than PI4P when oocytes are returned to normoxia. They also show that the intracellular vesicular pools of PI4P are depleted ahead of the PM pools and the PI4P recovery occurs first in PM, then in the vesicles. They show that the disappearance and recovery of the polarity protein Lethal giant larvae (Lgl) parallels that of PIP2 during hypoxia and subsequent normoxia, with a very slight delay. The authors then go on to show the RNAi knockdown of the PM enzyme (PI4KIIIa) that phosphorylates PIP delays the recovery of PI4P at the membrane, with recovery first occurring in the vesicular pools. This knockdown also delays the recovery of PIP2 and, as with recovery of PI4P, the recovery of PIP2 now occurs first in vesicular pools. Lgl recovery follows that of PI4P and PIP2 with RNAi knockdown of PI4KIIIa. The knockdown of all three of the enzymes that phosphorylate PIP to generate PI4P delays recovery of PI4P, PIP2 and Lgl at the membrane even more. The authors show that proteins required for the PM localization of PI4KIIIa have similar effects on the recovery of PM PI4P, PIP2 and Lgl (with delays and recovery of vesicular pools before PM pools). Independently, the authors show that ATP depletion in HEK293 cells result in similar reversible depletion of PI4P, PIP2 and Lgl from the PM. From these studies and their previous findings, the authors conclude that pools of PI4P and PIP2 are likely rapidly turned over in the membrane even during normoxia and that this rapid recovery is dependent on the PM localized enzyme that phosphorylates phosphoinositol.

Major comments: Overall, the data are beautifully presented; it is quite helpful to have a video of each experimental treatment showing the corresponding response of all three molecules that are being monitored. Signal quantification over time is carefully documented. With the exception that a link between hypoxia and depletion of ATP has not been demonstrated here, the key conclusions are convincing. However, as pointed out below (in the significance section), some of the major points have already been published by this group. Their conclusion that hypoxia induces acute and reversible reduction of cellular ATP levels (which are then proposed to affect the activities of the enzymes required for PI4P and, consequently, PIP2 production) was not shown. They did demonstrate that acute depletion of ATP had the same consequences on PM phospholipids as acute hypoxia (in HEK293 cells). And, indeed, it makes sense that hypoxia could affect enzymes required for ATP synthesis, but the authors would have to show that acute hypoxia results in acute reduction in cellular ATP pools to make the links they suggest. This is something they should be able to do in the HEK293 cells now that they have their ATP sensor. Just to note, this group did show that hypoxia can reduce levels of ATP in Drosophila oocytes in their previous paper (Dong et al., 2015, Figure S3), but it is unclear if this is reversible and happens in the time frame of the experiments presented in this current manuscript.

In Dong et al 2015 we biochemically measured ATP levels in embryos treated by hypoxia, but due to lack of ATP biosensors it was not possible then to show real time ATP level changes in cells undergoing hypoxia and reoxygenation. Instead, we showed that direct ATP inhibition by antimycin treatment mimics the effect of hypoxia, to support the hypothesis that hypoxia acts through ATP inhibition.

In the current manuscript, we demonstrated for the first time that hypoxia triggered acute and reversible ATP level reduction in Drosophila follicular epithelial cells (Figure S3). In the finalized manuscript we plan to add data to show ATP level changes in HEK293 cells under hypoxia, as suggested by the reviewer.

My suggestions are the following: (1) The authors need to make it absolutely clear what was already known, including the following: (A) hypoxia reversibly affects PM pools of PI4P, PIP2, and Lgl (and other membrane associated proteins), (B) that hypoxia can affect ATP levels in Drosophila oocytes (although these previous studies do not show anything about the dynamics) and (C) that reducing ATP levels affects PM pools of PI4P, PIP2 and Lgl.

We agree with reviewer and have revised the introduction (p3 line 24-34) to make clearer what we previously published on the hypoxia/ATP and PIP2 turnover. It should be noted though that our previous publications did not contain any data regarding PI4P under hypoxia or ATP inhibition, as the current manuscript is the first time we reported the making and use of PI4P sensor such as P4Mx2::GFP.

(2) They should demonstrate that acute hypoxia and return to normoxia has acute and reversible effects on cellular ATP levels - they now have the tools to do this, at least in HEK293 cells.

We agree with reviewer and will add this data to the final revision. Such experiments require significantly modified setup for imaging live HEK293 cells with controlled hypoxia/reoxygenation, but we are reasonably confident that such experiments are feasible.

Minor comments:

The manuscript is too long and the discussion unnecessarily repeats everything already presented in the results. The authors should find a way to streamline the discussion.

We will revise the final manuscript to make the discussion more concise and streamlined.

N values should be given for all figures and experiments, and the N=23/24 versus N=24/24 needs to be explained the first time it is used.

We have revised manuscript so all N values are clearly provided and easier to understand.

There are a few mismatches in terms of plural nouns and singular verbs and vice versa sprinkled into the manuscript, so some careful editing would be useful.

We have revised the manuscript to eliminate such errors/typos, especially with the help of the generous and comprehensive list of by reviewer#3.

Significance: I was initially quite excited about the novelty of their findings and the potential insight into the dynamics of PM pools of the two phospholipids that are critical to cell polarity and that play important signaling roles. However, at least a subset of their conclusions were either published in their earlier work or do not necessarily follow from what they have done in this manuscript. Their statement that hypoxia in Drosophila induces acute and reversible depletion of PM PI4P and PIP2 was presented in a previous publication (See Figure 8 of Dong et al., 2015).

We greatly appreciate reviewer’s comments on the significance of our discovery. Again all data regarding PI4P are new in this manuscript and have not been published before. We only published very preliminary data suggesting the reversible depletion of PIP2 and PIP3 (but not PI4P) under hypoxia (Dong et al, 2015). The current manuscript provides a comprehensive set of quantitative live imaging data with high spatial and temporal resolution that demonstrate for the first time the dynamic turnover of PM PI4P under hypoxia and ATP inhibition, the correlation between such turnovers of PM PI4P and PIP2, and the direct correlations between PI4P/PIP2 turnover and Lgl electrostatic PM targeting and intracellular ATP levels. In addition, studies on the role of PI4KIIIa complex in such process have not been done before.

This manuscript would appeal to an audience interested in the mechanisms of cell polarity and phosphoinositide signaling.

I am a Drosophila developmental geneticist quite familiar with the topics that this paper addresses.

REVIEWER #2

(Evidence, reproducibility and clarity (Required)): Summary: This manuscript describes the effect of hypoxia on the levels of PI4P and PI45P2 , two key PPIs that are enriched on the inner leaflet of the plasma membrane. These PPIs are synthesized by the sequential phosphorylation of PI by a PI-4 kinase and subsequently a PI4P 5 kinases, both of which use ATP. The relevant PI-4 kinase at the plasma membrane, PI4KIIIa has been conclusively identified previously in mammalian cells by the DeCamilli lab (Nakatsu et.al JCB 2012) and its role in regulating the synthesis of PI4P and PI(4,5)P2 in two Drosophila cell types in vivo shown by two previous studies. Balakrishanan (photoreceptors during PLC signalling) and Basu et.al Dev.Biol 2020 ( in multiple larval cell types ). PI4KIIIa has been shown to exist as a complex of the enzymatic polypeptide, EFR3 and TTC7. The studies by Nakatsu, Balakrishnan and Basu have shown the importance of the complex subunits is regulating PI4P and PI45P2 levels in cultured mammalian cells and Drosophila cell types in vivo.

We thank reviewer for pointing out the work of Balakrishanan et.al J.Cell Sci 2018 which showed similar results to our manuscript and Basu et al 2020. We have added a brief summary this reference to the Discussion in revised manuscript (p13, line 17)

In the present study, Lu et. al build on their previous work showing that the polarity protein Lgl undergoes hypoxia induced translocation. They show that hypoxia also induces loss of PI4P and PI45P2 at the plasma membrane in these cells correlated with loss of Lgl localization to the PM. The manuscript then goes on to establish the requirement of the PI4KIIIa complex in regulating Lgl localization as well as PI4P and PI45P2 levels at the plasma membrane during hypoxia and the subsequent recovery of these at the plasma membrane.

The strength of the manuscript is twofold. (i) The work is done to a high technical standard and the investigators have carried out the measurements of LGL localization, PI4P and PI45P2 levels along with simultaneous measurements of ATP levels in vivo. The work would be strengthened further if the authors could show the level of depletion of PI4K isoforms or PI4KIIIa complex subunits units induced in ovarian tissue under their experimental conditions by the GAL4 drivers used in this study. This is not a persnickety detail as RNAi lines can have very different effectiveness in Drosophila ovarian tissue compared to other fly cell types. This point is, in particular, important in cases where an RNAi line is being used and the conclusion is a lack of impact on a phenotype being studied.

We are fully aware of the potential caveata of RNAi. In our previous publications we were able to validate RNAi knock-down efficiency against endogenously or ectopically expressed GFP-tagged target proteins (Dong et al., 2020; Dong et al., 2015; Lu et al., 2021) or endogenous proteins with available antibodies (Dong et al., 2015). It is regrettable that presently such reagents are not available for directly examining the level of RNAi knock-down for PI4KIIIa and PI4KIIa etc. We did show that rbo-RNAi efficiently knocked down the expression levels of Rbo::GFP (Figure S1C). In current manuscript, we have been very careful to draw conclusions based on negative RNAi results.

(ii) A second strength is that the authors now illuminate a further in vivo cell type where the function of the PI4KIIIa complex in regulating PI4P and PI45P2 levels. This adds to the earlier work of Nakatsu, Balakrishnan and Basu.

A key difficulty with the current story is the lack of specificity of the phenotype they demonstrate under hypoxia. Of course, hypoxia is expected to deplete cellular ATP levels but PI4KIIIa is not the only enzyme that this lack of ATP will impact. There will be dozens or more other kinases, both protein and lipid kinases whose function will be impacted by the drop in ATP levels. Therefore, it is hard to attribute a specific/particular role to the PI4KIIIa complex under these conditions. The mislocalization of LGL::mCherry while correlated with PI4P and PI45P2 levels at the plasma membrane may be just that- a correlation. It is quite possible, indeed likely, that the mislocalization of LGL-mCherry under hypoxia conditions is due to the reduction of the activity of another lipid or protein kinase due to the drop in ATP levels due to hypoxia (PKC is a possibility too).

We agree with reviewer that PI4KIIIa almost certainly is only one of the enzymes that are involved in regulating the PI4P/PIP2 turnover triggered by hypoxia. This manuscript is our first effort to investigate the potential regulatory network underlying the hypoxia-triggered turnover of PM PI4P and PIP2, and it is our long term goal to identify more components in the regulatory network.

As to underlying mechanisms of the loss of PM Lgl under hypoxia, we previously showed before that PM targeting of Lgl dependents on both PI4P and PIP2 and acute depletion of PI4P and PIP2 in cultured cells completely blocks the PM targeting of Lgl (Dong et al, 2015). Thus, although we cannot exclude the contribution from other lipids, it is highly plausible that loss of PM PI4P and PIP2 triggered by hypoxia is the main driving force disrupting the electrostatic PM targeting of Lgl.

Lgl is phosphorylated by aPKC and such phosphorylation inhibits Lgl PM targeting by neutralizing the positive charges on Lgl’s polybasic motif (Dong et al, 2015, Bailey et al, 2015). Thus, potential inhibition of aPKC activity by hypoxia should not inhibit the PM targeting of Lgl. Consistently, we previously showed that aPKC-/- mutant cells showed same acute and reversible loss of PM Lgl under hypoxia (Dong et al, 2015).

Minor comments:

The authors must reference all published work on the PI4KIIIa complex in the literature. Some of it is excluded in the present version

We apologize for the missing references and in the revised manuscript we have already added several additional references based on the suggestions of reviewer#1 and #3. In the finalized manuscript we will make our best effort to cover all the relevant studies.

The Drosophila work, particularly cell types used, etc are not accessible to people who are not fly experts. This should be done.

We added a sentence to the first paragraph of Results to specifically make it clear that all Drosophila studies used follicular epithelial cells from female ovary (p4, line 26).

Significance: Adds to knowledge on the PI4KIIIa complex. Builds on existing knowledge in the PI4KIIIa field and maybe also cell polarity field.

REVIEWER #3

(Evidence, reproducibility and clarity (Required)):

Summary: Phosphatidylinositol phosphates (PIPs) are key determinants of membrane identity and regulate crucial cellular processes such as polarization, lipid transfer and membrane trafficking. Despite decades of study, surprisingly little is known about how levels of PIPs are regulated in response to cellular stress. Here, using Drosophila ovarian follicular epithelial cells and human HEK293 cells, the authors show that levels of plasma membrane (PM) PI4P and PIP2 decrease rapidly in response to hypoxia, resulting in loss of polybasic proteins from the PM. These effects are reversed in response to reoxygenation. Similarly, hypoxia leads to acute depletion of ATP levels, which also regenerate following reoxygenation. Using a combination of quantitative image analysis and genetic analysis, they show that PI4KIIIalpha and its binding partners Rbo/ EFR3 and TTC7 are needed to maintain PI4P and PIP2 at the PM under normal and hypoxic conditions, whereas the other two Drosophila PI 4-kinases, Fwd/PI4KIIIbeta and PI4KII, play a less important role in PM PIP homeostasis. Their results suggest that manipulations with indirect effects on PIPs (hypoxia, ATP depletion, ischemia) can have a profound impact on electrostatic charge at the PM, as well as downstream processes that require PM PI4P and PIP2.

Major Comments: 1. In general, the authors' conclusions are convincing. However, some of the results are less evident from the still images and graphs provided in the figures than from the movies that accompany the figures. Some suggested improvements are below.

No additional experiments are essential to support the claims of the paper, although some additional quantitation would be helpful to the reader, as detailed below. Data and methods are generally presented in such a way that they can be reproduced, although some additional details would be helpful, as listed below. Experiments were adequately replicated, and statistical analysis appears adequate.

We are extremely grateful to the generous efforts of the reviewer providing such a detailed list of suggested improvements. We have incorporated all the text revisions into the revised manuscript and will revise the figures accordingly the final revision too.

Minor comments: 1. Although the data are generally quantified quite well, there are two instances in the first full paragraph on p. 5 where this is not the case. First, PM PI4P is described as "oftentimes" as showing a transient increase in the early phase of hypoxia. However, this is not quantified. How often did this occur among the samples examined? How large is the transient increase when it occurs (Fig. 1A' error bars are not obvious on the colored background)? Second, the authors state that the P4Mx2-GFP puncta "often" became brighter after recovery. How often did this occur? No quantitation is provided.

Upon close inspection of the data, we conclude that during the early phase of hypoxia PM P4Mx2::GFP always showed an initial drop followed a transient increase. Thus we revised the sentence to delete “oftentimes”.

We did not specifically quantify the transient increase of the PM P4Mx2::GFP during the early phase of hypoxia, as we consider it likely an artifact as discussed in the manuscript, making its quantification less meaningful.

As to the P4Mx2::GFP puncta, regretfully we do not have imaging tools that can accurately and automatically recognize and measure such puncta in our live recordings. We are actively developing such software using trainable Weka segmentation tool (https://imagej.net/plugins/tws/) and hopefully such puncta quantifications will be possible in our future experiments.

The authors conclude that "PI4KIIalpha and Fwd contribute significantly to the maintenance of PM PI4P" (bottom of p. 7), yet they did not validate their RNAi knockdowns of these two genes, so they do not know whether it is one or both of these PI4Ks that contribute.

We agree with reviewer that our RNAi knockdowns on PI4KIIa and Fwd are not sufficient to tell whether one or both contribute to the PM PI4P maintenance. We revised the sentence to “Our data support that PI4KIIα and/or FWD contribute significantly to the maintenance of PM PI4P...”

In Fig. 4B, a subset of the cells "show failed recovery of PM Lgl::GFP". However, some cells did recover. This average percentage of cells that recovered should be quantified, if possible.

Added numbers of PI4K-3KD cells that show normal or failed hypoxia response of Lgl::GFP and revised the sentences accordingly (p8, line18)

In Fig. 7A, B, the bottom cell in each example lags behind the top cell in recovery of the MaLionR sensor. The frequency of observed cells in each class for 7A, B should be quantified.

Added n numbers of each cell classs to Figure 7A, B legend.

In most cases, prior studies were referenced appropriately. However, two previous studies in Drosophila showing the effects of Sktl/PIP2 reduction on localization of polybasic proteins Lgl, Baz/Par-3 and Par-1 were not cited (relevant to the first paragraph of the Introduction, p. 3): Gervais et al., Development (2008), Claret et al. Curr Biol (2014). In addition, two studies showing the importance of Drosophila PI4KIIIalpha in synthesizing PM PI4P and PIP2 were not cited (relevant to the description of this enzyme, top of p. 6): Yan et al., Development (2011), Tan et al., J Cell Sci (2014). Data showing fwd null mutants are not lethal (relevant to top of p. 7) were published in Brill et al., Development (2000).

We thank reviewer for suggesting the additional references. We added Yan et al and Tan et al for referencing PI4PIIIa, and Brill et al for referencing the original characterization of fwd. We discussed work from Gervais et al and Claret et al in the revised discussion (p14, line 9).

For the most part, text and figures are clear and accurate. However, there are quite a few typos and grammatical mistakes, as well as instances of lack of clarity in the writing that should be addressed. In addition, there are a number of improvements to presentation of data that would make the figures easier to understand. These are listed below. Suggestions to improve presentation of data and conclusions are below.

Again we greatly appreciate such generous efforts from the reviewer and have incorporated all the text revisions into the revised manuscript. We will revise the figures accordingly the final revision too.

Significance: Overall, the authors do a nice job of showing that hypoxia leads to previously unappreciated effects on levels of PM PI4P and PIP2, resulting in loss of PM association of proteins important for normal cellular physiology. This finding is quite novel. Moreover, the authors provide insight into the identity of the PI4Ks that are responsible for regenerating PM PIP2 following return to normoxia. Their analysis of the dynamics of these changes provides multiple interesting insights, including the potential roles of intracellular pools of PI4P in replenishing PM PIP2 and the observation that intracellular accumulation of PIP2 is occasionally observed in association with the appearance of intracellular PI4P puncta, suggesting a novel route for PIP2 replenishment in response to hypoxic stress. Their results will provide the basis for future studies examining the cellular mechanisms involved. This study will be of interest to those studying phosphoinositide biology as well as cellular responses to hypoxic stress and recovery, such as occur during ischemia and reperfusion. Reviewer expertise: Drosophila molecular genetics, cell biology, developmental biology, phosphoinositides, PIP pathway enzymes, PIP effectors

---

**REFEREES CROSS-COMMENTING** This session includes the comments of all reviewers.

__Reviewer 3: __I agree with reviewer #1 that the authors did not do a good job of clarifying what they and others had previously shown, and I must confess I didn't carefully examine their previous papers carefully enough before preparing my review. In fact, they previously showed that hypoxia affects localization of Dlg at the plasma membrane and that its recovery depends on PI4KIIIalpha and PIP2 (Lu et al., Development 2021). This is in addition to their previous data showing effects of hypoxia on Lgl (Dong et al., J Cell Biol 2015). Thus, less of the information in the current manuscript is novel than I thought when I initially read it.

I also agree with reviewer #2 that they need to do a better job of citing the relevant literature and considering the possibility that hypoxia and reduced levels of ATP might affect many different enzymes. In addition, as suggested by reviewer #1, it seems important

__Reviewer 1: __I agree with what Reviewer 3 is suggesting and with reviewer 2 that the authors should do a better job of citing all of the relevant literature. I also appreciate the detailed edits provided by Reviewer 3 - it was very generous of them to do this.

__Reviewer 2: __The points raised by reviewer 1 and 3 with regard to the citing or prior work (from the authors or other labs) also applies to their citing of literature on PI and PI4K signalling. Here too citing or prior work has been less than satisfactory making it difficult to do this.

We want to thank all three reviewers for their thoughtful and constructive comments. We have revised the introduction to make it clear what we had observed in our in our previous studies. On the other hand, in this manuscript we presents a systematic study on the hypoxia-triggered turnover of PM PI4P and PIP2, the correlation between PI4P/PIP2 turnover and electrostatic PM targeting of Lgl, as well as a potential role of PI4KIIIa and its PM targeting mechanism in regulating the turnover of the PM PI4P and PIP2 under hypoxia. Although the latter by no means indicates that PI4KIIIa is the only enzyme in regulating such process, its characterization is the beginning for us to further identify additional enzymes and regulators in this hypoxia triggered phenomenon.

We have already added additional references as suggested by the reviewers in the revised manuscript, and once additional experiments are completed we will finalize the manuscript to make sure all relevant references are cited.

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

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

We have incorporated nearly all of the suggestions from reviewer#3 into the current revision, with few exceptions as listed at the end of this letter. Below are point-to-point response to selected suggestions involving data interpretation and comprehensive text revisions:

1. • 5, first paragraph, line 2: replace "oftentimes" with "often" and provide quantitation (see above) *

     Deleted the “oftentimes”. Upon close inspection of our data we conclude that PI4P always showed transient increase of PM signal in early hypoxia.

2. • 5, first paragraph, line 6: the claim of "often" should be quantified (see above) *

     Deleted the “often”. PI4P puncta actually were consistently brighter after recovering from hypoxia.

3. • 5, second paragraph: the extent of recovery of Lgl is less when Lgl-RFP is coexpressed with PLC-PH-GFP, potentially due to titration of PIP2 by PLC-PH; the authors should comment on this*

     This is a good suggestion from the reviewer. Revised by adding to the end of paragraph:* “Note that in Figure 1B Lgl::RFP recovery appears lower than in wild type, possibly due to the titration of PIP2 by PLC-PH::GFP expression.” *

4. • 5, last line: the authors should provide information about the "targeted RNAi screen"; which genes were tested? did any others give relevant phenotypes? a table showing the results of the screen should be provided as supplementary information*

     Added Table S1 which summarize the results of RNAi screen.

5. • 11, first full paragraph, line 6: what about PI4KIIIbeta? is the KmATP for this enzyme known?*

     Based on literature, KmATP of PI4KIIIbeat is similar to PI4KIIIa’s (~400uM, Balla and Balla 2006). We added the PI4KIIIb KmATP value to the revised discussion (p11, line22)

6. • 11, last paragraph, line 2: what is meant by "etc." is unclear; remove "etc." and include specific information related to what was reported in the literature (with proper references)*

     Revised the sentence to “…that KmATP of PI4KIIIα was measured decades ago using purified PI4KIIIα enzymes from tissues such as bovine brains and uterus (Carpenter and Cantley, 1990)”. The reference (Carpenter and Cantley, 1990) is a review which contains detailed of biochemical characterizations of PI4K kinases from numerous publications.

7. • 12, line 3: why do the authors claim that the intracellular pool of PI4P is first synthesized by PI4KIIalpha? what about PI4KIIIbeta? their results do not distinguish between these enzymes*

     We favor the hypothesis that PI4KIIalpha is responsible for synthesizing the intracellular pool of PI4P because the very low KmATP of PI4KIIa. PI4KIIIbeta has high KmATP just like PI4KIIIa.

8. • 12, last paragraph, lines 6-7: for the reader, please clarity the mechanism that was invoked to explain how PIP5K can make PIP2 from PI in E. coli (Botero et al., 2019)*

     Revised the sentence to “PIP5K can be sufficient to make PIP2 from PI in E.coli by phosphorylating its third, fourth and fifth positions (Botero et al., 2019)”

9. • 13, first paragraph, last line: cannot conclude that components of PI4KIIIalpha are "highly interdependent" without testing effect of knockdown of PI4KIIIalpha on Rbo and TTC7, etc.; instead, can conclude that the data are consistent with all of the components acting in the same process; also, delete "the" before "proper"*

     Revised the sentence to “ .. supporting that components in PI4KIIIαa complex may act interdependently for the proper PM targeting in vivo.”

10. • 14, second paragraph, lines 3-5: expand on this idea; what additional lipids could be important here? are there examples of other proteins that require these additional lipids?*
11. • 16, line 6: explain in brief what "pNP plasmid" is and how the multi-RNAi method works (what promoters drive expression of the shRNAs, how many shRNAs are included in the plasmid, etc.)*

      Added a section in Material and Methods to describe in details the generation of pNP constructs and fly stocks.

12. • 16, lines 8-3: appropriate references should be included for each stock where available*

      Added references to stocks cy2-Gal4, rbo::GFP and UAS-AT1.03NL1.

13. • 16, line 11: explain what UAS-AT1.03NL1 is*

      Added: “UAS-AT1.03NL1(DGRC#117011) was used to express the Drosophila-optimized ATeam ATP sensor AT[NL] in follicle cells (Tsuyama et al., 2017)”

14. • 16, lines 16-17: Gerry Hammond should not be listed as providing these constructs if he is a coauthor on the manuscript; appropriate references should be cited for these constructs*

      Revised the sentence to “Mammalian constructs of P4M::GFP, P4Mx2::GFP, and PLC-PH::GFP were as previously described (Hammond et al., 2014; Hammond et al., 2012).”

15. • 17, lines 3-4: sentence fragment "Images were further" is not complete*

      This was a typo, deleted.

16. • 19, lines 5-6 from bottom: title doesn't accurately reflect that PI4P doesn't appear to recover in WT control; why is this the case? recovery was observed in other experiments*

      Live recording showed that PM PI4P did recover during reoxygenation (Figure 3A, Movie S7). This particular recording in Figure 3A/Movie S7 was a bit difficult for automatic quantification by our custom software due to that P4Mx2::GFP signal somehow was weak and noisy, resulting in less than “ideal” recovery curves.

17. • 22, line 16: fix typo in "uncalibrated"; spell out what "AT[NL] sensor shows*

      Revised to “Heat map of the (uncalibrated) FRET ratio of ATeam ATP sensor AT[NL] in follicle cells”

Simple text revisions have been made for the following suggestions:

1. • 3, first paragraph, line 11, and second paragraph, line 2: combine references (Dong, Dong, Lu) with (Bailey and Prehoda, Hong)*
2. • 4, 8th line from bottom: PLC-PH is generally referred to as PLCdelta-PH in the literature and should be defined this way the first time "PLC-PH" is used p. 5, 5th line from top: I believe the callout should be to Fig. 1A' rather than Fig. 1B*
3. • 5, first paragraph, line 4: delete "to" after "sensor" p. 5, first paragraph, line 7: insert "the" before "intracellular" p. 6, first full paragraph, line 4: change "cells" to "cell" p. 6, second full paragraph, line 1: change "was" to "were" p. 6, second full paragraph, line 4: delete "ref" at beginning of line) p. 7, first paragraph, line 4: explain briefly (either here or in the Methods) the "newly published multi-RNAi tools" reported by Qiao et al., 2018 p. 7, first paragraph, lines 9-10: suggest changing to "...(Fig. 3A,B); the latter...sensors being bound..." p. 7, second paragraph, line 5: replace "is" with "are" p. 7, last paragraph, line 2: "under normal conditions" is vague; replace "normal" with "aerobic" or "normoxic" p. 8, first paragraph, line 4: replace "normal" with "aerobic" or "normoxic" for clarity p. 8, last line: callout to Fig. 5A appears incorrect; should be Fig. S1A instead p. 9, first full paragraph, lines 4 and 5: callouts should be to Fig. 6A (line 4) and 6B (line 5) p. 10, second full paragraph, line 1: insert "human" or "mammalian" after "cultured" p. 10, second full paragraph, line 3: delete "Although" and start sentence with "For reasons unknown, out..." p. 10, second full paragraph, lines 4-5: replace "in HEK293" with "to HEK293" and add a period after "present"; start new sentence on line 5 with "However, our data strongly..." p. 10, third line from bottom: fix typo in "Discussion" p. 11, line 9: replace "require" with "requires" p. 11, line 10: perhaps the authors mean to refer to "PI and PIP kinases" rather than "PIP and PIP kinases" p. 11, line 10: delete comma at end of line and replace with "and" p. 11, first full paragraph, lines 3 and 4: replace "PI4KIIIa" with "PI4KIIIalpha" (two instances) p. 11, first full paragraph, lines 5 and 7: the "m" in "Km" should be a subscript p. 11, first full paragraph, lines 8-9: add "the before "intracellular PI4P pool" and replace "slower" with "more slowly" and "recovers faster" with "recover more quickly" p. 11, first full paragraph, line 10: replace "loss" with "lose" p. 11, last paragraph, line 1: insert "a" after "such" p. 12, line 4: insert "is" after "PM PI4P" and delete "is" after "PI4P pool"; also replace "the transfer" with "this transfer" p. 12, line 5: delete "the before "intracellular"; delete the "is" after "(Dickson et al., 2014)" and replace with a comma"; replace "our data showing the loss" with "our data show loss" p. 12, first full paragraph, line 2: "interconversion" does not require a hyphen" p. 12, first full paragraph, line 5: why is PIP5K in red? p. 12, first full paragraph, line 6: replace "attracted" with "recruited" p. 12, first full paragraph, line 8: insert commas around "however"; also, "knockdown" does not require a hyphen p. 12, last paragraph, line 5: delete "the" before "PI4P" p. 12, 5th line from bottom: replace "are" with "is" p. 12, 4th line from bottom: start sentence with "In this regard," rather than "To this regard," p. 13, first paragraph, line 1: delete "the" and insert "a" before "large" p. 13, second paragraph, line 2: add comma after "localization" p. 13, second paragraph, line 3: delete "the" before "PM PI4P" p. 13, second paragraph, line 10: replace "is" with "are" before "supported" p. 13, third line from bottom: should read "PIP2" rather than "PIP" p. 13, second line from bottom: "must have" is awkward and unclear; perhaps change to "is predicted to have a" p. 13, last line; replace "polybasic proteins or" with "polybasic and" p. 14, first paragraph, line 1: delete "the" before "PM PI4P" p. 14, first paragraph, line 3: replace "domains" with "motifs" p. 14, first paragraph, line 6: insert "the idea" before "that Lgl" p. 14, first paragraph, last line: replace "or" with "and" p. 14, second paragraph, line 2: insert "of" after "plenty" p. 14, last paragraph, line 4: replace "While" with "Although" p. 14, last paragraph, line 6: delete "the" p. 16, line 5: insert comma after "chromosome" p. 16, line 7: insert "Additional stocks used were" p. 16, line 12: fix typo in David Bilder's name p. 16, line 14: italicize lgl::GFP p. 16, line 18: briefly explain what the pGU vector is and how it works p. 16, line 21: delete "the" before "young females"; also, the age of the females should be specified (newly eclosed? 1-3 days old? yeasted or not?) p. 16, 7th line from bottom: replace "ensures" with "ensure" p. 17, line 10: "overexposure" does not require a hyphen; add a comma after "necessary" p. 17, line 12: insert "in" before "visualizing" p. 17, line 14: reword to say "imaged live in a temperature-controlled chamber at 37{degree sign}C." p. 17, line 15: delete "which" after "serum" p. 17, lines 16-17: replace "washout by replaced with chamber with normal serum" with "washed out by replacing with normal serum" p. 17, 10th line from bottom: replace "were" with "was" p. 17, second line from bottom: change "ROIs" to "ROI" p. 18, line 5: replace "were" with "was" p. 19, line 4: replace "the follicular cells" with "follicle cells" p. 19, line 6: fix typo in "reoxygenation" (here and in legends to Figs 2-5, multiple instances) p. 19, line 14: no RNAi is shown in Fig 1, so delete "(WT/RNAi); also, for consistency with other figure legends (e.g., Fig. 3), suggest changing text here to: "P4Mx2, PLC-PH (n=18,18), pLC-PH, Lgl (n=20,20), P4M2, Lgl (n=20,20)" p. 19, 7th line from bottom: should read "hr:min:sec"*
4. • 20, line 3: replace "follicular" with "follicle" p. 20, line 5: replace "regulates" with "regulate"; also, it is not clear what is meant by "targeting and retargeting" (would be simpler to replace with "localization") p. 20, Figs 4-6 legends: the numbers of samples examined in these experiments are missing (n=?) p. 20, 10th line from bottom: replace "follicular" with "of follicle" p. 20, 7th line from bottom: delete "changes" p. 20, third line from bottom: replace "shown" with "seen" p. 21, lines 6 and 8: replace "(B)" with "(A')" and "(C)" with "(B)" p. 21, line 11: replace "(B)" with "(A)" and "(C)" with "(B)" p. 21, lines 18 and 23: insert "cells" after "12" p. 21, line 25: bold "(C')" p. 22, line 3: incomplete phrase should be replaced with "regardless of the transient loss of PM PM PIP2" (or "...PM PLC-PH-RFP") p. 22, line 4: replace "PLC-PH::GFP" with "PLC-PH::RFP" p. 22, line 6: replace "express" with "expression" p. 22, line 10: should this be "P4Mx2::GFP"? p. 22, line 12: add comma before "PLC-PH::GFP" pp 23-26: for clarity, delete word "samples" in legends to Movies S2-S16 p. 23: fix typo in "Lgl::mCherry" in legends to Movies S2 and S3 p. 24: change "Movie S09" to "Movie S9" p. 26: change "Time intervals is" to "Time intervals are" in legends to Movies S17-S19 Reviewer#3’s suggestions to improve the figures and movies: - show single-color images in grayscale, which is easier to see on black and helpful for colorblind readers (applies to all figures except Fig. S3); movies and merged still images should be shown in green and magenta for colorblind (not sure if channels in movies are difficult to change)*

We converted Fig. 1A to gray scale but found it visually inferior to the color version, as the gray images make the temporal differences between green and red channels less pronounced. We will change red channel in movies to magenta color but to convert red channels in all figures requires a very large amount of work to recapture and recrop all the frames used. We hope that reviewer understand our decision to keep colors in figures unchanged.

- replace colored labels on black boxes with colored labels on white background (Fig. 5A (left), Fig. 5B-D (top), Fig. 6A (left), Fig. 7A-C (left), Fig. S1A (left), Fig. S1C (top), Fig. S2A (left), Fig. S4 (left))

We have revised the figures accordingly.

- provide scale bars throughout (Figs 2-7, S1-S4)

We have revised the figures accordingly.

- replace pale colored boxes under labels for "hypoxia" and "air" with slightly darker boxes (Fig. 1A-C, Fig. 2A, Fig. 3A, Fig. 5A', Fig. 6A', Fig. S2B)

We tested many different combination of colors and the current set appears to give the best contrast so far. We decided to keep the original color.

- provide vertical lines similar to those in Fig. 4A' in all of the time-course graphs and/or making the background colors slightly darker (Figs 2A', 3A', 5A', 6A'); also make the error bars darker (Fig. 1A'-C', Fig. 4, Fig. 5A', Fig. S2B)

We revised all backgrounds in charts to make them similar to Fig. 4A’.

- for consistency, label PM index graphs in Fig. 4 and Fig. S2 as Fig. 4A' and Fig. S2A'

We revised figures 4 and S2 accordingly.

- why are some of the PM index graphs labeled "PM index" and others labeled "PM index-1" on the Y-axis? this should be explained or changed for consistency

The mixed use of “PM index” and “PM index-1” are relics due to different versions of software used throughout the project. We revised all graphs to make Y-axis “PM Index” label consistent.

- "blank diamonds" described in figure legend for Fig. 7B' are barely visible when printed

Revised Figure 7B’ by filling blank diamonds with grey color to increase their visibility.

- Fig. 7C is mislabeled (MaLionR label should be replaced with PLC-PH-RFP)

We corrected this error in revised Figure 7C.

- in Fig. S3A, it would help to know the size of the cells (i.e., how many were present in the area examined)

Revised the Figure S3A legend to clarify that each circle covers approximately three to four cells.

- movies should be referred to in order (current Movies S7 and S8 should be renamed S6 and S7, and current Movie S6 should be renamed S8)

We appreciate reviewer’s suggestion but decided to keep the movies in current order. This manuscript contains a large number of movies (total of 19) and to make them better organized we purposefully grouped movies based on the RNAi experiments (e.g. all three PI4KIIIa-RNAi movies are named consecutively). Although this makes the call out of two Lgl movies slightly out of order, we consider it a reasonable compromise for easier movie browsing for readers.

- in Movie S19, "PLC-PH::RFP" is mislabeled "PLC-PH::GFP" (both P4MX2 and PLC-PH are labeled GFP in the movie)

We renamed to movie file to correct this typo.

Description of analyses that authors prefer not to carry out

Please include a point-by-point response explaining why some of the requested data or additional analyses might not be necessary or cannot be provided within the scope of a revision. This can be due to time or resource limitations or in case of disagreement about the necessity of such additional data given the scope of the study. Please leave empty if not applicable.

Reivewer#3 suggested the following movie order changes:

1. • 6, second full paragraph, lines 6 and 9: the callouts should be to Movies S5, not Movie S4 p. 7, second paragraph, line 3: change name of Movie S7 to S6, and call out Movie S6 here p. 7, third paragraph, line 2: change name of Movie S8 to S7, and call out Movie S7 here*
2. • 8, first paragraph, line 8: change name of Movie S6 to S8, and call out Movie S8 here* We appreciate reviewer’s suggestion but decided to keep the movies in current order. This manuscript contains a large number of movies (total of 19) and to make them more accessible we purposefully grouped movies based on the RNAi experiments (e.g. all three PI4KIIIa-RNAi experiment movies are named consecutively). Although this makes call out of two Lgl movies slightly out of order, we consider it a reasonable compromise for easier movie browsing for readers..

We are regretful that we are unable to directly evaluate the RNAi knock-down efficiency of several genes such as PI4KIIIa. We have nonetheless been careful to draw the conclusions in the manuscript in accordance with the potential caveat of RNAi experiments. We did directly show that rbo-RNAi directly knocked down the Rbo::GFP (Figure S1C). In addition, although we could not confirm the knock-down of ttc7-RNAi, we showed it can reduced level of Rbo::GFP, which is likely due to an effective knock-down of TTC7 (Figure 6B).

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:

Phosphatidylinositol phosphates (PIPs) are key determinants of membrane identity and regulate crucial cellular processes such as polarization, lipid transfer and membrane trafficking. Despite decades of study, surprisingly little is known about how levels of PIPs are regulated in response to cellular stress. Here, using Drosophila ovarian follicular epithelial cells and human HEK293 cells, the authors show that levels of plasma membrane (PM) PI4P and PIP2 decrease rapidly in response to hypoxia, resulting in loss of polybasic proteins from the PM. These effects are reversed in response to reoxygenation. Similarly, hypoxia leads to acute depletion of ATP levels, which also regenerate following reoxygenation. Using a combination of quantitative image analysis and genetic analysis, they show that PI4KIIIalpha and its binding partners Rbo/ EFR3 and TTC7 are needed to maintain PI4P and PIP2 at the PM under normal and hypoxic conditions, whereas the other two Drosophila PI 4-kinases, Fwd/PI4KIIIbeta and PI4KII, play a less important role in PM PIP homeostasis. Their results suggest that manipulations with indirect effects on PIPs (hypoxia, ATP depletion, ischemia) can have a profound impact on electrostatic charge at the PM, as well as downstream processes that require PM PI4P and PIP2.

Major Comments:

1. In general, the authors' conclusions are convincing. However, some of the results are less evident from the still images and graphs provided in the figures than from the movies that accompany the figures. Some suggested improvements are below.
2. No additional experiments are essential to support the claims of the paper, although some additional quantitation would be helpful to the reader, as detailed below.
3. Data and methods are generally presented in such a way that they can be reproduced, although some additional details would be helpful, as listed below.
4. Experiments were adequately replicated, and statistical analysis appears adequate.

Minor comments:

1. Although the data are generally quantified quite well, there are two instances in the first full paragraph on p. 5 where this is not the case. First, PM PI4P is described as "oftentimes" as showing a transient increase in the early phase of hypoxia. However, this is not quantified. How often did this occur among the samples examined? How large is the transient increase when it occurs (Fig. 1A' error bars are not obvious on the colored background)? Second, the authors state that the P4Mx2-GFP puncta "often" became brighter after recovery. How often did this occur? No quantitation is provided.
2. The authors conclude that "PI4KIIalpha and Fwd contribute significantly to the maintenance of PM PI4P" (bottom of p. 7), yet they did not validate their RNAi knockdowns of these two genes, so they do not know whether it is one or both of these PI4Ks that contribute.
3. In Fig. 4B, a subset of the cells "show failed recovery of PM Lgl::GFP". However, some cells did recover. This average percentage of cells that recovered should be quantified, if possible.
4. In Fig. 7A, B, the bottom cell in each example lags behind the top cell in recovery of the MaLionR sensor. The frequency of observed cells in each class for 7A, B should be quantified.
5. In most cases, prior studies were referenced appropriately. However, two previous studies in Drosophila showing the effects of Sktl/PIP2 reduction on localization of polybasic proteins Lgl, Baz/Par-3 and Par-1 were not cited (relevant to the first paragraph of the Introduction, p. 3): Gervais et al., Development (2008), Claret et al. Curr Biol (2014). In addition, two studies showing the importance of Drosophila PI4KIIIalpha in synthesizing PM PI4P and PIP2 were not cited (relevant to the description of this enzyme, top of p. 6): Yan et al., Development (2011), Tan et al., J Cell Sci (2014). Data showing fwd null mutants are not lethal (relevant to top of p. 7) were published in Brill et al., Development (2000).
6. For the most part, text and figures are clear and accurate. However, there are quite a few typos and grammatical mistakes, as well as instances of lack of clarity in the writing that should be addressed. In addition, there are a number of improvements to presentation of data that would make the figures easier to understand. These are listed below.
7. Suggestions to improve presentation of data and conclusions are below.

Suggestions to improve the text:

p. 3, first paragraph, line 11, and second paragraph, line 2: combine references (Dong, Dong, Lu) with (Bailey and Prehoda, Hong)

p. 4, 8th line from bottom: PLC-PH is generally referred to as PLCdelta-PH in the literature and should be defined this way the first time "PLC-PH" is used

p. 5, 5th line from top: I believe the callout should be to Fig. 1A' rather than Fig. 1B

p. 5, first paragraph, line 2: replace "oftentimes" with "often" and provide quantitation (see above)

p. 5, first paragraph, line 4: delete "to" after "sensor"

p. 5, first paragraph, line 6: the claim of "often" should be quantified (see above)

p. 5, first paragraph, line 7: insert "the" before "intracellular"

p. 5, second paragraph: the extent of recovery of Lgl is less when Lgl-RFP is coexpressed with PLC-PH-GFP, potentially due to titration of PIP2 by PLC-PH; the authors should comment on this

p. 5, last line: the authors should provide information about the "targeted RNAi screen"; which genes were tested? did any others give relevant phenotypes? a table showing the results of the screen should be provided as supplementary information

p. 6, first full paragraph, line 4: change "cells" to "cell"

p. 6, second full paragraph, line 1: change "was" to "were"

p. 6, second full paragraph, line 4: delete "ref" at beginning of line)

p. 6, second full paragraph, lines 6 and 9: the callouts should be to Movies S5, not Movie S4

p. 7, first paragraph, line 4: explain briefly (either here or in the Methods) the "newly published multi-RNAi tools" reported by Qiao et al., 2018

p. 7, first paragraph, lines 9-10: suggest changing to "...(Fig. 3A,B); the latter...sensors being bound..."

p. 7, second paragraph, line 3: change name of Movie S7 to S6, and call out Movie S6 here

p. 7, second paragraph, line 5: replace "is" with "are"

p. 7, third paragraph, line 2: change name of Movie S8 to S7, and call out Movie S7 here

p. 7, last paragraph, line 2: "under normal conditions" is vague; replace "normal" with "aerobic" or "normoxic"

p. 8, first paragraph, line 4: replace "normal" with "aerobic" or "normoxic" for clarity

p. 8, first paragraph, line 8: change name of Movie S6 to S8, and call out Movie S8 here

p. 8, last line: callout to Fig. 5A appears incorrect; should be Fig. S1A instead

p. 9, first full paragraph, lines 4 and 5: callouts should be to Fig. 6A (line 4) and 6B (line 5)

p. 10, second full paragraph, line 1: insert "human" or "mammalian" after "cultured"

p. 10, second full paragraph, line 3: delete "Although" and start sentence with "For reasons unknown, out..."

p. 10, second full paragraph, lines 4-5: replace "in HEK293" with "to HEK293" and add a period after "present"; start new sentence on line 5 with "However, our data strongly..."

p. 10, third line from bottom: fix typo in "Discussion"

p. 11, line 9: replace "require" with "requires"

p. 11, line 10: perhaps the authors mean to refer to "PI and PIP kinases" rather than "PIP and PIP kinases"

p. 11, line 10: delete comma at end of line and replace with "and"

p. 11, first full paragraph, lines 3 and 4: replace "PI4KIIIa" with "PI4KIIIalpha" (two instances)

p. 11, first full paragraph, lines 5 and 7: the "m" in "Km" should be a subscript

p. 11, first full paragraph, line 6: what about PI4KIIIbeta? is the KmATP for this enzyme known?

p. 11, first full paragraph, lines 8-9: add "the before "intracellular PI4P pool" and replace "slower" with "more slowly" and "recovers faster" with "recover more quickly"

p. 11, first full paragraph, line 10: replace "loss" with "lose"

p. 11, last paragraph, line 1: insert "a" after "such"

p. 11, last paragraph, line 2: what is meant by "etc." is unclear; remove "etc." and include specific information related to what was reported in the literature (with proper references)

p. 12, line 3: why do the authors claim that the intracellular pool of PI4P is first synthesized by PI4KIIalpha? what about PI4KIIIbeta? their results do not distinguish between these enzymes

p. 12, line 4: insert "is" after "PM PI4P" and delete "is" after "PI4P pool"; also replace "the transfer" with "this transfer"

p. 12, line 5: delete "the before "intracellular"; delete the "is" after "(Dickson et al., 2014)" and replace with a comma"; replace "our data showing the loss" with "our data show loss"

p. 12, first full paragraph, line 2: "interconversion" does not require a hyphen"

p. 12, first full paragraph, line 5: why is PIP5K in red?

p. 12, first full paragraph, line 6: replace "attracted" with "recruited"

p. 12, first full paragraph, line 8: insert commas around "however"; also, "knockdown" does not require a hyphen

p. 12, last paragraph, line 5: delete "the" before "PI4P"

p. 12, last paragraph, lines 6-7: for the reader, please clarity the mechanism that was invoked to explain how PIP5K can make PIP2 from PI in E. coli (Botero et al., 2019)

p. 12, 5th line from bottom: replace "are" with "is"

p. 12, 4th line from bottom: start sentence with "In this regard," rather than "To this regard,"

p. 13, first paragraph, line 1: delete "the" and insert "a" before "large"

p. 13, first paragraph, last line: cannot conclude that components of PI4KIIIalpha are "highly interdependent" without testing effect of knockdown of PI4KIIIalpha on Rbo and TTC7, etc.; instead, can conclude that the data are consistent with all of the components acting in the same process; also, delete "the" before "proper"

p. 13, second paragraph, line 2: add comma after "localization"

p. 13, second paragraph, line 3: delete "the" before "PM PI4P"

p. 13, second paragraph, line 10: replace "is" with "are" before "supported"

p. 13, third line from bottom: should read "PIP2" rather than "PIP"

p. 13, second line from bottom: "must have" is awkward and unclear; perhaps change to "is predicted to have a"

p. 13, last line; replace "polybasic proteins or" with "polybasic and"

p. 14, first paragraph, line 1: delete "the" before "PM PI4P"

p. 14, first paragraph, line 3: replace "domains" with "motifs"

p. 14, first paragraph, line 6: insert "the idea" before "that Lgl"

p. 14, first paragraph, last line: replace "or" with "and"

p. 14, second paragraph, line 2: insert "of" after "plenty"

p. 14, second paragraph, lines 3-5: expand on this idea; what additional lipids could be important here? are there examples of other proteins that require these additional lipids?

p. 14, last paragraph, line 4: replace "While" with "Although"

p. 14, last paragraph, line 6: delete "the"

p. 16, line 5: insert comma after "chromosome"

p. 16, line 6: explain in brief what "pNP plasmid" is and how the multi-RNAi method works (what promoters drive expression of the shRNAs, how many shRNAs are included in the plasmid, etc.)

p. 16, line 7: insert "Additional stocks used were"

p. 16, lines 8-3: appropriate references should be included for each stock where available

p. 16, line 11: explain what UAS-AT1.03NL1 is

p. 16, line 12: fix typo in David Bilder's name

p. 16, line 14: italicize lgl::GFP

p. 16, lines 16-17: Gerry Hammond should not be listed as providing these constructs if he is a coauthor on the manuscript; appropriate references should be cited for these constructs

p. 16, line 18: briefly explain what the pGU vector is and how it works

p. 16, line 21: delete "the" before "young females"; also, the age of the females should be specified (newly eclosed? 1-3 days old? yeasted or not?)

p. 16, 7th line from bottom: replace "ensures" with "ensure"

p. 17, lines 3-4: sentence fragment "Images were further" is not complete

p. 17, line 10: "overexposure" does not require a hyphen; add a comma after "necessary"

p. 17, line 12: insert "in" before "visualizing"

p. 17, line 14: reword to say "imaged live in a temperature-controlled chamber at 37{degree sign}C."

p. 17, line 15: delete "which" after "serum"

p. 17, lines 16-17: replace "washout by replaced with chamber with normal serum" with "washed out by replacing with normal serum"

p. 17, 10th line from bottom: replace "were" with "was"

p. 17, second line from bottom: change "ROIs" to "ROI"

p. 18, line 5: replace "were" with "was"

p. 19, line 4: replace "the follicular cells" with "follicle cells"

p. 19, line 6: fix typo in "reoxygenation" (here and in legends to Figs 2-5, multiple instances)

p. 19, line 14: no RNAi is shown in Fig 1, so delete "(WT/RNAi); also, for consistency with other figure legends (e.g., Fig. 3), suggest changing text here to: "P4Mx2, PLC-PH (n=18,18), pLC-PH, Lgl (n=20,20), P4M2, Lgl (n=20,20)"

p. 19, 7th line from bottom: should read "hr:min:sec"

p. 19, lines 5-6 from bottom: title doesn't accurately reflect that PI4P doesn't appear to recover in WT control; why is this the case? recovery was observed in other experiments

p. 20, line 3: replace "follicular" with "follicle"

p. 20, line 5: replace "regulates" with "regulate"; also, it is not clear what is meant by "targeting and retargeting" (would be simpler to replace with "localization")

p. 20, Figs 4-6 legends: the numbers of samples examined in these experiments are missing (n=?)

p. 20, 10th line from bottom: replace "follicular" with "of follicle"

p. 20, 7th line from bottom: delete "changes"

p. 20, third line from bottom: replace "shown" with "seen"

p. 21, lines 6 and 8: replace "(B)" with "(A')" and "(C)" with "(B)"

p. 21, line 11: replace "(B)" with "(A)" and "(C)" with "(B)"

p. 21, lines 18 and 23: insert "cells" after "12"

p. 21, line 25: bold "(C')"

p. 22, line 3: incomplete phrase should be replaced with "regardless of the transient loss of PM PM PIP2" (or "...PM PLC-PH-RFP")

p. 22, line 4: replace "PLC-PH::GFP" with "PLC-PH::RFP"

p. 22, line 6: replace "express" with "expression"

p. 22, line 10: should this be "P4Mx2::GFP"?

p. 22, line 12: add comma before "PLC-PH::GFP"

p. 22, line 16: fix typo in "uncalibrated"; spell out what "AT[NL] sensor shows

pp 23-26: for clarity, delete word "samples" in legends to Movies S2-S16

p. 23: fix typo in "Lgl::mCherry" in legends to Movies S2 and S3

p. 24: change "Movie S09" to "Movie S9"

p. 26: change "Time intervals is" to "Time intervals are" in legends to Movies S17-S19

Suggestions to improve the figures and movies:

• show single-color images in grayscale, which is easier to see on black and helpful for colorblind readers (applies to all figures except Fig. S3); movies and merged still images should be shown in green and magenta for colorblind (not sure if channels in movies are difficult to change)
• replace colored labels on black boxes with colored labels on white background (Fig. 5A (left), Fig. 5B-D (top), Fig. 6A (left), Fig. 7A-C (left), Fig. S1A (left), Fig. S1C (top), Fig. S2A (left), Fig. S4 (left))
• provide scale bars throughout (Figs 2-7, S1-S4)
• replace pale colored boxes under labels for "hypoxia" and "air" with slightly darker boxes (Fig. 1A-C, Fig. 2A, Fig. 3A, Fig. 5A', Fig. 6A', Fig. S2B)
• provide vertical lines similar to those in Fig. 4A' in all of the time-course graphs and/or making the background colors slightly darker (Figs 2A', 3A', 5A', 6A'); also make the error bars darker (Fig. 1A'-C', Fig. 4, Fig. 5A', Fig. S2B)
• for consistency, label PM index graphs in Fig. 4 and Fig. S2 as Fig. 4A' and Fig. S2A'
• why are some of the PM index graphs labeled "PM index" and others labeled "PM index-1" on the Y-axis? this should be explained or changed for consistency
• "blank diamonds" described in figure legend for Fig. 7B' are barely visible when printed
• Fig. 7C is mislabeled (MaLionR label should be replaced with PLC-PH-RFP)
• in Fig. S3A, it would help to know the size of the cells (i.e., how many were present in the area examined)
• movies should be referred to in order (current Movies S7 and S8 should be renamed S6 and S7, and current Movie S6 should be renamed S8)
• in Movie S19, "PLC-PH::RFP" is mislabeled "PLC-PH::GFP" (both P4MX2 and PLC-PH are labeled GFP in the movie)

Significance

Overall, the authors do a nice job of showing that hypoxia leads to previously unappreciated effects on levels of PM PI4P and PIP2, resulting in loss of PM association of proteins important for normal cellular physiology. This finding is quite novel. Moreover, the authors provide insight into the identity of the PI4Ks that are responsible for regenerating PM PIP2 following return to normoxia. Their analysis of the dynamics of these changes provides multiple interesting insights, including the potential roles of intracellular pools of PI4P in replenishing PM PIP2 and the observation that intracellular accumulation of PIP2 is occasionally observed in association with the appearance of intracellular PI4P puncta, suggesting a novel route for PIP2 replenishment in response to hypoxic stress. Their results will provide the basis for future studies examining the cellular mechanisms involved. This study will be of interest to those studying phosphoinositide biology as well as cellular responses to hypoxic stress and recovery, such as occur during ischemia and reperfusion. Reviewer expertise: Drosophila molecular genetics, cell biology, developmental biology, phosphoinositides, PIP pathway enzymes, PIP effectors

Referees cross-commenting

This session includes the comments of all reviewers.

Reviewer 3

I agree with reviewer #1 that the authors did not do a good job of clarifying what they and others had previously shown, and I must confess I didn't carefully examine their previous papers carefully enough before preparing my review. In fact, they previously showed that hypoxia affects localization of Dlg at the plasma membrane and that its recovery depends on PI4KIIIalpha and PIP2 (Lu et al., Development 2021). This is in addition to their previous data showing effects of hypoxia on Lgl (Dong et al., J Cell Biol 2015). Thus, less of the information in the current manuscript is novel than I thought when I initially read it.

I also agree with reviewer #2 that they need to do a better job of citing the relevant literature and considering the possibility that hypoxia and reduced levels of ATP might affect many different enzymes. In addition, as suggested by reviewer #1, it seems important

Reviewer 1

I agree with what Reviewer 3 is suggesting and with reviewer 2 that the authors should do a better job of citing all of the relevant literature. I also appreciate the detailed edits provided by Reviewer 3 - it was very generous of them to do this.

Reviewer 2

The points raised by reviewer 1 and 3 with regard to the citing or prior work (from the authors or other labs) also applies to their citing of literature on PI and PI4K signalling. Here too citing or prior work has been less than satisfactory making it difficult to do this.

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

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

This manuscript describes the effect of hypoxia on the levels of PI4P and PI45P2 , two key PPIs that are enriched on the inner leaflet of the plasma membrane. These PPIs are synthesized by the sequential phosphorylation of PI by a PI-4 kinase and subsequently a PI4P 5 kinases, both of which use ATP. The relevant PI-4 kinase at the plasma membrane, PI4KIIIa has been conclusively identified previously in mammalian cells by the DeCamilli lab (Nakatsu et.al JCB 2012) and its role in regulating the synthesis of PI4P and PI(4,5)P2 in two Drosophila cell types in vivo shown by two previous studies. Balakrishanan et.al J.Cell Sci 2018 (photoreceptors during PLC signalling) and Basu et.al Dev.Biol 2020 ( in multiple larval cell types ). PI4KIIIa has been shown to exist as a complex of the enzymatic polypeptide, EFR3 and TTC7. The studies by Nakatsu, Balakrishnan and Basu have shown the importance of the complex subunits is regulating PI4P and PI45P2 levels in cultured mammalian cells and Drosophila cell types in vivo.

In the present study, Lu et. al build on their previous work showing that the polarity protein Lgl undergoes hypoxia induced translocation. They show that hypoxia also induces loss of PI4P and PI45P2 at the plasma membrane in these cells correlated with loss of Lgl localization to the PM. The manuscript then goes on to establish the requirement of the PI4KIIIa complex in regulating Lgl localization as well as PI4P and PI45P2 levels at the plasma membrane during hypoxia and the subsequent recovery of these at the plasma membrane.

The strength of the manuscript is twofold.

• (i) The work is done to a high technical standard and the investigators have carried out the measurements of LGL localization, PI4P and PI45P2 levels along with simultaneous measurements of ATP levels in vivo. The work would be strengthened further if the authors could show the level of depletion of PI4K isoforms or PI4KIIIa complex subunits units induced in ovarian tissue under their experimental conditions by the GAL4 drivers used in this study. This is not a persnickety detail as RNAi lines can have very different effectiveness in Drosophila ovarian tissue compared to other fly cell types. This point is, in particular, important in cases where an RNAi line is being used and the conclusion is a lack of impact on a phenotype being studied.
• (ii) A second strength is that the authors now illuminate a further in vivo cell type where the function of the PI4KIIIa complex in regulating PI4P and PI45P2 levels. This adds to the earlier work of Nakatsu, Balakrishnan and Basu.

A key difficulty with the current story is the lack of specificity of the phenotype they demonstrate under hypoxia. Of course, hypoxia is expected to deplete cellular ATP levels but PI4KIIIa is not the only enzyme that this lack of ATP will impact. There will be dozens or more other kinases, both protein and lipid kinases whose function will be impacted by the drop in ATP levels. Therefore, it is hard to attribute a specific/particular role to the PI4KIIIa complex under these conditions. The mislocalization of LGL::mCherry while correlated with PI4P and PI45P2 levels at the plasma membrane may be just that- a correlation. It is quite possible, indeed likely, that the mislocalization of LGL-mCherry under hypoxia conditions is due to the reduction of the activity of another lipid or protein kinase due to the drop in ATP levels due to hypoxia (PKC is a possibility too).

Minor comments:

The authors must reference all published work on the PI4KIIIa complex in the literature. Some of it is excluded in the present version

The Drosophila work, particularly cell types used, etc are not accessible to people who are not fly experts. This should be done.

Significance

Adds to knowledge on the PI4KIIIa complex. Builds on existing knowledge in the PI4KIIIa field and maybe also cell polarity field.

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

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

Summary:

This manuscript takes a closer look at how hypoxia affects the accumulation of PI4P and PI4,5P2 (PIP2) in the plasma membrane of Drosophila ovarian follicular epithelial cells and how ATP depletion similarly affects the localization of the same phospholipids in HEK293 cells. They demonstrate that hypoxia results in the reversible loss of plasma membrane (PM) association of both lipids, with PIP2 disappearing ahead of PI4P, and recovering more slowly than PI4P when oocytes are returned to normoxia. They also show that the intracellular vesicular pools of PI4P are depleted ahead of the PM pools and the PI4P recovery occurs first in PM, then in the vesicles. They show that the disappearance and recovery of the polarity protein Lethal giant larvae (Lgl) parallels that of PIP2 during hypoxia and subsequent normoxia, with a very slight delay. The authors then go on to show the RNAi knockdown of the PM enzyme (PI4KIIIa) that phosphorylates PIP delays the recovery of PI4P at the membrane, with recovery first occurring in the vesicular pools. This knockdown also delays the recovery of PIP2 and, as with recovery of PI4P, the recovery of PIP2 now occurs first in vesicular pools. Lgl recovery follows that of PI4P and PIP2 with RNAi knockdown of PI4KIIIa. The knockdown of all three of the enzymes that phosphorylate PIP to generate PI4P delays recovery of PI4P, PIP2 and Lgl at the membrane even more. The authors show that proteins required for the PM localization of PI4KIIIa have similar effects on the recovery of PM PI4P, PIP2 and Lgl (with delays and recovery of vesicular pools before PM pools). Independently, the authors show that ATP depletion in HEK293 cells result in similar reversible depletion of PI4P, PIP2 and Lgl from the PM. From these studies and their previous findings, the authors conclude that pools of PI4P and PIP2 are likely rapidly turned over in the membrane even during normoxia and that this rapid recovery is dependent on the PM localized enzyme that phosphorylates phosphoinositol.

Major comments:

Overall, the data are beautifully presented; it is quite helpful to have a video of each experimental treatment showing the corresponding response of all three molecules that are being monitored. Signal quantification over time is carefully documented. With the exception that a link between hypoxia and depletion of ATP has not been demonstrated here, the key conclusions are convincing. However, as pointed out below (in the significance section), some of the major points have already been published by this group. Their conclusion that hypoxia induces acute and reversible reduction of cellular ATP levels (which are then proposed to affect the activities of the enzymes required for PI4P and, consequently, PIP2 production) was not shown. They did demonstrate that acute depletion of ATP had the same consequences on PM phospholipids as acute hypoxia (in HEK293 cells). And, indeed, it makes sense that hypoxia could affect enzymes required for ATP synthesis, but the authors would have to show that acute hypoxia results in acute reduction in cellular ATP pools to make the links they suggest. This is something they should be able to do in the HEK293 cells now that they have their ATP sensor. Just to note, this group did show that hypoxia can reduce levels of ATP in Drosophila oocytes in their previous paper (Dong et al., 2015, Figure S3), but it is unclear if this is reversible and happens in the time frame of the experiments presented in this current manuscript.

My suggestions are the following:

1. The authors need to make it absolutely clear what was already known, including the following: (A) hypoxia reversibly affects PM pools of PI4P, PIP2, and Lgl (and other membrane associated proteins), (B) that hypoxia can affect ATP levels in Drosophila oocytes (although these previous studies do not show anything about the dynamics) and (C) that reducing ATP levels affects PM pools of PI4P, PIP2 and Lgl.
2. They should demonstrate that acute hypoxia and return to normoxia has acute and reversible effects on cellular ATP levels - they now have the tools to do this, at least in HEK293 cells.

Minor comments:

The manuscript is too long and the discussion unnecessarily repeats everything already presented in the results. The authors should find a way to streamline the discussion.

N values should be given for all figures and experiments, and the N=23/24 versus N=24/24 needs to be explained the first time it is used.

There are a few mismatches in terms of plural nouns and singular verbs and vice versa sprinkled into the manuscript, so some careful editing would be useful.

Significance

I was initially quite excited about the novelty of their findings and the potential insight into the dynamics of PM pools of the two phospholipids that are critical to cell polarity and that play important signaling roles. However, at least a subset of their conclusions were either published in their earlier work or do not necessarily follow from what they have done in this manuscript. Their statement that hypoxia in Drosophila induces acute and reversible depletion of PM PI4P and PIP2 was presented in a previous publication (See Figure 8 of Dong et al., 2015).

This manuscript would appeal to an audience interested in the mechanisms of cell polarity and phosphoinositide signaling.

I am a Drosophila developmental geneticist quite familiar with the topics that this paper addresses.

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

Major comments:

Are the key conclusions convincing?

We discuss 4 key conclusions.

__# 1 __A PRC of the segmentation clock was constructed.

Although the authors have produced an interesting phase map, the regulation function F(\phi) of the circle map does not give the phase response curve (PRC) (Hoppensteadt & Keener 1982, Guevara & Glass 1982). This holds only when the system is stimulated with very short pulses (ideally Dirac delta), but the experimental pulses here are a quarter of the intrinsic period.

There are several definitions of the PRC (Dirac pulses PRCs, linear PRCs, etc.). We use the general definition from Izhikievich, 2007: “In contrast to the common folklore, the function PRC (θ) can be measured for an arbitrary stimulus, not necessarily weak or brief. The only caveat is that to measure the new phase of oscillation perturbed by a stimulus, we must wait long enough for transients to subside“.

The corresponding equation from Izhikievich (section 10.1.3) is

PRC(θ)= θ_new-θ

which is equivalent to our Equation 1.

Hence, the key assumption we make is that after perturbing the system, we are back on the limit cycle as pointed out by Izhikievich. We think this is a reasonable assumption, because the perturbation we impose is relatively weak, despite pulsing for almost one quarter of the intrinsic period. The concentrations of DAPT we used in this current study are just enough to elicit a measurable response, and further lowering the concentration does not result in entrainment within our experiment time (0.5uM, Figure S7B in submitted version of the manuscript). Additionally, we previously reported that periodic pulsing with 2uM DAPT did not result in change of the Notch signaling activity with respect to control samples (Sonnen et al., 2018). Along similar lines, the DAPT drug concentrations we used are much lower compared to what has been used in previous studies aiming to perturb signaling levels, e.g. 100uM and 50uM used in study of segmentation clock in zebrafish embryos (Özbudak and Lewis, 2008 and Liao et al., 2016, respectively), and 25uM used in study of the segmentation clock in mouse PSM cells (Hubaud et al., 2017). Combined, we reason that we apply weak perturbations that allow to extract the PRC of the segmentation clock during entrainment. Additional evidence that indeed we have revealed a meaningful PRC is provided below, please see our response to point #3.

__# 2 __Furthermore, in eq. 1 T_ext must be the winding number, and the modulus must be in units of

phase, either one or two pi, for the circle map to be correct. Thus, calling the measured response of the system a PRC is not convincing.

We thank the reviewer for pointing this out. We indeed rescaled everything to express the PRC in units of phase. We made this more explicit and updated equations throughout the text.

__# 3 __The system is being entrained. Technically, It would also be easier to get the stroboscopic maps

in the quasi-periodic regime since all the points in the circle will be sampled. Since no quasi-periodic response was demonstrated, the claim of entrainment is not convincing.

While, in principle, PRC can be indeed obtained from responses in the “quasi-periodic” regime, such an approach is, in practice, challenging due to the intrinsic noise. The closest approximation to this is the phase response after the first pulse, that we reproduce below and compare to our inferred PRC, where we indeed clearly see a high noise level. Nevertheless, also the PRC based on the 1st pulse is in agreement with the PRC we derived from the entrainment data.

In the entrained regime, one can get a much more reliable estimate of the phase response despite the noise. The level of noise in the stroboscopic map lowers as the samples approach entrainment (Figure S12), and the entrainment phase itself is a reliable statistical quantity that can be used to infer regions of the PRC as the detuning is varied.

In addition, and maybe even more importantly, we identify several key features characteristics of entrainment, such as the change of entrainment phase as a function of detuning (Figure 7, Figure S6-S7 in submitted version of the manuscript) and the dependency of the time to entrainment as a function of initial phase (Figure 6). While additional features can be linked, in theory, with entrainment, i.e. period-doubling, higher harmonics (Figure 5), quasi-periodicity, we do not agree with the reviewer that all of these need, or in fact, can be found in the experimental data, in particular because of the influence of the noise. Conversely the positive experimental evidence that we provide for the presence of entrainment, combined with the theoretical framework we develop, justifies, in our view, the conclusions we make.

__# 4 __The response of the system to external pulses is compatible with a SNIC. This is compatible, but

it is equally compatible with other explanations. Assuming that the PRC is the same as the regulation function F(\phi), the PRC in Kotani 2012 (PRL 2012 fig. 3C) would be a similar shape as that shown by the authors. Similar models to that in Kotani et al., have been studied, but a SNIC has not been found (an der Heiden & Mackey 1982). It is relatively straightforward to construct a phenomenological model with a SNIC, but having underlying biological insight is not guaranteed. No argument for choosing a SNIC is given, so this emphasis of the paper is not convincing.

It is true that the mapping of PRCs to oscillators is undetermined, in the sense that many systems could potentially give rise to similar PRCs. That said, there is value in parsimonious models, which often generalize very well despite their simplicity. This explains why in neuroscience, constant sign PRCs are generally associated with SNIC. There is a mathematical reason for this : 1-D oscillators with resetting (such as the quadratic fire-and-integrate model) are the simplest models displaying constant sign PRCs, and are the “normal” form for SNICs. In other words, SNIC bifurcations are among the simplest ones compatible with constant sign PRCs, and we think it is informative to point this out. In our manuscript, we go one step further by actually fitting the experimental PRC with a simple, analytical model that allows us to compute Arnold tongue for any values of the perturbation (contrary to more complex models).

Other models such as Kotani 2012 can display similar PRC shapes, but they are of mathematically higher complexity, and furthermore it is not clear how such systems might behave when entrained. For instance that model in particular uses delayed differential equations, and as such contains long term couplings, so that a perturbation might have effects over many cycles, which is not consistent with the hypothesis we here make of a relatively rapid return to the limit cycle. Furthermore, for more complex models, PRCs are analytical only in the linear regime, while our model is analytical for all perturbations. That said, we agree that other types of oscillators can be associated with constant sign PRCs, and we have given more details in this part, in particular we better emphasize the Class I vs Class II oscillators as a way to broaden our discussion on PRC, and emphasize the “infinite period” bifurcation category which is more intuitive and further includes saddle node homoclinic bifurcations.

__# 5 __The work demonstrates coarse graining of complex systems.

This conclusion is correct, but coarse graining theory-driven analysis and control of dynamical systems has been established for many years. What is new here is that it is applied specifically to the in vitro culture system of the mouse segmentation clock.

We agree it is new to successfully apply coarse-graining analysis and, importantly, control, to the in vitro culture system of the mouse segmentation clock. We also agree that such an approach has been pioneered and established for many years, especially in (theoretical) physics, but indeed, the key question is whether and how this can be applied to complex biological systems. Insights coming from theoretical considerations on idealized physical systems might not necessarily apply to biology, as already pointed out by Winfree.

There are still very few examples in biology with coarse graining similar to what we do here. We think there is immense value in demonstrating that quantitative insights, and control of the biological systems, can be obtained without precise knowledge of molecular details, which is still counter-intuitive to many biologists. In this sense, we think our report will be of interest to both colleagues within the field of the segmentation clock and also to anyone interested to in the question, how theory and physics guided approaches can enable novel insight into biological complexity.

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

Following on the points above, each of these needs to be corrected or re-done, and/or the conclusions need to be modified accordingly.

We have modified the manuscript in response to all those points.

# 6 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 the authors wish to make the strong claim of determining a true PRC, Dirac delta-like perturbation needs to be applied, or approximated by short time duration pulses compared to the intrinsic period.

Please refer to our response to point #1 and #3..

# 7 *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.

It's not clear to this reviewer if it is feasible to deliver a very short pulse and record a response. But this may not be relevant, see above.

Please refer to our response to point #1 and #3 .

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

Yes.

Are the experiments adequately replicated and statistical analysis adequate?

Yes.

Minor comments:

Specific experimental issues that are easily addressable.

No issues.

Are prior studies referenced appropriately?

Yes.

# 8 Are the text and figures clear and accurate?

Figure 1D illustrates how a PRC should be obtained, but doesn't show the experimental protocol applied in the paper.

Figure 1D is a general introduction on the phase description of oscillators and phase response. It demonstrates how a perturbation can change the phase and is not supposed to represent the experimental protocol. We describe how data are analyzed and how phases are extracted in Supplementary Note 1.

__# 9 __In Figure 5B, 10 uM DAPT, the traces are already synchronized before the pulse train starts,

which makes the subsequent behavior difficult to interpret.

It appears here that by chance, the samples were already almost synchronized. We notice however that the establishment of a stable rhythm with the pulses (which here is not a multiple of the natural period) supports entrainment, and is already evident when looking at the timeseries with respect to the perturbation. The temporal evolution of the instantaneous period further confirms this, showing a change in period close to ½ zeitgeber period (which is very different from the natural period of ~140 mins). This also relates to point #35, in reply to both comments we have further expanded this figure to better show the 2:1 entrainment, adding statistics on the measured period and period evolution for a zeitgeber period of 300 mins.

# 10 Do you have suggestions that would help the authors improve the presentation of their data and Conclusions? The text includes several paragraphs reviewing broad principles of coarse graining and making general conclusions. This is confusing, because, as mentioned above, there is no new general advance in this paper. The interesting contributions here are specific to the applications to the segmentation clock, and the text should be focused on this aspect.

As commented above for #3 , we respectfully disagree that there is no “new general advance” in this paper. It is far from obvious that a complex ensemble of coupled oscillators implicated in embryonic development would be amenable to such coarse-graining theory. Of note, we still do not have a full understanding of neither the core oscillators in individual cells, nor what slows these down and eventually stops the oscillations, and multiple recent works suggest that both phenomena are under transient nonlinear control (e.g. our own work in Lauschke 2013). It is remarkable that despite this lack of detailed mechanistic insight, general entrainment theory can be applied to the segmentation process at the tissue level. We further show that classical entrainment theory alone is not sufficient to account for the experimental findings. Specifically, we need to account for a period change that we interpret as an internal feedback, an insight that would be impossible without our coarse-graining approach. While the results might of course be specific to the segmentation process, we think our approach motivated by coarse-graining theory and leading to new insights into the process is of general interest. We tried to make these points explicit in our conclusion.

Reviewer #1 (Significance (Required)):

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

Description of the complex mouse segmentation clock in terms of a simple model and its PRC is an interesting, original and non-trivial result. The proposal that the segmentation clock is close to a SNIC bifurcation provides a consistent dynamical explanation of slowing behavior that has been recognized for some time, but not fully understood. This proposal also raises a hypothesis about the behavior of the underlying molecular regulatory networks, which may be tested in the future. The increase or decrease of the intrinsic period due to the zeitgeber period is not expected from theory, pointing to structures in internal biochemical feedback loops, an idea which again may be tested in the future. Also surprising from a theoretical perspective, the spatial gradient of period in the system persisted after entrainment. Although the categorization of the generic behavior is interesting, by its nature there is little from this that might give a typical developmental biologist any conclusions about pathways or molecules. The successes and limits of the theoretical description do nevertheless focus future attention on interesting behaviors.

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

Such an analysis of the segmentation clock is based strongly on the experimental system and results in Sonnen et al., 2018, and goes well beyond it in terms of the dynamical analysis. It provisionally categorizes the mouse segmentation clock as a Class I excitable system, allowing its dynamics at a coarse grained level to be compared to other oscillatory systems. In this aspect of simplification, it is similar to approach of Riedel-Kruse et al., 2007 who used a mean-field model of oscillator coupling to explain the synchrony dynamics observed in the zebrafish segmentation clock in response to blockade of coupling pathways, thereby allowing a high-level comparison to other synchronizing systems.

It is interesting the reviewer sees similarities with the work of Riedel-Kruse et al, which uses a mean-field variable Z that corresponds to a classical approach, as described in Pikovsky’s textbook, to quantify synchronization of oscillators. In our view, while of course we work in the same context of coupled oscillators in the PSM, our approach based on perturbing and monitoring the system’s PRC in real-time provides a novel strategy to gain insight. This is evidenced by the fact that our quantifications of synchronization and insight into the PRC is the basis to exert precise control of the pace and rhythm of segmentation.

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

Developmental biologists, biophysicists

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.

Developmental biology, somitogenesis, dynamical systems theory, biophysics, cell signaling

---

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

Summary: This is a beautifully elegant study that tests how previously published theoretical predictions about entraining nonlinear oscillators applies to a biological oscillator, the segmentation clock. The authors use a combination of state of the art experimental techniques, signal processing and analytical theory to reach a series of interesting and novel conclusions.

They show that the segmentation clock period can be entrained through Notch inhibitor (DAPT) pulses acting as an external clock (referred to as zeitgeber) using a previously developed and sophisticated microfluidic perfusion system. Pulsing DAPT every 120 to 180min can change the internal clock period while entrainment beyond this range leads to higher order coupling to the zeitgeber period, i.e. entrainment of every other pulse. They then perform entrainment experiments where the concentration of DAPT is changed to elicit a change in the strength of interaction between the internal clock and the external stimulus (referred to as zeitgeber strength); interestingly at low strength response to entrainment is more variable leading to entrainment occurring in some samples while others remain unaffected (Figure 4A); overall, higher concentration leads to faster entrainment (Figure 4C). The experimental data is then analysed using stroboscopic maps to reveal that a stable entrainment phase shift is achieved between the internal clock and the external zeitgeber. Phase response curve (PRC) analysis indicates that the system response is not sinusoidal but predominantly characterised by negative PRC, a behaviour consistent with saddle-node on invariant cycle (SNIC); it also reveals that the intrinsic period changes in a non-linear way and that this effect is reversible when external stimulation stops. Finally, a theoretical model is proposed to represent the segmentation clock as a dynamical system; this is based upon Radial Isochron Cycle with Acceleration (ERICA), an extension motivated by the PRC analysis results which are incompatible with a Radial Isochron Cycle (RIC); this model has predictive capability and could be used to design new control strategies for entrainment of the segmentation clock.

This study makes a series of key conclusions which are of particular importance in understanding the dynamic response of a biological oscillators. Firstly, given it's the characteristics of the dynamic response to entrainment, the segmentation clock is likely close to a SNIC bifurcation and this can explain the tendency for relaxation of the period over time. Secondly, the clock period was changed in a non-linear way in the direction of the zeitgeber period, a finding which is interpreted to indicate the presence of feedback of the segmentation clock onto itself, potentially via Wnt. This makes an excellent prediction that if tested experimentally would greatly improve the impact of the study. It is also noted that the entrainment of the segmentation clock does not abolish spatial periodicity and phase wave emergence suggesting that single cell oscillators can adjust to periodic perturbation while maintaining emergent properties. This is also a significant result that would need to be followed up with experiments and computation however would be best suited to a separate study.

Major comments:

__# 12 __The coarse graining is a major point that would need to be clarified since the rest of the analysis

and theoretical modelling in the paper flow from this. Firstly, the interpretation of the schematic in Figure 1A on experimental data collection is not immediately obvious to the reader, lacks a clear flow between the different panels or steps (which could be numbered for example) and does not have a legend to indicate the different colour mapping.

We are grateful to the reviewer for this comment. We have implemented in Figure 1A all the changes suggested by the reviewer: we numbered the different steps and have added a colour mapping. In addition we have rephrased the caption of Fig 1A to better connect the experimental steps.

__# 13 __Secondly, Figure 2A which explicitly addresses coarse graining is not clear enough. Is the

message here that by excluding the inner parts of the sample with a radial ROI, a similar dynamic response is observed over time?

Yes, indeed this is the point and we have adjusted the figure and text to explain this better. Our goal is to focus on the quantification of segmentation pace and rhythm. This is best captured by reporters such as LuVeLu, which has maximum intensity in regions where segment forms, and which dynamics is known to be strongly correlated to segmentation (Aulehla et al., 2007; Lauschke and Tsiairis et al., 20132). The global ROI is thus expected to precisely capture these segmentation and clock dynamics and we have now included more validation data and have also edited the text to make this very important point clearer:

“To perform a systematic analysis of entrainment dynamics, we first introduced a single oscillator description of the segmentation clock. We used the segmentation clock reporter LuVeLu, which shows highest signal levels in regions where segments form \cite{Aulehla_A_2007}. Hence, we reasoned that a global ROI quantification, averaging LuVeLu intensities over the entire sample, should faithfully report on the segmentation rate and rhythm, essentially quantifying 'wave arrival' and segment formation in the periphery of the sample.”

Figure 2A indeed shows that the dynamics (from the timeseries) is very similar when considering the entire field of view (global ROI) or when considering only the periphery of the 2D-assay (excluding central regions). We modified Figure 2A to clarify this point by indicating each measurement as either global ROI or global ROI minus the diameter of the excluded circular region (e.g. global ROI - 50px). We also emphasized in the caption that timeseries are obtained using global ROI, unless otherwise specified. We included a link (https://youtu.be/fRHsHYU_H2Q) in the caption to a movie of 2D-assay subjected to periodic pulses of DAPT (or DMSO) and corresponding timeseries from global ROI.

Since the inner part of the sample corresponds to the posterior side how do we interpret similarities and differences between signals with different ROIs?

As stated above, the global ROI measurements essentially capture the signal at the periphery where segments form and faithfully mirrors segmentation rate and rhythm. We have now included a comparison to the center ROI, also in response to reviewer’s comments, see our response #34.

The result shows that the period and PRC in the center matches the one found in the periphery, i.e. global ROI. We have shown previously that center and periphery differ in their oscillation phase by 2pi, i.e., one full cycle (Lauschke et al., 2013). We interpret these findings as confirmation of our analysis strategy, i.e. the global ROI allows a very reproducible, unbiased quantification that reports on segmentation clock and period.

__# 14 __A quantitative analysis of essential coarse-grained properties such as period and amplitude

should be performed for different ROIs and across multiple samples. As this effectively masks any spatial differences, limitations of this approach should be clearly stated in the Discussion. For example in lines 466-470 where it is difficult to interpret the slowing down tendency and relate back to single cell level.

As outlined in our response to comment #13 and also #34, we chose an analysis that allows to determine the segmentation pace and rhythm, i.e. segment formation, which is well captured by LuVeLu signal and a global ROI analysis. We agree that a spatially resolved analysis of dynamic behaviour is important (and indeed a gradient of amplitude might be relevant in such context), but we think this is beyond the scope of the current study focused on the system level segmentation clock behaviour. We have revised the discussion as suggested by the reviewer to make this point approach and the need for future studies clearer.

__# 15 __The functional characterisation of the sample using LFNG, AXIN2 and MESP2 is unclear. The

images included in Figure 2D representing expression observed when tissue explants are grown within the microfluidic chip are difficult to interpret and would require a more detailed description of anterior-posterior, pillars etc; it is also difficult to view the bright-field since it is presented as a merged image.

It is particularly difficult to see the somite boundaries for the same reason. In lines 113-117 the authors state that the global oscillation period matches the periodic boundary formation. How do we reach this conclusion from these images? What is the variability between samples?

If these two issues would be addressed it would increase confidence in the coarse graining argument and thus would strengthen the importance of the findings in the study.

We thank the reviewer for this feedback, and we have added more quantifications to address this point directly in the modified Figure 2. Importantly, we added the quantification of the rate of segmentation in multiple samples based on segment boundary formation (new Figure 2D) and compared this to the global ROI quantifications using the reporter lines LuVeLu. This data provides clear evidence that the quantification of global ROI reporter intensities closely matches the rate of morphological segment boundary formation. In addition, we show that segment formation and also Wnt-signaling oscillations (Axin2-Achilles) and the segmentation marker Mesp2 (Mesp2-GFP) are all entrained to the zeitgeber period. We have also revised the text to clarify this important validation of our quantitative approach.

In addition, we provide, in the revised Figure Suppl. 2, details of entrained samples, focusing on the segmenting regions. The brightfield and reporter channels were separated, emphasizing the segment boundaries and the expression pattern of the reporters. For ease of visualization, these samples were also re-oriented so that the tissue periphery (corresponding to anterior PSM) is at the top while the tissue center (corresponding to the posterior PSM) is at the bottom. This now additionally better shows the localization of the different reporters with respect to the segment boundary. We also included supplementary movies showing timelapse of samples expressing either Axin2-GSAGS-Achilles or Mesp2-GFP that were subjected to periodic DAPT pulses, with their respective controls.

Several minor points could be addressed to improve the manuscript and are listed below:

# 16 Figure 1 A the colormap and axes for the oscillatory traces should be defined

We thank the reviewer, and we have modified the figure accordingly (related to point # 12). A colormap and axes for the illustrated timeseries are now included.

# 17 Strength of zeitgeber is not defined and there is no analytical expression provided; how does it

relate to DAPT concentration? Is the fact that low DAPT concentration corresponds to weak strength expected or is it a result?

Zeitgeber strength generally refers to the magnitude of the perturbation periodically applied to an oscillator. With DAPT pulses, our expectation was that both the duration of the pulse and the drug concentration could influence the strength. Practically, the pulse duration was kept constant for all experiments and the concentration was varied. We thus expected that DAPT concentration would indeed be correlated to zeitgeber strength. We have discussed multiple evidence supporting this assumption in the main text, and this is indeed a result. In particular, as explained in the section “The pace of segmentation clock can be locked to a wide range of entrainment periods”, higher DAPT concentration gives rise to faster and better entrainment, as expected from classical theory. In the context of Arnold tongue, weaker zeitgeber strength corresponds to narrower entrainment region, which is experimentally observed (Fig 8F, showing regions where the clock is entrained).

From a modelling standpoint, Zeitgeber strength corresponds to parameter A which is the amplitude of the perturbation. Possible zeitgeber strength was inferred from the model by matching the experimental entrainment phase with that obtained from the model isophases. As explained in Supplementary Note 2, we tested four concentrations of DAPT (0.5, 1, 2, and 3 uM) respectively corresponding to A values of 0.13, 0.31,0.43, 0.55. As we can see, those A values are not linear in DAPT concentrations, which is expected since multiple effects (such as saturation) can occur.

__# 18 __In some figures it looks like the amplitude of oscillations may change with DAPT concentration

and hence zeitgeber strength? Is this expected?

We have not systematically analyzed the amplitude effect and have, intentionally, focused on the period and phase readout as most robust and faithful parameters to be quantified. Regarding the amplitude of LuVeLu reporter, we are cautious given that it is influenced, potentially, by the (artificial) degradation system that we included in LuVeLu, i.e. a PEST domain. This effect concerns the amplitude, but not the phase and period, explaining our strategy.

That said, we agree with the referee that DAPT concentrations might change the amplitude of oscillations. Such change could even play a role in the change of intrinsic period (in fact a similar mechanism drives overdrive suppression for cardiac oscillators, Kunysz et al., 1995). But since the change of period can be more easily measured and inferred, we prefer to directly model it instead of introducing a new hypothesis on amplitude/period coupling, at least for this first study of entrainment.

__# 19 __Figure 2A including the black area creates confusion and it is unclear which ROI is used in the

rest of the study; consider moving this to a supplementary figure perhaps

We thank the reviewer for this feedback (related to point #13), and we have modified the figure accordingly. As we responded to point # 13: We modified Figure 2A, by indicating each measurement as either global ROI or global ROI minus the diameter of the excluded circular region (e.g. global ROI - 50px). We also emphasized in the caption that timeseries are obtained using global ROI, unless otherwise specified.

__# 20 __What type of detrending is used in Figure 2 and throughout (include info in the figure legend)?

We used sinc-filter detrending, described and validated in detail previously (Mönke et al., 2020), as specified in Supplementary Note 1: Materials and methods > H. Data analysis > Monitoring period-locking and phase-locking: In this workflow, timeseries was first detrended using a sinc filter and then subjected to continuous wavelet transform. We thank the reviewer for pointing out that this detail is lacking in the figure captions, and we have modified the captions accordingly.

__# 21 __Figure 2D merged images are difficult to read/interpret (see major comments)

We thank the reviewer for this comment, and we have modified the figure accordingly (please see response to related point #15).

__# 22 __Kuramoto order parameter is used to quantify the level of synchrony across the different samples

however it is not defined in the text. Is it also possible to assess variability in each sample? For example how quickly does entrained occur in each sample? How faithfully the peaks of expression beyond 80min (to exclude initial unsynchronised state) match with zeitgeber time? This would help make the point that weak strength leads to a more variable response which is an interesting finding.

We have now added a mathematical definition of the Kuramoto parameter in Supplementary Note 1.

A high order parameter corresponds to coherence between samples, as also elaborated in respective figure captions (e.g. in the caption for polar plots in Figure 4D).

In terms of variability in response to entrainment, we thank the reviewer for the comments, which has prompted us to perform an additional analysis, now included as Figure S13 in the Supplement.

Briefly, we represent below figures showing how different samples get synchronized with the zeitgeber. To do this, we first represent the zeitgeber signal as a continuous uniformly increasing phase (“zeitgeber time”) with period : . The initial condition for is chosen so that the zeitgeber phase at the moment of last pulse is matching the experimental entrainment phase for each . We plot for each sample (dotted lines) and the zeitgeber phase (magenta line). To quantify how well each sample is following the zeitgeber time, we compute the Kuramoto parameter: . By the end of experiment most samples reach , indicating entrainment. Most samples need zeitgeber cycles to become entrained. For min the entrainment takes much longer (edge of the Arnold tongue). For min there is much variability, which can be explained by the horizontal region in the PRC around the entrainment phase. As suggested by the referee, synchronization is faster for higher DAPT concentration. So those dynamics are indeed consistent with the expectation from classical PRC theory.

# 23 Do samples change period to Tzeit in similar ways - i.e. patterns over time. It looks like the

kuramoto order parameter and period drop initially - why?

We do not have a direct answer as to why the Kuramoto first order parameter and the period drop for the condition the reviewer specified. It has to be noted though that because of how wavelet analysis is done (cross-correlation of the timeseries with wavelets), the period and phase determination at the boundaries of the time series are less reliable (edge effects, see Mönke et al., 2020). Because of this, we should take caution when considering data to and from the first and last pulses, respectively. This was explicitly stated in the generation of stroboscopic maps: “As wavelets only partially overlap the signal at the edges of the timeseries, resulting in deviations from true phase values (Mönke et al., 2020), the first and last pulse pairs were not considered in the generation of stroboscopic maps.”

# 24 In Figure 4C why is the Kuramoto order parameter already higher in the 2uM DAPT conditions at

the start of the experiment?

Samples can, by chance, start synchronously and this results in a high Kuramoto first order parameter. Because of this likelihood, it is thus important to interpret the entrainment behaviour of multiple samples using various readouts, in addition to a high Kuramoto first order parameter. We investigated entrainment of the samples based on several measures: multiple samples remaining (or becoming more) synchronous (because each sample actively synchronizes with the zeitgeber), period-locking (where the pace of the samples match the pace of the zeitgeber, which can be distinct from natural pace), and phase-locking (where there is an establishment of a stable phase relationship between the samples and the zeitgeber).

# 25 Figure 3C and Figure S2 require statistical testing between CTRL and DAPT in each condition

p-values were calculated for the specified conditions and were added in the caption of the figures. These values are enumerated here:

• Figure 3C
• 170-min 2uM DAPT (vs DMSO control): p
• Figure S2
• 120-min 2uM DAPT (vs DMSO control): p = 0.064
• 130-min 2uM DAPT (vs DMSO control): p = 0.003
• 140-min 2uM DAPT (vs DMSO control): p = 0.272
• 150-min 2uM DAPT (vs DMSO control): p = 0.001
• 160-min 2uM DAPT (vs DMSO control): p To calculate p-values, two-tailed test for absolute difference between medians was done via a randomization method (Goedhart, 2019). This confirms that the period of samples subjected to pulses of DAPT is not equal to the controls, except for the 140-min condition (where the zeitgeber period is equal to the natural period, i.e. 140 mins).

# 26 Figure 3A gray shaded area not clearly visible on the graph

We have decided to remove the interquartile range (IQR) in the specified figure as it does not serve a crucial purpose in this case. By removing it in Figure 3A, the timeseries of individual samples are now clearer.

# 27 Figure 6C colour maping of time progression is not clearly visible on the graph; the interpretation

of this observation is unclear in the text and the figure

We agree that the low quality of the image is unfortunate, and it seems that our file was greatly compressed upon submission. We have checked the proper quality of figures in the resubmitted version of the manuscript.

Regarding the interpretation of Figure 6C, we conclude that in our experiments the entrainment phase is an attractor or stable fixed point, in line with theory (Granada and Herzel, 2009; Granada et al., 2009),. We had elaborated this in the text (lines 248-252 of the submitted version of the manuscript): at the same zeitgeber strength and zeitgeber period, faster (or slower) convergence towards this fixed point (i.e. entrainment) was achieved when the initial phase of the endogenous oscillation (φinit) was closer or farther to φent.

# 28 Figure 7A circular spread not clearly visible on the graph

Similar to point #27, we have provided a high resolution graph for the re-submission and hopefully resolved this issue.

# 29 Figure S7A difficult to see the difference between colours

See point #28.

# 30 Is it possible to compare the PRC and the plots of period over time during entrainment? The PRC

is mainly negative (Fig 8A1,A2), in my understanding this means a delay, however the periods seem to decrease over time before entraining to the Tzeit (Fig 3B). Is this reflective of a decrease in Kuramoto parameter and potential de-synchronisation of single cells before re-synchronisation at Tzeit?

To address this question, we now plot the Phase response with colors indicating pulse number in new Supplementary Figure S13. While capturing the entire PRC as a function of time would require many more experiments (in particular to sample the phases far from entrainment phase), we still clearly see that the PRCs appear to translate vertically as the oscillator is being entrained, i.e. the latter time points are shifted up (down) for T_zeit = 120 (170) min, respectively.

# 31 Fig 8A What is the importance/meaning of the PRC being similar shape between different

entrainment periods? Does this reflect that the underlying gene network is the same?

If one single gene network is responsible for oscillations, we expect from dynamical systems theory that the PRC are not only of similar shape but actually the same, independent of the entrainment period. What is surprising is that the PRC for different entrainment periods do not overlap, and the simplest explanation for this is that the intrinsic period changes with entrainment, all things being kept equal (including the underlying gene networks). This relates to the previous point since we indeed observe that the PRC “translates” vertically with the pulse number for longer periods. The change of period might be due to a long-term regulation as detailed in the discussion.

# 32 The spatial period gradient and wave propagation under DAPT (Figure S8) should be included in

the results and not just the discussion.

We fully agree with the reviewer that both the establishment and the maintenance of a spatial phase gradient is of great interest. However, many more experiments would be required to fully quantify and understand the processes at play here, which we believe to be out of the scope of the current manuscript. To keep the focus of the paper on the global segmentation clock itself, we prefer to keep this figure in Supplement.

Reviewer #2 (Significance (Required)):

We currently do not have a detailed understanding of how biological oscillators integrate local signals from their neighbours as well global external signals to give rise to complex patterning that is important for embryonic development. Main bottlenecks that hinder our understanding are lack of real-time endogenous dynamic response together with known global inputs as well as comprehensive models that can explain emergent behaviour in a variety of tissues.

This study goes a long way in addressing these bottlenecks in the embryonic tissue responsible for somite formation, a dynamical and oscillatory system also known as the segmentation clock. Firstly, they rely on a state-of-the-art previously developed system to entrain endogenous response in live tissue explants using precise microfluidic control. They test the complete range of exogenous perturbation periods and use an existing live reporter (LuVeLu) to monitor endogenous response. They also identify higher order coupling relationships whereby every other LuVeLu peak is entrained through external stimulation.

As the stimulation system does not control but rather perturb the endogenous response, the observations from LuVeLu provide a unique opportunity in understanding input-output relationships and thus describing the dynamic response of the segmentation clock. Authors propose to study dynamic behaviour of the clock using coarse-graining and focus on describing the overall response over time while amalgamating spatial information. Appropriate coarse-graining is an important strategy in addressing complex problems and is widely used. They use sophisticated methodology such as phase response curves and Arnold tongue mapping to make several important observations. For example the nonlinear shortening and elongation of the period in response to stimulation is particularly interesting since this may indicates a feedback of the clock onto itself potentially via Wnt. Another key observation is that the spatial periodicity and phase wave activity persists in the perturbed conditions suggesting that individual single cell oscillators can adjust their behaviour to external input while retaining coordination with their neighbours. Finally, the authors go on to construct a general dynamical model of the segmentation clock and use this to conclude that the intrinsic period of the oscillator is altered and that the oscillator can be considered excitable.

This work sheds light onto mechanisms of coordination of Notch activity in assemblies of cells observed in living tissue, an area of research that is important not only for somitogenesis but also for understanding gene expression patterning in many other tissues where Notch plays a critical role, for example in the development of the neural system and organs. As a study of a real-world nonlinear oscillator this work is directly of interest to theoreticians and synthetic biology experts interested in understanding complex patterning and emergence.

---

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

In this manuscript, authors studied the system-level responses of the somite segmentation clock by the coarse-grained theoretical-experimental approach, applying the theory of entrainment to understanding the phase responses of mouse pre-somitic mesoderm (PSM) tissues in the presence of periodic perturbation of Notch inhibitor DAPT generated by micro-fluidics technique. It was demonstrated that the segmentation clock is responsive to diverse range of the perturbation-periods from 120 to 180 min, can be period- and phase-locked, and the efficiency is dependent of the DAPT concentration (input-strength). The authors also observed two cycles of the segmentation-clock ticking in single cycles of 300 or 350 min period-perturbation, suggesting that higher order (2:1 mode) entrainment. They also applied stroboscopic maps to analysis and found that entrainment-phases are dependent of period of DAPT pulses, which is recapitulating theoretical predictions. The estimation of the phase response curve (PRC) of the segmentation clock revealed that the inferred PRC is an asymmetrical and mainly negative function, which represents characteristic features in oscillators that emerge after saddle-node on invariant cycle (SNIC) bifurcation. These results also indicated that the the segmentation clock changed the intrinsic period during entrainment.

Major comments:

# 33 I have major concerns about the relevance of the global time-series analysis proposed in Fig.2

and conclusion about the changes of the intrinsic period during entrainment. The validity of the global time-series analysis should be carefully analyzed, because it could bring artifacts in estimated values of the intrinsic period. The authors concluded (page 3, line 172) that the period calculated by the global analysis represents similar values with the rate of segment formation, but there is no data about the quantification of the periods of segmentation, such as the frequency of Mesp2 reporter expression.

We thank the reviewer for this feedback. We have now added the quantification of the period of segment formation (new Figure 2E) and show its strong correspondence to the dynamics of reporters used (Lfng, Axin2, and Mesp2). Please see also our response to point #15 with additional comments regarding the validation of the global time-series analysis.

# 34 Another related issue is the presence of spatial period gradient as mentioned (page 13, line 524).

One possible approach to circumvent this issue would be "local" time-series analysis; for instance, just focusing on the "putative posterior" regions that are close to source-positions of waves. Authors can re-compute and estimate PRCs by using such a method.

We thank the reviewer for this suggestion and have accordingly now included the analysis of a localized ROI at the center (center ROI) of the 2D-assays (new Figures S5-S6). We also computed the PRC from center ROIs as shown below. We note strong correspondence between the global ROI and the center ROI.

# 35 I have another major concern about the evidence of higher order entrainment shown in Fig.5. If

the 1:2 entrainment is successful, we can expect that the values of observed period is close to the half of the period of pulses; However, the period shown in Fig.5B looks like 185 min longer than the half of 350 min. Is this gap due to the temporal accuracy of time-lapse movies?

We do not think the discrepancy comes from a problem of temporal accuracy as the temporal accuracy is the same for all movies and there is no reason why there would be a specific issue for this set of experiments. In addition, we have re-analyzed the data to calculate the period from the stroboscopic maps. Mathematically speaking, we take the stroboscopic map as (see PDF) and use this to estimate the period of oscillation in entrained samples , in particular inverting the formula for 1:2 entrainment we have : see PDF.

The advantage of this method is that it gives a more ``instantaneous” estimation of the period.

The results are as follows:

350 10uM: 187 +- 8 min (average across entrained samples from the last zeitgeber period)

350 5uM: 193 +- 13 min (average across entrained samples from the last zeitgeber period)

300 2uM: 148 +- 8 min (averaged across entrained samples and from two last periods)

This additional analysis is in agreement with the wavelet analysis.

The reviewer is right that for 350 minutes, entrained samples show an observed period that is higher than expected, also based on this new additional analysis. The reason for this is not known. One explanation is the relatively short observation time, especially considering for pulses separated by as much as 350-minutes, i.e. only 3 pulses are applied. [We notice that for 300 minutes pulses, the period converges to 150 mins between the 3rd and the 4th pulse]. We have adjusted the text in the results section to reflect that for 350min entrained samples, the observed period ‘approaches’ the predicted value, while for 300min entrained samples, the observed period is very close to it, i.e. 147mins In addition, we comment that the phase distribution narrows with time, another indication supporting higher order entrainment.

# 36 Also, authors showed the period evolution towards 1:2 locking with just one condition (350 min).

Authors can show the data for multiple conditions as in Fig. 3D, at least for 300 min and 325 min pulses and add the data about final entrained period with statistic analysis that supports the difference between the entrained period and the natural period (140 min).

We thank the reviewer for this feedback and have modified the figure accordingly. In particular, in Figure 5A, we have added the period evolution plot for samples subjected to 300-min periodic pulses of 2uM DAPT (or DMSO for control). Additionally, we have added Figure 5D, which plots the average period in the 300-min and 350-min conditions. We summarize the median average period here with computed p-values:

• 300-min pulses of 2uM DAPT (or DMSO for control): p-value = 0.191
• CTRL: 130.39 mins

• DAPT: 146.45 mins

• 350-min pulses of 5uM DAPT (or DMSO for control): p-value = 0.049

• CTRL: 127 mins

• DAPT: 174.86 mins

• 350-min pulses of 10uM DAPT (or DMSO for control): p-value = 0.016

• CTRL: 142.82 mins
• DAPT: 185.12 mins

Minor comments:

# 37 The authors can draw vertical lines indicating the T_zeit in Fig.3B, Fig.4B and Fig.5B in order to

help comparisons between T_zeit and patterns of period (solid lines).

We thank the reviewer for this comment. We have accordingly added a horizontal line indicating Tzeit in Figures 3B, 4B, S4A, and S5A (figure panel numbers based on the submitted version of the manuscript). We similarly added a horizontal line indicating 0.5Tzeit in the period evolution plots of 300-min and 350-min conditions in Figures 5A and 5B, respectively.

# 38 In Fig.5A, the authors can show period evolution in the case of 300 min DAPT-pulses as shown

in Fig.5B.

We thank the reviewer for this feedback (related to point #36), and we have modified the figure accordingly.

# 39 In Fig.6B DAPT panel, the authors can draw the points of phi_ent as shown in Fig.7A.

We thank the reviewer for this comment, and we have modified the figure accordingly.

# 40 In Fig. 8F, authors can put the information about DAPT concentration at the right y-axis.

This is a similar comment as point #17, see above. In brief, we do not know the precise relation between the strength of the perturbation in our model and DAPT concentration, zeitgeber strength was inferred from the model by matching the experimental entrainment phase with that obtained from the model isophases.

# 41 In Fig. 8G, the PRC in the panel "170 mins" does not have any fixed point (cross sections with

horizontal lines of "0" phase response). If entrainment is successful, there should be stable and unstable fixed points, but those are absent, although 170 min pulses succeeded in the entrainment as shown in Fig.3D. Authors can explain where the fixed points are.

The fixed points are indeed defined by the intersection with a horizontal line, but not with the ‘0’ line. They are found where the phase response compensates for the detuning/period mismatch, not at ‘0’ phase response. (See PDF for more details).

Note however on Fig 8G that we further observe a vertical shift of the PRC, which prompted us to propose a change of the intrinsic period with (as explained in the text when we introduce Figs 8A1-2).

Another way to visualize fixed points is offered in Fig 16 D-E, where we plot the inferred corrected PTC and the stroboscopic maps: there, fixed points correspond to intersections with the diagonal.

Reviewer #3 (Significance (Required)):

Although the phase-analysis has been widely applied to various biological systems, such as circadian clocks, cardiac tissues and neurons, this paper represents the first detailed experimental analysis of the segmentation clock based on the theory of phase dynamics. The major results are inline with theoretical predictions, whereas the suggestion about the SNIC bifurcation is attractive not only to the theoretical researchers but also to the experimental biologists; it has been believed that the segmentation clock consists of negative-feedback oscillator that emerge by Hopf bifurcation, whereas this paper proposes another possibility of the molecular network structure for the clockwork. This issue is related to recently proposed hypothesis about the excitable system in the segmentation clock based on the Yap signaling (Hubaud et al. Cell 171, 668 (2017)). However, unfortunately, discussion about detailed molecular networks are not abundant.

# 42 Thus, maybe the main readers are computational biologists and systems biologists.

We thank the reviewer for his/her significance comment. We have added comments on the bifurcation structure of the segmentation clock and on excitable systems in the discussion. While our focus is on coarse-graining so that we do not and cannot infer precise molecular details, we can still infer some properties of the underlying networks. In particular we now cite several papers explaining how systems with tunable periods/excitable are indicative of the interplay between positive and negative feedbacks. We think those considerations are of interest to a broad range of biologists interested in connecting experiments to theory.

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, authors studied the system-level responses of the somite segmentation clock by the coarse-grained theoretical-experimental approach, applying the theory of entrainment to understanding the phase responses of mouse pre-somitic mesoderm (PSM) tissues in the presence of periodic perturbation of Notch inhibitor DAPT generated by micro-fluidics technique. It was demonstrated that the segmentation clock is responsive to diverse range of the perturbation-periods from 120 to 180 min, can be period- and phase-locked, and the efficiency is dependent of the DAPT concentration (input-strength). The authors also observed two cycles of the segmentation-clock ticking in single cycles of 300 or 350 min period-perturbation, suggesting that higher order (2:1 mode) entrainment. They also applied stroboscopic maps to analysis and found that entrainment-phases are dependent of period of DAPT pulses, which is recapitulating theoretical predictions. The estimation of the phase response curve (PRC) of the segmentation clock revealed that the inferred PRC is an asymmetrical and mainly negative function, which represents characteristic features in oscillators that emerge after saddle-node on invariant cycle (SNIC) bifurcation. These results also indicated that the the segmentation clock changed the intrinsic period during entrainment.

Major comments:

1. I have major concerns about the relevance of the global time-series analysis proposed in Fig.2 and conclusion about the changes of the intrinsic period during entrainment. The validity of the global time-series analysis should be carefully analyzed, because it could bring artifacts in estimated values of the intrinsic period. The authors concluded (page 3, line 172) that the period calculated by the global analysis represents similar values with the rate of segment formation, but there is no data about the quantification of the periods of segmentation, such as the frequency of Mesp2 reporter expression. Another related issue is the presence of spatial period gradient as mentioned (page 13, line 524). One possible approach to circumvent this issue would be "local" time-series analysis; for instance, just focusing on the "putative posterior" regions that are close to source-positions of waves. Authors can re-compute and estimate PRCs by using such a method.
2. I have another major concern about the evidence of higher order entrainment shown in Fig.5. If the 1:2 entrainment is successful, we can expect that the values of observed period is close to the half of the period of pulses; However, the period shown in Fig.5B looks like 185 min longer than the half of 350 min. Is this gap due to the temporal accuracy of time-lapse movies? Also, authors showed the period evolution towards 1:2 locking with just one condition (350 min). Authors can show the data for multiple conditions as in Fig. 3D, at least for 300 min and 325 min pulses and add the data about final entrained period with statistic analysis that supports the difference between the entrained period and the natural period (140 min).

Minor comments:

1. The authors can draw vertical lines indicating the T_zeit in Fig.3B, Fig.4B and Fig.5B in order to help comparisons between T_zeit and patterns of period (solid lines).
2. In Fig.5A, the authors can show period evolution in the case of 300 min DAPT-pulses as shown in Fig.5B.
3. In Fig.6B DAPT panel, the authors can draw the points of phi_ent as shown in Fig.7A.
4. In Fig. 8F, authors can put the information about DAPT concentration at the right y-axis.
5. In Fig. 8G, the PRC in the panel "170 mins" does not have any fixed point (cross sections with horizontal lines of "0" phase response). If entrainment is successful, there should be stable and unstable fixed points, but those are absent, although 170 min pulses succeeded in the entrainment as shown in Fig.3D. Authors can explain where the fixed points are.

Significance

Although the phase-analysis has been widely applied to various biological systems, such as circadian clocks, cardiac tissues and neurons, this paper represents the first detailed experimental analysis of the segmentation clock based on the theory of phase dynamics. The major results are inline with theoretical predictions, whereas the suggestion about the SNIC bifurcation is attractive not only to the theoretical researchers but also to the experimental biologists; it has been believed that the segmentation clock consists of negative-feedback oscillator that emerge by Hopf bifurcation, whereas this paper proposes another possibility of the molecular network structure for the clockwork. This issue is related to recently proposed hypothesis about the excitable system in the segmentation clock based on the Yap signaling (Hubaud et al. Cell 171, 668 (2017)). However, unfortunately, discussion about detailed molecular networks are not abundant. Thus, maybe the main readers are computational biologists and systems biologists.

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

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

Summary:

This is a beautifully elegant study that tests how previously published theoretical predictions about entraining nonlinear oscillators applies to a biological oscillator, the segmentation clock. The authors use a combination of state of the art experimental techniques, signal processing and analytical theory to reach a series of interesting and novel conclusions.

They show that the segmentation clock period can be entrained through Notch inhibitor (DAPT) pulses acting as an external clock (referred to as zeitgeber) using a previously developed and sophisticated microfluidic perfusion system. Pulsing DAPT every 120 to 180min can change the internal clock period while entrainment beyond this range leads to higher order coupling to the zeitgeber period, i.e. entrainment of every other pulse. They then perform entrainment experiments where the concentration of DAPT is changed to elicit a change in the strength of interaction between the internal clock and the external stimulus (referred to as zeitgeber strength); interestingly at low strength response to entrainment is more variable leading to entrainment occurring in some samples while others remain unaffected (Figure 4A); overall, higher concentration leads to faster entrainment (Figure 4C). The experimental data is then analysed using stroboscopic maps to reveal that a stable entrainment phase shift is achieved between the internal clock and the external zeitgeber. Phase response curve (PRC) analysis indicates that the system response is not sinusoidal but predominantly characterised by negative PRC, a behaviour consistent with saddle-node on invariant cycle (SNIC); it also reveals that the intrinsic period changes in a non-linear way and that this effect is reversible when external stimulation stops. Finally, a theoretical model is proposed to represent the segmentation clock as a dynamical system; this is based upon Radial Isochron Cycle with Acceleration (ERICA), an extension motivated by the PRC analysis results which are incompatible with a Radial Isochron Cycle (RIC); this model has predictive capability and could be used to design new control strategies for entrainment of the segmentation clock.

This study makes a series of key conclusions which are of particular importance in understanding the dynamic response of a biological oscillators. Firstly, given it's the characteristics of the dynamic response to entrainment, the segmentation clock is likely close to a SNIC bifurcation and this can explain the tendency for relaxation of the period over time. Secondly, the clock period was changed in a non-linear way in the direction of the zeitgeber period, a finding which is interpreted to indicate the presence of feedback of the segmentation clock onto itself, potentially via Wnt. This makes an excellent prediction that if tested experimentally would greatly improve the impact of the study. It is also noted that the entrainment of the segmentation clock does not abolish spatial periodicity and phase wave emergence suggesting that single cell oscillators can adjust to periodic perturbation while maintaining emergent properties. This is also a significant result that would need to be followed up with experiments and computation however would be best suited to a separate study.

Major comments:

The coarse graining is a major point that would need to be clarified since the rest of the analysis and theoretical modelling in the paper flow from this. Firstly, the interpretation of the schematic in Figure 1A on experimental data collection is not immediately obvious to the reader, lacks a clear flow between the different panels or steps (which could be numbered for example) and does not have a legend to indicate the different colour mapping. Secondly, Figure 2A which explicitly addresses coarse graining is not clear enough. Is the message here that by excluding the inner parts of the sample with a radial ROI, a similar dynamic response is observed over time? Since the inner part of the sample corresponds to the posterior side how do we interpret similarities and differences between signals with different ROIs? A quantitative analysis of essential coarse-grained properties such as period and amplitude should be performed for different ROIs and across multiple samples. As this effectively masks any spatial differences, limitations of this approach should be clearly stated in the Discussion. For example in lines 466-470 where it is difficult to interpret the slowing down tendency and relate back to single cell level.

The functional characterisation of the sample using LFNG, AXIN2 and MESP2 is unclear. The images included in Figure 2D representing expression observed when tissue explants are grown within the microfluidic chip are difficult to interpret and would require a more detailed description of anterior-posterior, pillars etc; it is also difficult to view the bright-field since it is presented as a merged image. It is particularly difficult to see the somite boundaries for the same reason. In lines 113-117 the authors state that the global oscillation period matches the periodic boundary formation. How do we reach this conclusion from these images? What is the variability between samples?

If these two issues would be addressed it would increase confidence in the coarse graining argument and thus would strengthen the importance of the findings in the study.

Several minor points could be addressed to improve the manuscript and are listed below: -Figure 1 A the colormap and axes for the oscillatory traces should be defined -strength of zeitgeber is not defined and there is no analytical expression provided; how does it relate to DAPT concentration? Is the fact that low DAPT concentration corresponds to weak strength expected or is it a result? - In some figures it looks like the amplitude of oscillations may change with DAPT concentration and hence zeitgeber strength? Is this expected? -Figure 2A including the black area creates confusion and it is unclear which ROI is used in the rest of the study; consider moving this to a supplementary figure perhaps -what type of detrending is used in Figure 2 and throughout (include info in the figure legend) -Figure 2D merged images are difficult to read/interpret (see major comments) -Kuramoto order parameter is used to quantify the level of synchrony across the different samples however it is not defined in the text. Is it also possible to assess variability in each sample? For example how quickly does entrained occur in each sample? How faithfully the peaks of expression beyond 80min (to exclude initial unsynchronised state) match with zeitgeber time? This would help make the point that weak strength leads to a more variable response which is an interesting finding. - Do samples change period to Tzeit in similar ways - i.e. patterns over time. It looks like the kuramoto order parameter and period drop initially - why? -In Figure 4C why is the Kuramoto order parameter already higher in the 2uM DAPT conditions at the start of the experiment? -Figure 3C and Figure S2 require statistical testing between CTRL and DAPT in each condition -Figure 3A gray shaded area not clearly visible on the graph -Figure 6C colour maping of time progression is not clearly visible on the graph; the interpretation of this observation is unclear in the text and the figure -Figure 7A circular spread not clearly visible on the graph -Figure S7A difficult to see the difference between colours -Is it possible to compare the PRC and the plots of period over time during entrainment? The PRC is mainly negative (Fig 8A1,A2), in my understanding this means a delay, however the periods seem to decrease over time before entraining to the Tzeit (Fig 3B). Is this reflective of a decrease in Kuramoto parameter and potential de-synchronisation of single cells before re-synchronisation at Tzeit? -Fig 8A What is the importance/meaning of the PRC being similar shape between different entrainment periods? Does this reflect that the underlying gene network is the same? -The spatial period gradient and wave propagation under DAPT (Figure S8) should be included in the results and not just the discussion.

Significance

We currently do not have a detailed understanding of how biological oscillators integrate local signals from their neighbours as well global external signals to give rise to complex patterning that is important for embryonic development. Main bottlenecks that hinder our understanding are lack of real-time endogenous dynamic response together with known global inputs as well as comprehensive models that can explain emergent behaviour in a variety of tissues.

This study goes a long way in addressing these bottlenecks in the embryonic tissue responsible for somite formation, a dynamical and oscillatory system also known as the segmentation clock. Firstly, they rely on a state-of-the-art previously developed system to entrain endogenous response in live tissue explants using precise microfluidic control. They test the complete range of exogenous perturbation periods and use an existing live reporter (LuVeLu) to monitor endogenous response. They also identify higher order coupling relationships whereby every other LuVeLu peak is entrained through external stimulation.

As the stimulation system does not control but rather perturb the endogenous response, the observations from LuVeLu provide a unique opportunity in understanding input-output relationships and thus describing the dynamic response of the segmentation clock. Authors propose to study dynamic behaviour of the clock using coarse-graining and focus on describing the overall response over time while amalgamating spatial information. Appropriate coarse-graining is an important strategy in addressing complex problems and is widely used. They use sophisticated methodology such as phase response curves and Arnold tongue mapping to make several important observations. For example the nonlinear shortening and elongation of the period in response to stimulation is particularly interesting since this may indicates a feedback of the clock onto itself potentially via Wnt. Another key observation is that the spatial periodicity and phase wave activity persists in the perturbed conditions suggesting that individual single cell oscillators can adjust their behaviour to external input while retaining coordination with their neighbours. Finally, the authors go on to construct a general dynamical model of the segmentation clock and use this to conclude that the intrinsic period of the oscillator is altered and that the oscillator can be considered excitable. This work sheds light onto mechanisms of coordination of Notch activity in assemblies of cells observed in living tissue, an area of research that is important not only for somitogenesis but also for understanding gene expression patterning in many other tissues where Notch plays a critical role, for example in the development of the neural system and organs. As a study of a real-world nonlinear oscillator this work is directly of interest to theoreticians and synthetic biology experts interested in understanding complex patterning and emergence.

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

Major comments:

• Are the key conclusions convincing?

We discuss 4 key conclusions.

A PRC of the segmentation clock was constructed. Although the authors have produced an interesting phase map, the regulation function F(\phi) of the circle map does not give the phase response curve (PRC) (Hoppensteadt & Keener 1982, Guevara & Glass 1982). This holds only when the system is stimulated with very short pulses (ideally Dirac delta), but the experimental pulses here are a quarter of the intrinsic period. Furthermore, in eq. 1 T_ext must be the winding number, and the modulus must be in units of phase, either one or two pi, for the circle map to be correct. Thus, calling the measured response of the system a PRC is not convincing.

The system is being entrained.

Technically, entrainment requires the existence of quasi-periodic response (outside Arnold tongues). It would also be easier to get the stroboscopic maps in the quasi-periodic regime since all the points in the circle will be sampled. Since no quasi-periodic response was demonstrated, the claim of entrainment is not convincing.

The response of the system to external pulses is compatible with a SNIC.

This is compatible, but it is equally compatible with other explanations. Assuming that the PRC is the same as the regulation function F(\phi), the PRC in Kotani 2012 (PRL 2012 fig. 3C) would be a similar shape as that shown by the authors. Similar models to that in Kotani et al., have been studied, but a SNIC has not been found (an der Heiden & Mackey 1982). It is relatively straightforward to construct a phenomenological model with a SNIC, but having underlying biological insight is not guaranteed. No argument for choosing a SNIC is given, so this emphasis of the paper is not convincing.

The work demonstrates coarse graining of complex systems.

This conclusion is correct, but coarse graining theory-driven analysis and control of dynamical systems has been established for many years. What is new here is that it is applied specifically to the in vitro culture system of the mouse segmentation clock.

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

Following on the points above, each of these needs to be corrected or re-done, and/or the conclusions need to be modified accordingly.

• 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 the authors wish to make the strong claim of determining a true PRC, Dirac delta-like perturbation needs to be applied, or approximated by short time duration pulses compared to the intrinsic period.

• 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.

It's not clear to this reviewer if it is feasible to deliver a very short pulse and record a response. But this may not be relevant, see above.

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

Yes.

• Are the experiments adequately replicated and statistical analysis adequate?

Yes.

Minor comments:

• Specific experimental issues that are easily addressable.

No issues.

• Are prior studies referenced appropriately?

Yes.

• Are the text and figures clear and accurate?

Figure 1D illustrates how a PRC should be obtained, but doesn't show the experimental protocol applied in the paper.

In Figure 5B, 10 uM DAPT, the traces are already synchronized before the pulse train starts, which makes the subsequent behavior difficult to interpret.

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

The text includes several paragraphs reviewing broad principles of coarse graining and making general conclusions. This is confusing, because, as mentioned above, there is no new general advance in this paper. The interesting contributions here are specific to the applications to the segmentation clock, and the text should be focused on this aspect.

Significance

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

Description of the complex mouse segmentation clock in terms of a simple model and its PRC is an interesting, original and non-trivial result. The proposal that the segmentation clock is close to a SNIC bifurcation provides a consistent dynamical explanation of slowing behavior that has been recognized for some time, but not fully understood. This proposal also raises a hypothesis about the behavior of the underlying molecular regulatory networks, which may be tested in the future. The increase or decrease of the intrinsic period due to the zeitgeber period is not expected from theory, pointing to structures in internal biochemical feedback loops, an idea which again may be tested in the future. Also surprising from a theoretical perspective, the spatial gradient of period in the system persisted after entrainment. Although the categorization of the generic behavior is interesting, by its nature there is little from this that might give a typical developmental biologist any conclusions about pathways or molecules. The successes and limits of the theoretical description do nevertheless focus future attention on interesting behaviors.

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

Such an analysis of the segmentation clock is based strongly on the experimental system and results in Sonnen et al., 2018, and goes well beyond it in terms of the dynamical analysis. It provisionally categorizes the mouse segmentation clock as a Class I excitable system, allowing its dynamics at a coarse grained level to be compared to other oscillatory systems. In this aspect of simplification, it is similar to approach of Riedel-Kruse et al., 2007 who used a mean-field model of oscillator coupling to explain the synchrony dynamics observed in the zebrafish segmentation clock in response to blockade of coupling pathways, thereby allowing a high-level comparison to other synchronizing systems.

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

Developmental biologists, biophysicists,

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.

Developmental biology, somitogenesis, dynamical systems theory, biophysics, cell signaling

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 excellent comments. The comments raised by the reviewers have tremendously improved our manuscript and allowed us to provide more clarity of our findings. Please find below the point-by-point answers to the reviewer's comments.

---

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

The manuscript from Savva et al. focuses on a long-standing and unresolved challenge of metabolic (and not only) health in mammals: the sexual dimorphism. Authors couple transcriptomics and metabolomics to in-depth molecular phenotyping in offspring of dams fed HFD before conception and throughout pregnancy and lactation to isolate molecular determinants of sexually dimorphic response to maternal obesity.

**Major comments:**

1. While the manuscript present a compelling and exhaustive amount of data, which accurately describes the sexually dimorphic responses to maternal obesity, it lacks mechanistic insights and I personally think this to mainly be a timing issue. For example, I tried hard to find experimental details on the RNA sequencing and could not find much: when is the RNA sequencing performed? If, as I suspect, the sequencing experiments match the metabolomics experiments, I don't think they add much mechanistic insights onto the observed phenomena. They rather contribute to better describe them. Indeed, both metabolomic and transcriptomic profiles might be consequence of the observed phenotypes, rather than be causative (as the authors try to argue). Are these differences already present at birth? what happens to placenta and fetal tissues?

ANSWER: We have clarified the experimental setting and added information about the maternal status. We did not collect the placenta and fetal tissues as the main goal of the study was to investigate the offspring metabolism in a longitudinal study, using the animal as its own control. Therefore, we looked at glucose tolerance, insulin sensitivity and lipid profile in the liver at two different timepoints on the same mouse. At end we sacrificed the mouse and collected tissues for further analysis (lipidomic, RNA sequencing and histology).

Some adult phenotypes - especially metabolic and neurological phenotypes - might also be influenced by different maternal care early postnatal. Are the litters balanced by number and sex ratio? Would cross-fostering maintain the phenotypes?

ANSWER: We agree with the reviewer that phenotype can be influenced by maternal care early postnatal. Each dam delivered litters in different proportions. Dams on CD delivered in average seven to eight littermates (same ratio of female and male offspring) whereas dams on HFD delivered less (five in average). Some of the offspring died at very early age of unknown reasons. The final number of offspring used for this study was 11 females and 12 males born from dams on CD and 11 females and 10 males born from dams on HFD. We did not perform cross-fostering to limit the stress in the animals.

Of the two points above, I would love to see more details on the RNA sequencing, as well as placental and fetal tissues analysed. It would be also interesting to know about any litter balancing measure or at least have more statistics on the litter size and sex ratios.

ANSWER: Each individual was followed over the course of the experimentation for body weight and food intake (six months). However, not all animals were used for every in vivo and ex vivo experiment justified under consideration of the 3Rs, sample throughput capacity and financial constraints (magnetic resonance, lipidomic, qPCR, and tolerance tests). However, all experimentation was performed according to a prior power calculation and published reports (PMID: 30808418, 23446231, 31811898, 25694038 and 31820027). Throughout the study, we opted for randomized experimental design (random selection of the animals for in vivo and ex vivo experiments).

After weaning, up to five littermates were housed per cage. However, some male individuals had to be separated due to aggressive behaviors. When females showed hierarchical behaviors in the cage, we separated them to be sure that each individual had full access to food. Since mice are social animals, no individual was housed alone.

Reviewer #2 (Significance (Required)):

The manuscript from Savva et al. revolves around the unresolved challenge of how sexually dimorphic phenotypes are established. The topic is actual - although already a lot has been published, as acknowledged by the authors as well - and of broad interest to the community of mouse geneticists and physiologists. To understand the molecular underpinnings of sexually dimorphic phenotypes, the authors use in-depth molecular phenotyping in the mouse coupled to metabolomics and transcriptomics. While extremely informative and exhaustive, the actual dataset is - at least for me - purely descriptive, which might reduce its overall impact. I'm a mouse geneticist and a metabolic physiologist and I find the topic of sexual dimorphism extremely interesting.

**Referee Cross-commenting**

I generally agree with the other reviewers' comments. I think the ms is interesting and the dataset compelling although to a certain extent overlapping with previously published studies. There is general agreement on the lack of mechanistic data and the authors should definitely address this point.

ANSWER: In mammals, including humans, biological sex is determined by a pair of sex chromosomes. Genes on the X chromosome (X-linked genes) have distinctive inheritance patterns because they are present in different number between females (XX) and males (XY). Moreover, a plethora of attributes can be determined by sex chromosomes and more importantly by X-linked chromosomes. Therefore, we extracted the sex-linked chromosomes that are affected by the maternal diet or by sex and presented them in figure 5 to give more insight into the molecular underpinning of the sexual dimorphism in metabolism homeostasis.

We have presented the data in figures 4e-4g and commented on page 12 L300-L315 in the result section and in the last paragraph of the discussion section (page 18).

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

In this paper Savva et al. explore how maternal obesity influence hepatic metabolism in a sex-specific fashion. They first assess the contribution of the adipose tissue to the development of insulin resistance and glucose intolerance focusing on inflammation and oxidative phosphorylation pathways. Then they proceed to asses if maternal obesity could remodel the hepatic triglyceride levels and phospholipids using proton magnetic resonance spectroscopy and LC-MS lipidomic, respectively. Finally, they explore hepatic lipid metabolism and genes promoting cancer development.

Despite the methodology is correct and elegant, the study does not explore a possible mechanism of action and some results are contradictory. Indeed, some of the results seems to be driven by the sex of the offspring independently of the maternal feeding.

There are indeed, some limitations for the authors and editors to consider. To this reviewer the manuscript is difficult to read, particularly the results section in which the data are listed without discussing their relevance or their connection to previous research from other groups. Moreover, the discussion could benefit from an extensive rewrite. Indeed, this section lacks of clarity and references that could help elucidate the novel finding of the authors.

ANSWER: We have rewritten the manuscript and we believe that it has been extensively improved in clarity. We have added references and commented on the differences observed, if any. We also have added one complete paragraph on the potential role of sex-biased modulators, especially the sex chromosomes XX or XY. These data are presented in figures 4e-4g and commented on page 12 L300-L315 in the result section and in the last paragraph of the discussion section (page 18).

**Major comments:**

Line 97: How the authors explain similar weight gain in the F-C/HF vs. F-HF/HF? A large body of literature reports that maternal high fat diet influences offspring weight gain, independently of sex, when compared to maternal standard diet (PMID: 23973955; 29872021; 31076636; 3036829).

ANSWER: The literature is still controversial, possibly due to slightly different experimental settings (e.g. the exact composition of the control and high fat diet, exposure time to the different diets). In the current study we used a match control diet of the high fat diet to minimize potential diet-derived signaling molecule effect. We found several publications in line with our findings (PMID: 30405201, 31690792 and 29972240). In addition, in our study, we assessed the changes in body weight over a long time in the offspring, whereas only few studies show detailed measurements over time, which makes it more difficult to compare across studies.

Line 99: Which is the explanation for the reduced body weight in M-HF/HF from birth until 9 week of age? Can the authors show the timeline for food intake?

ANSWER: Excess gestational weight gain is associated with health risk for both the infant and the mother. A large number of epidemiological studies have demonstrated a direct relationship between birth weight and BMI attained in later life. Lower birth weight seems to be associated with later risk for central obesity, which also confers increased cardiovascular risk. It has been previously demonstrated that offspring born from obese mother have lower body weight at birth than control diet/lean mother’s littermates (PMID: 15116085 and 24936914). Here, offspring after weaning are all put on the HFD until six months of age (before sacrifice).

We do not have timeline of food intake but we measured average food intake twice a week during three weeks at about 4-month of age, we presented the data in Fig.S1a and in the result section page 5 (Line 104). Interestingly, food intake tended to be induced and reduced by MO in female and male, respectively.

Line 101: How the authors explain the increase in final weight of the male compared to the female if no differences in total fat, VAT or SAT was observed between the offspring?

ANSWER: We have presented the graphs in relative amount body fat (% TF), and TF is corrected for the total body weight. Figs.1d-1f. When fat is reported to the body weight males have lower relative fat content than females. However, males are taller (bigger) and have bigger bone and muscle mass than females. Therefore, males weight more than females despite lower relative amount of fat.

Line 116: The authors state: "The ratio between the total SAT and the Abd SAT revealed that MO redistributed SAT outside of the abdominal region in females but not in males" but Fig.1g displays no significant differences between F-C/HF and F-HF/HF. Please explain.

ANSWER: Sex differences are observed (at END, MO tended to increase the ratio in F and decrease in M) to a lower level than F. Lower SAT on VAT ratio in obese males than in females is well recognized, however here we observed that MO tends to redistribute fat differently between sexes, and fat distribution has been strongly correlated to metabolic risks.

Line 128-130: The authors state: "At MID, glucose tolerance was highly diet- and sex-dependent, and males but not females showed impaired glucose tolerance by MO." However, in Fig.1h no significant differences in glucose peak or glucose AUC were observed between M-C/HF and M-HF/F. Please explain. This is correct, however fasting glucose (T0) was higher and T60 and T120 as well. In addition, Ins levels during the OGTT was increased at T0, T30 and T120.

ANSWER: We have clarified the sentence and agree with the reviewer that the males are not affected by MO but are already insulin resistant when born from lean mothers. We added the quantitative insulin-sensitivity check index (QUICKI) in Fig.1n and demonstrate that indeed males are less insulin sensitive than females, but MO does not impact the insulin sensitivity in both sexes.

Line 130: OGTT only provides information on insulin secretion and action but does not directly yield a measure of insulin sensitivity. Please rephrase.

ANSWER: We have measured the insulin level during the OGTT, this information, combined with the glucose disappearance curve gives important information on the ability of pancreatic beta cell to release insulin in response to a glucose load and on the ability of insulin to store the glucose into the cells i.e. insulin sensitivity.

We have added the quantitative insulin-sensitivity check index (QUICKI) in Fig.1n and confirmed that males are less insulin sensitive than females with no effect of MO. Line 125

Authors should rephrase the conclusion of the paragraph since MO does not seems to influence fat distribution or insulin resistance. Looking at figure 1 it seems that the only differences observed are driven by the sex of the offspring independently of maternal feeding.

ANSWER: We have changed the conclusion and focused on the sex differences.

Do the dams were insulin resistant? Indeed, hyperinsulinemia and insulin resistance are key programming factor of offspring metabolic syndrome.

ANSWER: We agree that it would have been good to have the insulin levels of the mothers before mating. Unfortunately, we took only the body weight and the glucose level after 2 h fasting. We saw no differences between CD and HFD fed mothers. We included these results in the Figure 1b. Of note, we have performed long term HFD in young female mice for other purposes and noticed that females are resistant to hyperinsulinemia when fed a HFD in a long term, so we assumed that feeding a HFD for 6 weeks before mating would not affect insulin levels.

Line 162: "There were no significant differences between the sexes. However, it is interesting to note that several pathways were regulated differently between sexes between the C/HF and HF/HF groups." Can the authors rephrase the concept, it is unclear.

ANSWER: We have revised the sentences and gave some examples in P6, Lines 132-139.

Line 170: The authors state: "insulin secretion pathway is reduced in males only". How this results are in line with the data reported in Fig.1i were both M-C/HF and M-HF/HF display increased insulin secretion?

ANSWER: We agree with the reviewer that this is controversial. However, males from obese mothers showed slightly increased insulin levels during the OGTT and slightly reduced QUICKI as compared to males from lean mothers. Moreover, here we measured the total insulin level and not the C-peptide level that is more representative of the “active insulin”. One can be that males have higher insulin level but lower active insulin.

Lines 174-175: All the genes reported in Fig. 1o, except for LPIN1, do not seems to be altered by MO. Please rephrase.

ANSWER: Lpin1 and Pdk1 are reduced by MO in females. We have rephrased P6, Lines 142-147

Line 184: The authors state that the signaling pathways was assessed both at transcriptional and post-transcriptional levels. Where are depicted the data of the post-transcriptional levels?

ANSWER: We have corrected this sentence and rephrased it.

Similarly to glucose metabolism and fat depot results, also in the case of the liver steatosis the increased number of lipid droplets seems to be linked to the sex of the animal rather than the maternal diet. Since the authors also investigated inflammatory pathways it could be of interest to assess CD68 infiltration by immunohistochemistry and Picrosirius red staining for the assessment of fibrosis.

ANSWER: We agree with the reviewer that immune cells infiltration quantification would have been excellent. However, due to the lock down and the moving of our lab we could not performed the IHC.

Liver histology in Fig.5a (M-C/HFD) is completely different from the one depicted in Fig.4a in terms of steatosis. Can the authors please explain this difference and report the magnification used in Fig.4a. Please also report the scale in Fig.5a.

ANSWER: We have corrected this mistake and we apologize for the confusion. The Fig.4a described a M-moHF offspring liver. The pictures have been changed and magnification has been added.

Changes in placental function are thought to be a key link between the maternal and intrauterine milieu and long-term health of the offspring (PMID:24107818). Alterations in placental function and structure in response to obesity and their underlying molecular mechanisms have been explored both in humans and in animal models (PMID:24484739; 22303323; 28291256). Others have shown that maternal hyperinsulinemia is strongly associated with offspring insulin resistance and excess placental lipid deposition and hypoxia (PMID: 28291256). Excessive lipid deposition leads to a lipotoxic placental environment that is associated with increased markers of inflammation and oxidative stress (PMID: 24333048). Can the authors could provide some data?

ANSWER: We agree with the reviewer that maternal and intrauterine milieu are crucial to determine long-term health status of the offspring. However, in the current study we wanted to explore the sex differences in offspring fed a HFD and born from either lean or obese mothers in the long term.

We did not focus the current project on the maternal status, but on the offspring and the sexual dimorphism in metabolic risks later in life.

**Minor:**

Fig. 2m Acox1 is not reduced by MO in female.

ANSWER: We have corrected the sentence.

Supp. TableS1 do not report Pdk1, Lpin1, Nox4 and Prlr.

ANSWER: The reason why these genes do not appear in the TableS1 is because the table S1 shows all the significantly regulated genes extracted from the KEGG pathways. The genes mentioned above Pdk1, Lpin1, Nox4 and Prlr do not appear in the KEEG pathway.

Supp. TableS2 do not report PCSK9 and PNPLA3

ANSWER: Same as above, the Table S2 is based on KEGG pathway analysis and the genes Pcsk9 and Pnpla3 do not appear in the KEGG pathway.

Reviewer #3 (Significance (Required)):

The prevalence of obesity during pregnancy continues to increase at alarming rates. This is concerning as in addition to immediate impacts on maternal wellbeing, obesity during pregnancy has detrimental effects on the long-term health of the offspring. This paper is connected to an extended research field aiming at prevent the detrimental effect of maternal obesity on the offspring.

An important limitation in the ability to design intervention strategies to prevent the detrimental effects of maternal obesity on offspring health is that it is currently unclear which of the many potential variables associated with obesity is the key programming factor mediating the effects on the offspring.

**Reviewer field of expertise:**

Molecular Biology, Type 2 Diabetes, Obesity and NAFLD.

**Referee Cross-commenting**

I agree with the other reviewers' comments, particularly on the lack of mechanistic data.

ANSWER: In mammals, including humans, biological sex is determined by a pair of sex chromosomes. Genes on the X chromosome (X-linked genes) have distinctive inheritance patterns because they are present in different number between females (XX) and males (XY). Moreover, a plethora of attributes can be determined by sex chromosomes and more importantly by X-linked chromosomes. Therefore, we extracted the sex-linked chromosomes that are affected by the maternal diet or by sex and presented them in figure 5 to give more insight into the molecular underpinning of the sexual dimorphism in metabolism homeostasis. We have presented the data page 12 L300-L315in the result section and in the last paragraph of the discussion section (page 18).

Reviewer #4 (Evidence, reproducibility and clarity (Required)):

Maternal obesity is a common condition in western society. There is abundant literature showing the deleterious metabolic consequences of MO in the offspring. In this manuscript, Savva et al. characterized the transcriptomic and lipidomic profiles of the liver in male and female progeny of female mice that were fed a high-fat diet during and before pregnancy. After weaning, mice were also fed a high-fat diet. They found that both transcription and lipid composition were different in males and females, and they show that females are protected to metabolic and liver disease, whereas males develop insulin resistance, liver steatosis, and are prone to develop liver cancer. The first part of the study where the authors characterize the metabolism of the progeny, including weight, fat mass in the distinct depots, glucose and insulin tolerance, is not novel. Several publications have previously reported these findings (Programming effects of maternal and gestational obesity on offspring metabolism and metabolic inflammation. Sci Rep. 2019 Nov 5;9(1):16027). It was also previously reported in the cited publication, increased liver weight, steatosis and TG content, similar to the results of the present study. Some novelty of the manuscript is the in-depth analysis of the lipid content of the liver in the models used, as well as the transcriptional profile. Despite the substantial amount of data that the authors generated to prove differences between the male and female offspring, there is not, however, any cross analysis that could link both omics data. Overall, as discussed below the results do not support some conclusions. In particular this reviewer has the following concerns and suggestions.

1. The metabolic status of the obese mothers has a direct impact on the offspring. It was previously reported that differences in glucose tolerance on the mother has a strong impact on the sex dimorphism in the metabolism of the progeny (please see the review: Sex and gender differences in developmental programming of metabolism. Molecular Metabolism 15 (2018) 8-19. What are the levels of glucose tolerance on the mothers used in the study? The weight of the mothers could also be shown.

ANSWER: We added the body weight and the Dam glucose level after 2 h fasting. We observed an increased BW in obese mother and no differences in glucose level compared to control diet fed mothers. We included these results in the Figure 1b.

The fact that mice were fed a high-fat diet after weaning could lead to confounding effects. Indeed, the lack of a group of mice fed a normal diet after weaning makes difficult to establish which is the relative contribution to the phenotype of the diet of the progeny, compared to the diet of the mother.

ANSWER: We published a paper in 2021 (PMID: 33398027) where we show that offspring born from obese mothers have sex specific hepatic modulation and provide molecular evidence of sex dependent in utero metabolic adaptation in the offspring born from obese mothers. In this study offspring were fed the CD after weaning and we demonstrated some reversal effect of CD feeding in male offspring.

Changes in food intake could also explain some differences.

ANSWER: We measured average food intake twice a week during three weeks at about 4-month of age, we presented the data in Fig.S1a and in the result section page 5 (Line 104). Males weighed significantly more than females regardless of the maternal diet (Fig.1c), with higher food intake (Fig.S1a). Interestingly, food intake tended to be induced and reduced by MO in female and male, respectively.

Lipidomics data show major differences between males e and females. However, the impact of these differences in the distinct metabolic phenotypes is not addressed.

ANSWER: We have reformulated the major lipidomic data to integrate them into the sex dependent metabolic phenotypes we observed.

The estrogen family and its two respective receptors, ERα and ERβ, have been widely suggested to be protective against obesity, diabetes, and cardiovascular disease. Does the transcriptional profile consistent with changes in the activity of estrogen receptor signaling?

ANSWER: We have at the expression level of Era (we did not find Erb) in the liver of offspring and presented the data in Figure S1b. Era was higher expressed in females than males, independently of the maternal diet. MO had no effect on the expression level of Era. Interestingly, androgen receptor (Ar) was higher expressed in the liver of F-moC than M-moC, these differences vanished in moHF-offspring because MO tended to reduce and induce the expression level of Ar in female and male, respectively.

Epigenetic modifications likely underlie the differences between males and females. Histone modifications or DNA methylation analysis could further improve the study.

ANSWER: We agree with the reviewer that there is likely a sex dependent epigenetic modification. How maternal in utero environment can differently affect female and male offspring born from the same mother remain to be elucidated.

In the last figure the authors claim that MO prevents HCC, but no data about HCC is shown, only gene expression analysis.

ANSWER: We agree with the reviewer that showing HCC markers would have strength the conclusion on the possible role of MO and sex in HCC development. However, we did not have the possibility to run western blot or IHC on our livers. Nevertheless, H&E staining showed bigger cell proliferation spots in females than in males and a reduction of the size of these spots in females born from obese mothers as compared to those born from lean mothers, in line with the pathways analysis and gene expression. Further studies focusing on HCC development in obesity would be needed to unravel the mechanism behind.

Reviewer #4 (Significance (Required)):

The first part of the study where the authors characterize the metabolism of the progeny, including weight, fat mass in the distinct depots, glucose and insulin tolerance, is not novel. Several publications have previously reported these findings (Programming effects of maternal and gestational obesity on offspring metabolism and metabolic inflammation. Sci Rep. 2019 Nov 5;9(1):16027). It was also previously reported in the cited publication, increased liver weight, steatosis and TG content, similar to the results of the present study.

**Referee Cross-commmenting**

I agree with the comments. I also think that the MS is difficult to read. No conexion between the OMICS data.

ANSWER: The current version of the manuscript has been extensively revised and we believe that it has improved in clarity and in novelty.

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

Maternal obesity is a common condition in western society. There is abundant literature showing the deleterious metabolic consequences of MO in the offspring. In this manuscript, Savva et al. characterized the transcriptomic and lipidomic profiles of the liver in male and female progeny of female mice that were fed a high-fat diet during and before pregnancy. After weaning, mice were also fed a high-fat diet. They found that both transcription and lipid composition were different in males and females, and they show that females are protected to metabolic and liver disease, whereas males develop insulin resistance, liver steatosis, and are prone to develop liver cancer. The first part of the study where the authors characterize the metabolism of the progeny, including weight, fat mass in the distinct depots, glucose and insulin tolerance, is not novel. Several publications have previously reported these findings (Programming effects of maternal and gestational obesity on offspring metabolism and metabolic inflammation. Sci Rep. 2019 Nov 5;9(1):16027). It was also previously reported in the cited publication, increased liver weight, steatosis and TG content, similar to the results of the present study. Some novelty of the manuscript is the in-depth analysis of the lipid content of the liver in the models used, as well as the transcriptional profile. Despite the substantial amount of data that the authors generated to prove differences between the male and female offspring, there is not, however, any cross analysis that could link both omics data. Overall, as discussed below the results do not support some conclusions. In particular this reviewer has the following concerns and suggestions.

1. The metabolic status of the obese mothers has a direct impact on the offspring. It was previously reported that differences in glucose tolerance on the mother has a strong impact on the sex dimorphism in the metabolism of the progeny (please see the review: Sex and gender differences in developmental programming of metabolism. Molecular Metabolism 15 (2018) 8-19. What are the levels of glucose tolerance on the mothers used in the study? The weight of the mothers could also be shown.
2. The fact that mice were fed a high-fat diet after weaning could lead to confounding effects. Indeed, the lack of a group of mice fed a normal diet after weaning makes difficult to establish which is the relative contribution to the phenotype of the diet of the progeny, compared to the diet of the mother.
3. Changes in food intake could also explain some differences.
4. Lipidomics data show major differences between males e and females. However, the impact of these differences in the distinct metabolic phenotypes is not addressed.
5. The estrogen family and its two respective receptors, ERα and ERβ, have been widely suggested to be protective against obesity, diabetes, and cardiovascular disease. Does the transcriptional profile consistent with changes in the activity of estrogen receptor signaling?
6. Epigenetic modifications likely underlie the differences between males and females. Histone modifications or DNA methylation analysis could further improve the study.
7. In the last figure the authors claim that MO prevents HCC, but no data about HCC is shown, only gene expression analysis.

Significance

The first part of the study where the authors characterize the metabolism of the progeny, including weight, fat mass in the distinct depots, glucose and insulin tolerance, is not novel. Several publications have previously reported these findings (Programming effects of maternal and gestational obesity on offspring metabolism and metabolic inflammation. Sci Rep. 2019 Nov 5;9(1):16027). It was also previously reported in the cited publication, increased liver weight, steatosis and TG content, similar to the results of the present study.

Referee Cross-commmenting

I agree with the comments. I also think that the MS is difficult to read. No conexion between the OMICS data.

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

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

In this paper Savvaet al. explore how maternal obesity influence hepatic metabolism in a sex-specific fashion. They first assess the contribution of the adipose tissue to the development of insulin resistance and glucose intolerance focusing on inflammation and oxidative phosphorylation pathways. Then they proceed to asses if maternal obesity could remodel the hepatic triglyceride levels and phospholipids using proton magnetic resonance spectroscopy and LC-MS lipidomic, respectively. Finally, they explore hepatic lipid metabolism and genes promoting cancer development.

Despite the methodology is correct and elegant, the study does not explore a possible mechanism of action and some results are contradictory. Indeed, some of the results seems to be driven by the sex of the offspring independently of the maternal feeding.

There are indeed, some limitations for the authors and editors to consider. To this reviewer the manuscript is difficult to read, particularly the results section in which the data are listed without discussing their relevance or their connection to previous research from other groups. Moreover, the discussion could benefit from an extensive rewrite. Indeed, this section lacks of clarity and references that could help elucidate the novel finding of the authors.

Major comments:

Line 97: How the authors explain similar weight gain in the F-C/HF vs. F-HF/HF? A large body of literature reports that maternal high fat diet influences offspring weight gain, independently of sex, when compared to maternal standard diet (PMID: 23973955; 29872021; 31076636; 3036829).

Line 99: Which is the explanation for the reduced body weight in M-HF/HF from birth until 9 week of age? Can the authors show the timeline for food intake?

Line 101: How the authors explain the increase in final weight of the male compared to the female if no differences in total fat, VAT or SAT was observed between the offspring?

Line 116: The authors state: "The ratio between the total SAT and the Abd SAT revealed that MO redistributed SAT outside of the abdominal region in females but not in males" but Fig.1g displays no significant differences between F-C/HF and F-HF/HF. Please explain.

Line 128-130: The authors state: "At MID, glucose tolerance was highly diet- and sex-dependent, and males but not females showed impaired glucose tolerance by MO." However, in Fig.1h no significant differences in glucose peak or glucose AUC were observed between M-C/HF and M-HF/F. Please explain.

Line 130: OGTT only provides information on insulin secretion and action but does not directly yield a measure of insulin sensitivity. Please rephrase.

Authors should rephrase the conclusion of the paragraph since MO does not seems to influence fat distribution or insulin resistance. Looking at figure 1 it seems that the only differences observed are driven by the sex of the offspring independently of maternal feeding. Do the dams were insulin resistant? Indeed, hyperinsulinemia and insulin resistance are key programming factor of offspring metabolic syndrome.

Line 162: "There were no significant differences between the sexes. However, it is interesting to note that several pathways were regulated differently between sexes between the C/HF and HF/HF groups." Can the authors rephrase the concept, it is unclear.

Line 170: The authors state: "insulin secretion pathway is reduced in males only". How this results are in line with the data reported in Fig.1i were both M-C/HF and M-HF/HF display increased insulin secretion?

Lines 174-175: All the genes reported in Fig. 1o, except for LPIN1, do not seems to be altered by MO. Please rephrase.

Line 184: The authors state that the signaling pathways was assessed both at transcriptional and post-transcriptional levels. Where are depicted the data of the post-transcriptional levels?

Similarly to glucose metabolism and fat depot results, also in the case of the liver steatosis the increased number of lipid droplets seems to be linked to the sex of the animal rather than the maternal diet. Since the authors also investigated inflammatory pathways it could be of interest to assess CD68 infiltration by immunohistochemistry and Picrosirius red staining for the assessment of fibrosis.

Liver histology in Fig.5a (M-C/HFD) is completely different from the one depicted in Fig.4a in terms of steatosis. Can the authors please explain this difference and report the magnification used in Fig.4a. Please also report the scale in Fig.5a.

Changes in placental function are thought to be a key link between the maternal and intrauterine milieu and long-term health of the offspring (PMID:24107818). Alterations in placental function and structure in response to obesity and their underlying molecular mechanisms have been explored both in humans and in animal models (PMID:24484739; 22303323; 28291256). Others have shown that maternal hyperinsulinemia is strongly associated with offspring insulin resistance and excess placental lipid deposition and hypoxia (PMID: 28291256). Excessive lipid deposition leads to a lipotoxic placental environment that is associated with increased markers of inflammation and oxidative stress (PMID: 24333048). Can the authors could provide some data?

Minor:

Fig. 2m Acox1 is not reduced by MO in female. Supp. TableS1 do not report Pdk1, Lpin1, Nox4 and Prlr. Supp. TableS2 do not report PCSK9 and PNPLA3

Significance

The prevalence of obesity during pregnancy continues to increase at alarming rates. This is concerning as in addition to immediate impacts on maternal wellbeing, obesity during pregnancy has detrimental effects on the long-term health of the offspring. This paper is connected to an extended research field aiming at prevent the detrimental effect of maternal obesity on the offspring.

An important limitation in the ability to design intervention strategies to prevent the detrimental effects of maternal obesity on offspring health is that it is currently unclear which of the many potential variables associated with obesity is the key programming factor mediating the effects on the offspring.

Reviewer field of expertise:

Molecular Biology, Type 2 Diabetes, Obesity and NAFLD.

Referee Cross-commenting

I agree with the other reviewers' comments, particularly on the lack of mechanistic data.

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

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

The manuscript from Savva et al. focuses on a long-standing and unresolved challenge of metabolic (and not only) health in mammals: the sexual dimorphism. Authors couple transcriptomics and metabolomics to in-depth molecular phenotyping in offspring of dams fed HFD before conception and throughout pregnancy and lactation to isolate molecular determinants of sexually dimorphic response to maternal obesity.

Major comments:

1. While the manuscript present a compelling and exhaustive amount of data, which accurately describes the sexually dimorphic responses to maternal obesity, it lacks mechanistic insights and I personally think this to mainly be a timing issue. For example, I tried hard to find experimental details on the RNA sequencing and could not find much: when is the RNA sequencing performed? If, as I suspect, the sequencing experiments match the metabolomics experiments, I don't think they add much mechanistic insights onto the observed phenomena. They rather contribute to better describe them. Indeed, both metabolomic and transcriptomic profiles might be consequence of the observed phenotypes, rather than be causative (as the authors try to argue). Are these differences already present at birth? what happens to placenta and fetal tissues?
2. Some adult phenotypes - especially metabolic and neurological phenotypes - might also be influenced by different maternal care early postnatal. Are the litters balanced by number and sex ratio? Would cross-fostering maintain the phenotypes?

Of the two points above, I would love to see more details on the RNA sequencing, as well as placental and fetal tissues analysed. It would be also interesting to know about any litter balancing measure or at least have more statistics on the litter size and sex ratios.

Significance

The manuscript from Savva et al. revolves around the unresolved challenge of how sexually dimorphic phenotypes are established. The topic is actual - although already a lot has been published, as acknowledged by the authors as well - and of broad interest to the community of mouse geneticists and physiologists. To understand the molecular underpinnings of sexually dimorphic phenotypes, the authors use in-depth molecular phenotyping in the mouse coupled to metabolomics and transcriptomics. While extremely informative and exhaustive, the actual dataset is - at least for me - purely descriptive, which might reduce its overall impact. I'm a mouse geneticist and a metabolic physiologist and I find the topic of sexual dimorphism extremely interesting.

Referee Cross-commenting

I generally agree with the other reviewers' comments. I think the ms is interesting and the dataset compelling although to a certain extent overlapping with previously published studies. There is general agreement on the lack of mechanistic data and the authors should definitely address this point.

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

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

Learn more at Review Commons

---

Reply to the reviewers

Reviewer #1 (Evidence, reproducibility and clarity):

The authors present further investigation of the Sox transcription factors in the model Cnidarian Hydractinia. They showcase the Hydractinia as now a relatively technically advanced model system to study animal stem cells, regeneration and the control of differentiation in animal cells. In this study they characterise the neural cells in hydractinia using FACS and sing cell transcriptome sequencing, investigate the sequential expression of SoxB genes in the i-cells and presumptive lineage giving rise to i-cells and investigate the neuronal regeneration making good use of transgenic rules. Finally, they investigate the role of SoxB genes in embryonic neurogenesis.

There are no major or minor issues effecting the conclusions

Reviewer #1 (Significance):

This study helps to confirm the role of an important group of transcription factors is conserved across the metazoan as well as showcasing an exciting model organism for regeneration and stem cell biology. This will of interest to a broad audience of developmental and biologists.

My own research is in the same field, using a different model system

Referees cross-commenting

I agree with the comments from the other reviewers, and am sure the authors can address these adequately with further explanation.

Reviewer #2 (Evidence, reproducibility and clarity):

Summary

Chrysostomou et al. investigate the role of three putative SoxB genes in embryonic neurogenesis in the colonial hydrozoan Hydractinia. They show that SoxB1 is co-expressed with Piwi in the multipotent i-cells and, using transgenics, they show that these Piwi/SoxB1 cells become neurons and gametes, consistent with the cell types that differentiate from i-cells. They further suggest that SoxB2 and SoxB3 are expressed downstream of SoxB1 in the progeny of the i-cells and, using shRNAs, investigate the role of SoxB genes on embryonic neurogenesis. The primary conclusions center on the similarity between neural differentiation in humans and Hydractinia as both systems pattern neurons using sequential expression of SoxB genes during the differentiation of neurons. The manuscript presents a large and diverse set of data derived from analysis of transgenic animals, single-cell sequencing, and investigation of gene function; despite this, the conclusions are either not particularly novel or not well-supported. The co-expression of SoxB1 in Piwi-expressing i-cells appears to be both novel and significant but the implications are not clearly indicated. Additional specific concerns are detailed below.

Major comments

1. SoxB genes act sequentially<br /> Knockdown of SoxB2 has already been shown to result in the loss of SoxB3, so the sequential action of SoxB genes in this animal does not seem to be a terribly novel conclusion.

Sequential expression of Soxb1-Soxb2 has not been demonstrated previously. Flici et al. did show some data on Soxb1 expression but these were not detailed. Furthermore, they have not shown in vivo transition to Soxb2. Our new single-molecule fluorescence in situ hybridization, and the transgenic reporter animals have been developed to address these issues.

While this manuscript does appear to report the most comprehensive analysis of SoxB1 expression, the evidence for sequential activation of SoxB1 and then SoxB2 in the same lineage (Figure 4) is a bit troubling. Panel A of this figure appears to show complete overlap between SoxB1 and SoxB2, suggesting all the cells in this field are synchronously passing through the transition point from SoxB1 to SoxB2 expression. While this may reflect reality, it would be more convincing to see adjacent cells expressing SoxB1 only or SoxB2 only, reflecting the dynamic progression of cell type specification along the main body axis.

As shown in Figures 1, Soxb1 is expressed by i-cells (together with Piwi1) in the lower body column of feeding polyps and in germ cells in sexual polyps. These cells do not express Soxb2. Figure 2 shows that Soxb2 is expressed more orally in a population of putative i-cell progeny as they migrate towards the head. These cells still express Soxb1. In the upper part of the body column, just under the tentacle line, there are Soxb2+ cells that do not express Soxb1. Therefore, cells expressing Soxb1 but not Soxb2 are present in the basal part of the polyp, Soxb1+/Soxb2+ double positive cells in the mid body region (i.e., the interface between the two domains where Soxb1+ cells start to express Soxb2 and downregulate Soxb1.), and cells expressing Soxb2 but not Soxb1 in the upper part of the polyp, just under the tentacle line. In Figure 4, we show the interface between these two domains using in vivo imaging of double transgenic reporter animals to visualize the Soxb1 to Soxb2 transition. Indeed, in the mid body area, most Soxb1+ cells also express Soxb2 (Figure 2). Hence, Figure 4 should be seen keeping Figure 2’s data in mind. At the mRNA level, the overlap between the Soxb1 and Soxb2 domains is smaller (Figure 2) than the one shown in Figure 4 because the latter constitutes a lineage tracing, showing fluorescent proteins with a long half-life. Therefore, when i-cells downregulate Soxb1 while starting to express Soxb2, the long half-life of tdTomato results in red fluorescence persisting longer than the mRNA encoding it. We have added cartoons to Figure 4 to indicate the position along the main body axis that are depicted.

Panel B is more concerning; while the authors have highlighted a cell that does appear to transition from SoxB1+ to SoxB1+/SoxB2+, there are several cells in the background that appear to gain SoxB2 expression without first expressing SoxB1. Do these cells constitute a fundamentally different, SoxB1-indpenendent, lineage of SoxB2+ cells? This would be noteworthy but is not mentioned or characterized.

The panels included in Figure 4 constitute selected confocal slices of stacks acquired in vivo. During imaging, cells move in three dimensions, making them appear and disappear in given optical planes over time. In other words, the individual time frames shown (T0-T5) were not always found in the same plane due to cell migration in the Z dimension. The cells that appear to gain Soxb2+ w/o having expressed Soxb1 first are an example of such cells. They are probably Soxb2+ cells that had already downregulated Soxb1 and migrated into the respective plane of image. We have added the explanation to Figure 4's legend.

Figure 7 shows the effect of SoxB1 knockdown (by shRNA) on the number of Piwi-expressing cells, nematocytes, etc but why not show that SoxB2 and SoxB3 are also knocked down in these experiments? Figure S11 shows no effect of SoxB2 and SoxB3 knockdown on SoxB1 expression but why wasn't the reciprocal experiment performed? If SoxB2 and SoxB3 are really downstream of SoxB1, the authors should demonstrate that with the shRNA experiments.

Our data show that Soxb1 is expressed in i-cells and its KD reduces the number of these stem cells (assessed by expression of Piwi1, an i-cell marker). Because i-cells give rise to all Hydractinia somatic lineages (and to germ cells), focusing specifically on Soxb2+ cells would provide no further insight because all cell types are expected to be affected. Indeed, injection of shRNA targeting Soxb1 resulted in smaller animals with multiple defects, including but not limited to the neural lineage.

1. Knockdown of SoxB genes resulted in complex defects in embryonic neurogenesis<br /> The manuscript aims to detail the roles of SoxB1, SoxB2, and SoxB3 in embryogenesis but only one of the main figures even shows pre-polyp life stages (Figure 7) and the results presented in in this figure are confusing. The authors suggest that knockdown of SoxB3 had no effect on embryonic neurogenesis but another interpretation of these data is that the SoxB3 shRNA simply did not work. The authors should provide additional support to show that this reagent is working as expected.

This information is included in Figure S11. Using mRNA in situ hybridization, we show that injection of shRNA targeting Soxb3 causes transcriptional downregulation of Soxb3 but not of Soxb2. The figure also shows the specificities of the shRNAs targeting Soxb1 and Soxb2.

Further, the results for SoxB1 and SoxB2 knockdown do not support the previous investigation of the role of SoxB2 in neurogenesis (Flici et al 2017). If SoxB1 is upstream of SoxB2, how does knockdown of SoxB1 have such a dramatic effect on RFamide neurons and nematocytes but knockdown of SoxB2 has an effect only on RFamide neurons? Is it possible the SoxB2 shRNA also wasn't working as expected? Can the results of the Flici et al 2017 paper showing SoxB2 knockdown in polyps be recapitulated using these shRNAs? If the point is to argue that embryos and adults (polyps) use fundamentally different mechanisms to drive neurogenesis, then the results presented in Figures 1-6 (which investigate SoxB genes in polyps) can't really be used to make inferences about embryonic neurogenesis. I think the authors have more work to do to demonstrate that embryonic and adult neurogenesis fundamentally differ.

The Soxb2 shRNA specificity is shown in Figure S11 (i.e., it KD Soxb2 but not Soxb1). We were equally surprised to discover that Soxb2 KD resulted in somewhat different phenotypes than the ones obtained by Flici et al. (2017) in polyps. At this stage, we cannot explain the difference. However, one could speculate that it resulted from slightly different regulation logic between embryonic and adult neurogenesis. More specifically, we propose different priorities for generating neural subtypes as explanation. Unfortunately, shRNAs work only with embryos, and long dsRNA mediated KD works only with polyps. CRISPR/Cas9-mediated KO is feasible in Hydractinia, but knocking out developmental genes, such as these Sox genes, would likely cause embryonic lethality. Other conditional KO/KD approaches are not available for Hydractinia. We believe we have made all possible efforts to clarify the roles of these genes using currently available techniques. Neurogenesis is a complex process that is only partially conserved among different animals and poorly studied in non-bilaterians. Furthermore, it is not possible to answer all questions in one study. As many studies before, our work contributes to the understanding of neurogenesis but also raises new questions. Addressing them is matter for future research. We have toned down the statement in the last sentence of the results and in the discussion and do not claim that embryonic and adult neurogenesis are fundamentally different.

Minor comments

Methods: A large bit of data from this manuscript relies on quantitative analysis of cell number but there's not enough information in the methods to understand how quantification was performed. How many slices from the z-stack were analyzed? Were counts made relative to the total tissue area in the X/Y dimension or relative to the number of total nuclei in the same section? How many individuals were examined for each analysis?

All cell counting analysis was performed using ImageJ/Fiji software. Counts were made relative to the total tissue area in the X/Y dimension (for the shRNA experiments). A Z-stack covering the whole depth of each larva was obtained. Counting was performed on cells positive for the respective cell type marker based on antibody staining and numbers were compared between shControl and shSoxb1/2/3 animals. At least 4 animals were counted per condition.

Page 11 - "Piwi2low cells, which are presumably i-cell progeny" - how were "high" and "low quantified?

“High” and “low” were not quantified. This is because i-cells progressively downregulate Piwi genes (i.e., Piwi1 and Piwi2) as they differentiate but this is a continuous process. Hence, it is difficult to put a threshold of Piwi1/Piwi2 protein level below which a cell ceases to be an i-cell while becoming a committed progeny. This is a similar process that is well documented in other animals where stemness markers are gradually downregulated during differentiation.

Page 13 - "a role in maintaining stemness" - this comment is not totally clear to me. Why would the number of EdU+ cells increase if the role of SoxB1 is to maintain stemness? Wouldn't SoxB1 knockdown then force stem cells to exit their program, resulting in early differentiation of i-cell progeny? This should be clarified.

KD of Soxb1 resulted in a decrease in the number of i-cells (i.e., Piwi1+ ones), suggesting that the gene is required for stemness maintenance. The increase in the numbers of cells in S-phase in this context was not related to i-cells because most of them were Piwi1-negative (Figure 7B). The identity of the cells in S-phase remains unknown, but a plausible explanation is that i-cell progeny (e.g., nematoblasts; see also next comment) increase their proliferative activity when i-cells numbers are low as a compensatory mechanism. This is merely a speculation. We have rephrased the paragraph to increase clarity.

Page 13 - "if progenitors are limiting" - if progenitors are limited why would there be an increase in nematocytes?

We do not have a definitive answer to this question but speculate that nematoblasts (i.e., stinging cell progenitors) account, at least in part, for the excessive proliferation seen under Soxb1 KD. This may constitute a mechanism allowing a depleted i-cell population to recover by self-renewal (instead of differentiation), moving temporarily the proliferation task to committed progeny (e.g., nematoblasts) until i-cell numbers return to normal. However, in the absence of evidence we refrain from expanding on this in the text.

Figures 1 and 2 claim to show "partial overlap" but they look perfectly overlapping to me. This makes the situation in Figure 4B difficult to interpret.

Figure 1 shows full overlap between Piwi1 and Sox1 expression and this is reflected in the text. Figure 2 shows no overlap between Soxb1 and Soxb2 in the lower body column (where only Soxb1 is expressed), overlap in the mid body region, and Soxb2 only expressing cells in the upper part of the body, just under the tentacle line. Similarly, the figure shows overlap between Soxb2/Soxb3 under the tentacle line, and predominantly Soxb3 above it in the head region. The small cartoons at the left side of each panel indicate its position along the oralaboral axis. See also our reply to the second part of comment #1.

Figure 4 - No indication of which part of the animal or which stage is shown in these images.

We have added cartoons to indicate the area in the polyp from where the images were taken.

Figure 5 - No indication of where these dissociated cells came from - polyps? Larvae?

All tissue samples were taken from feeding polyps; this is now mentioned in the Materials and Methods section.

Panel D is a bit perplexing - what are the "progeny" of Piwi+ cells if not SoxB2+ cells and their derivatives?

In Panel D, we show three cell fractions. One constitutes i-cells, based on high Piwi1 expression (green fluorescence of the Piwi1::GFP reporter transgene) and morphology; one fraction includes nematocytes, based on the characteristic nematocyst capsule, and one constitutes a mixture of other i-cell progeny. The latter includes different cell types, given that i-cells are thought to contribute to all lineages. They have only dim GFP fluorescence because the Piwi1 promoter-driven GFP shuts down upon i-cell differentiation. Soxb2+ cells are also among them but are not the only i-cell progeny.

Why are nematocytes but not neurons indicated?

Neurons are shown on Panels E & F. See also next comment.

Piwi seems to be maintained in Ncol-expressing cells but not in SoxB2- or RFamide-expressing cells? Does this suggest that Piwi is turned on in i-cells, off in SoxB2-expressing cells, and on again in terminally differentiating nematocytes? This would be quite surprising and should be verified with antibody labeling/imaging in Piwi transgenics to confirm the result. The resolution for Panel M is too low to evaluate this part of the figure.

The Piwi1i gene is downregulated upon i-cell differentiation. In the Piwi1:GFP reporter animal, residual GFP fluorescence persists post differentiation due to GFP's long half-life. The brightness of which depends on the time elapsed since differentiation. Because nematocytes are short living cells with high turnover, most nematocytes have recently differentiated and are therefore relatively bright green in the Piwi1::GFP animal. Neuron turnover is lower, making most neurons in the same transgenic animal appear dim. The resolution of the imaging flow cytometer is limited because the machine images 1000s of cells per second through all optical channels. However, it is high enough to allow the identification of features such as cell shape, some organelles (e.g., nematocytes), nuclear size and shape, and fluorescence intensity.

Figure 7 - the low magnification images provide nice overall context but the authors should also provide high magnification panels for the same images. Without them it is not possible to assess "defects in ciliation" or to determine if there are defects in GLWamide neurons from these knockdowns (e.g., neurite vs cell body defects). There's no mention of the fact that SoxB1 knockdown resulted in complete loss of RFamide cells, which is strange. Are there SoxB2-independent populations of RFamide? Panel B could be interpreted multiple ways - downregulation of Piwi in SoxB1 shRNA or upregulation in SoxB2/B3. The authors should provide an image of control shRNA-injected larvae with the same co-labeling of Piwi/EdU for context. From the images, it's not clear that there were differential effects of SoxB2 and SoxB3 on nematocytes.

The resolution of the images is, in fact, high, allowing it to be blown up on the screen. Even higher magnification of ciliation can be seen in Figure S12. KD of Soxb1 resulted in complete or nearly complete loss of Rfamide+ neurons. We have added this statement to the text as requested. Panel B shows the relative difference in Piwi1+ and S-phase cells between shSoxb1, shSoxb2, and shSoxb3-treated animals. The quantification relative to the control is presented in Figure 7C.

Figures 6 and S9 - why piwi2 and not piwi1?

In Figure 6, we co-stained the regenerates with two antibodies: one was a rabbit anti-GFP (to visualize the RFamide+ neurons), and the other was a guinea pig anti-Piwi2 (to visualize icells). The anti-Piwi1 antibody that was used in other images to visualize i-cells was raised in rabbit and could not be used in conjunction with the anti-GFP one.

Figure S1 - Kayal et al 2018 is the most recent phylogeny of cnidarians and should probably be cited in place of Zapata throughout the manuscript. Independent of this, the polytomy in Figure S1 panel A is not supported by either Zapata or Kayal and should be fixed.

We have cited Kayal et al. 2018 and revised the tree in Figure S1 as pointed.

Figure S3 - is this mRNA? Protein? Panels E-G are too small to interpret. Please provide stage/time for cartoons in panel H.

As per the legend, Panels A, B, D, E, F refer to protein; C is lectin staining (DSA), and G is EdU. The resolution of Panels E-G is actually high, allowing blowing up of the images on the screen to view the details. The stages of the cartoon in Panel H are now provided in the figure legend.

Figure S11 - please provide images of whole larvae as shown for Piwi knockdown in Fig S9 and some additional support (e.g., qPCR) to demonstrate the shRNAs are actually working.

Figure S9 represents immunostaining using the anti-Piwi1 antibody. In Figure S11, we show the specificity of the shRNA treatments; we used highly sensitive single-molecule mRNA in situ hybridization. Whole animal imaging is not informative due to the punctuated nature of the single-molecule staining.

Figure S12 - it's not clear what ciliary "defects" are being shown.

In the control, cilia are uniformly distributed along the oral-aboral axis whereas in the shSoxb1-injected animals, the pattern is patchy. Additionally, shSoxb1-injected larvae could not swim (planulae swim by coordinated cilia beat).

Reviewer #2 (Significance):

Generally, the results are either equivocal or the conclusions are not well supported by the results (as detailed above). The significance of this work to vertebrate neurobiology is somewhat weak. (Especially considering the orthology of these genes to bilaterian SoxB genes is not well supported.) Why not compare these results to other cnidarians - the expression patterns of SoxB1 and SoxB2 in corals and sea anemones seem to differ quite a lot (Shinzato et al 2008; Magie et al 2005), suggesting these genes are almost certainly not behaving in the same way across cnidarians. This is exciting! What's happening in Hydra? Seems like it should be possible to mine the single-cell data set from Siebert et al to test these hypothesized relationships between the Sox genes in another hydrozoan which constantly makes new neurons.

We have modified the concluding section in the discussion, in line with this comment. See also comment to Reviewer #3.

Reviewer #3 (Evidence, reproducibility and clarity):

This paper characterizes the role of Soxb genes in neurogenesis in Hydractinia. The authors use cutting edge approaches including FISH, transgenics, image flow cytometry, FACS and shRNA knock downs to characterize SoxB in Hydractinia. The images are beautiful, the data is sound and the interpretation of the data is appropriate.

I have only minor suggested listed by section below:

Abstract<br /> - The abstract and introduction should make clear that this is a colonial animal and the cell migration occurs from the aboral to the oral end of the polyp (not the animal, as there are many oral ends). This is relevant to the interpretation of the data as the polyps do not act in isolation as they interconnected and may communicate via the stolonal network that connects the polyps in the colony.

We have added a section to the Introduction to address the reviewer's comment. The Abstract, however, is too short to include this explanation.

• The human disease justification is a relatively weak one and does not need to be included. Using Hydractinia to understand the role of SoxB in the evolution of neurogenesis in animals is enough justification for the study.

We have adopted the reviewer's comment and modified the statement in the discussion (see also comment to Reviewer #2).

Introduction<br /> - Instead of Sox phylogenies (the term phylogeny is more appropriate for species trees), consider substituting, for Sox gene trees. And instead of "phylogenetic relation" use the term "orthology"

This has been done.

• The number of times the sentences that have the sentiment "....remain unknown." "....little is known.." "...unclear..." , "....difficult to establish...." etc. is distracting and detracts from what IS known about these genes. It is not necessary to continually justify the study throughout the introduction. Instead a clearer description of the background and setting up the question/hypothesis of SoxB paralog subfuctionalization in space and time - would be more informative to the reader.

We have reduced the number of occasions as recommended.

• The authors state that there are three SoxB genes in the Hydractinia genome? What genome? For several years there has been multiple papers published by subsets of these authors have used unpublished genome data, but the complete genome has yet to be released to the public. This is especially egregious because they cite their NSF funded EDGE proposal to CEF and UF which is supposed to develop tools to the community, and yet the community at large doesn't have access to the genome. If these data came from the genome, then the genome should be released. If these data came from a previously published transcriptome as in the previous SoxB paper then this should be stated explicitly.

The Hydractinia genome assembly, annotation, RNA-seq data, and genome browser are now available in the Hydractinia genome project portal at the National Human Genome Research Institute (NIH) website (https://research.nhgri.nih.gov/hydractinia/). The raw data have been deposited in the NCBI Sequence Read Archive (SRA) under BioProject PRJNA807936. This information has been added to the 'Resource availability' section.

Results<br /> - I assume there was no expression of Soxb2 and Soxb3 in the reproductive polyps? This should be stated explicitly.

Soxb2 expression in sexual polyps was consistent with the nervous system and with maternal deposition in oocytes. It was not detected in male germ cells. We have added a new in situ hybridization image of Soxb2 to Figure 12.

• The word "progeny" is used throughout to describe terminally differentiated cells. However, progeny implies offspring, but these are actually later stages of differentiation of the in a cell's ontogeny, thus the term should be changed to "differentiated cells"

We used "progeny" to indicate that the corresponding cells derived from a specific progenitor cell type. We did try replacing it with "differentiated cells" but this completely changes the meaning of the sentence: first, it does not include the cell of origin info and second, not all progeny are already fully differentiated.

• Typo on page 11 "This predictable generation of many new neurons provides an opportunity to study neurogenesis in [a ]regeneration." - Remove the "a"

Corrected.

• While the regeneration study is interesting, there is nothing revealed about the role of Soxb and there is not a lot of new information revealed about regenerations. Authors should better justify this section or consider omitting.

These sections demonstrate de novo neurogenesis in head regeneration. This was not known in this animal before.

Discussion<br /> - The authors assume that in the transgenic lineage, the fluorescent marker in differentiated cells is due to retention of fluorescence, but it is unclear if they can rule out that Soxb2 is still being expressed in those cells" Please clarify.

We conclude this by comparing the mRNA expression (Figures 1 & 2) with the fluorescent proteins (Figure 3).

• How did the authors determine that the shSoxb3 knockdown worked? Please discuss relevant controls and validation (either in discussion or methods). This is particularly important given that it didn't have an apparent phenotypic effect.

The efficacy of all shRNAs determined by in situ hybridization, showing that each shRNA downregulates its own target mRNA but not the others (Figure S11).

• Again, the connection to human health is a bit of a stretch. Instead, what is most interesting is the similarity of Soxb paralogs acting sequentially as has been found in vertebrates. This suggests a highly conserved mechanism of subfunctionization following gene duplication at the base of animals.

We agree. This is now also better highlighted in the discussion.

Figures<br /> - Its very hard to distinguish the overall abundance of Soxb2 and Soxb3 expression along the polyp body axis from the panels figure 2. A lower magnification or larger area in each region would be helpful

In Figure 2, we performed single-molecule in situ hybridization. While highly sensitive, this method generates spotty images because they highlight single molecules and are not coupled to an enzymatic reaction as in other methods. They mostly looks poor when showing low magnification images. Because a previous study (Flici et al. 2017) has already shown the general expression pattern, we aimed at providing the details of the transition.

• Figure 4 - either the figure is upside down or the text is upside down. It is also difficult to see the double staining (if any).

The figure is oriented to position the oral end up. The resolution of the panels is high, enabling blowing-up on the screen. The quality of in vivo time lapse images cannot match that of fixed and antibody stained ones, or of single in vivo images. This is because the animals are imaged for many hours during which they tend to bleach.

• Figure 5M is difficult to read due to the small print. Consider enlarging and moving it to Supplementary Material

The size of the text is small but the resolution is very high, enabling blowing up the image on the screen. We thought that the information was important enough to be presented in the main text and given that most readers would use the electronic version we preferred this option on another supplemental figure on top of the 12 we already have.

Reviewer #3 (Significance):

This is an interesting and important study because although it is well known that SoxB genes function in neurogenesis in animals, it is unclear how and if subfunctionalization occurs outside of vertebrates. Hydractinia is an excellent model to study SoxB genes because of its colonial organization and continuous development of nerve cells throughout the life of the animal. In addition, it is part of the early diverging cnidarian lineage and thus can provide insight into the relative conservation of SoxB genes across animals.

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

Learn more at Review Commons

---

Referee #3

Evidence, reproducibility and clarity

This paper characterizes the role of Soxb genes in neurogenesis in Hydractinia. The authors use cutting edge approaches including FISH, transgenics, image flow cytometry, FACS and shRNA knock downs to characterize SoxB in Hydractinia. The images are beautiful, the data is sound and the interpretation of the data is appropriate.

I have only minor suggested listed by section below:

Abstract<br /> - The abstract and introduction should make clear that this is a colonial animal and the cell migration occurs from the aboral to the oral end of the polyp (not the animal, as there are many oral ends). This is relevant to the interpretation of the data as the polyps do not act in isolation as they interconnected and may communicate via the stolonal network that connects the polyps in the colony.<br /> - The human disease justification is a relatively weak one and does not need to be included. Using Hydractinia to understand the role of SoxB in the evolution of neurogenesis in animals is enough justification for the study.

Introduction<br /> - Instead of Sox phylogenies (the term phylogeny is more appropriate for species trees), consider substituting, for Sox gene trees. And instead of "phylogenetic relation" use the term "orthology"<br /> - The number of times the sentences that have the sentiment "....remain unknown." "....little is known.." "...unclear..." , "....difficult to establish...." etc. is distracting and detracts from what IS known about these genes. It is not necessary to continually justify the study throughout the introduction. Instead a clearer description of the background and setting up the question/hypothesis of SoxB paralog subfuctionalization in space and time - would be more informative to the reader.<br /> - The authors state that there are three SoxB genes in the Hydractinia genome? What genome? For several years there has been multiple papers published by subsets of these authors have used unpublished genome data, but the complete genome has yet to be released to the public. This is especially egregious because they cite their NSF funded EDGE proposal to CEF and UF which is supposed to develop tools to the community, and yet the community at large doesn't have access to the genome. If these data came from the genome, then the genome should be released. If these data came from a previously published transcriptome as in the previous SoxB paper then this should be stated explicitly.

Results<br /> - I assume there was no expression of Soxb2 and Soxb3 in the reproductive polyps? This should be stated explicitly.<br /> - The word "progeny" is used throughout to describe terminally differentiated cells. However, progeny implies offspring, but these are actually later stages of differentiation of the in a cell's ontogeny, thus the term should be changed to "differentiated cells"<br /> - Typo on page 11 "This predictable generation of many new neurons provides an opportunity to study neurogenesis in [a ]regeneration." - Remove the "a"<br /> - While the regeneration study is interesting, there is nothing revealed about the role of Soxb and there is not a lot of new information revealed about regenerations. Authors should better justify this section or consider omitting.

Discussion<br /> - The authors assume that in the transgenic lineage, the fluorescent marker in differentiated cells is due to retention of fluorescence, but it is unclear if they can rule out that Soxb2 is still being expressed in those cells" Please clarify.<br /> - How did the authors determine that the shSoxb3 knockdown worked? Please discuss relevant controls and validation (either in discussion or methods). This is particularly important given that it didn't have an apparent phenotypic effect.<br /> - Again, the connection to human health is a bit of a stretch. Instead, what is most interesting is the similarity of Soxb paralogs acting sequentially as has been found in vertebrates. This suggests a highly conserved mechanism of subfunctionization following gene duplication at the base of animals.

Figures<br /> - Its very hard to distinguish the overall abundance of Soxb2 and Soxb3 expression along the polyp body axis from the panels figure 2. A lower magnification or larger area in each region would be helpful<br /> - Figure 4 - either the figure is upside down or the text is upside down. It is also difficult to see the double staining (if any).<br /> - Figure 5M is difficult to read due to the small print. Consider enlarging and moving it to Supplementary Material

Significance

This is an interesting and important study because although it is well known that SoxB genes function in neurogenesis in animals, it is unclear how and if subfunctionalization occurs outside of vertebrates. Hydractinia is an excellent model to study SoxB genes because of its colonial organization and continuous development of nerve cells throughout the life of the animal. In addition, it is part of the early diverging cnidarian lineage and thus can provide insight into the relative conservation of SoxB genes across animals.

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 present further investigation of the Sox transcription factors in the model Cnidarian Hydractinia. They showcase the Hydractinia as now a relatively technically advanced model system to study animal stem cells, regeneration and the control of differentiation in animal cells. In this study they characterise the neural cells in hydractinia using FACS and sing cell transcriptome sequencing, investigate the sequential expression of SoxB genes in the i-cells and presumptive lineage giving rise to i-cells and investigate the neuronal regeneration making good use of transgenic rules. Finally, they investigate the role of SoxB genes in embryonic neurogenesis.

There are no major or minor issues effecting the conclusions

Significance

This study helps to confirm the role of an important group of transcription factors is conserved across the metazoan as well as showcasing an exciting model organism for regeneration and stem cell biology. This will of interest to a broad audience of developmental and biologists.

My own research is in the same field, using a different model system

Referees cross-commenting

I agree with the comments from the other reviewers, and am sure the authors can address these adequately with further explanation.

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

Chrysostomou et al. investigate the role of three putative SoxB genes in embryonic neurogenesis in the colonial hydrozoan Hydractinia. They show that SoxB1 is co-expressed with Piwi in the multipotent i-cells and, using transgenics, they show that these Piwi/SoxB1 cells become neurons and gametes, consistent with the cell types that differentiate from i-cells. They further suggest that SoxB2 and SoxB3 are expressed downstream of SoxB1 in the progeny of the i-cells and, using shRNAs, investigate the role of SoxB genes on embryonic neurogenesis. The primary conclusions center on the similarity between neural differentiation in humans and Hydractinia as both systems pattern neurons using sequential expression of SoxB genes during the differentiation of neurons. The manuscript presents a large and diverse set of data derived from analysis of transgenic animals, single-cell sequencing, and investigation of gene function; despite this, the conclusions are either not particularly novel or not well-supported. The co-expression of SoxB1 in Piwi-expressing i-cells appears to be both novel and significant but the implications are not clearly indicated. Additional specific concerns are detailed below.

Major comments

1. SoxB genes act sequentially<br /> Knockdown of SoxB2 has already been shown to result in the loss of SoxB3, so the sequential action of SoxB genes in this animal does not seem to be a terribly novel conclusion. While this manuscript does appear to report the most comprehensive analysis of SoxB1 expression, the evidence for sequential activation of SoxB1 and then SoxB2 in the same lineage (Figure 4) is a bit troubling. Panel A of this figure appears to show complete overlap between SoxB1 and SoxB2, suggesting all the cells in this field are synchronously passing through the transition point from SoxB1 to SoxB2 expression. While this may reflect reality, it would be more convincing to see adjacent cells expressing SoxB1 only or SoxB2 only, reflecting the dynamic progression of cell type specification along the main body axis. Panel B is more concerning; while the authors have highlighted a cell that does appear to transition from SoxB1+ to SoxB1+/SoxB2+, there are several cells in the background that appear to gain SoxB2 expression without first expressing SoxB1. Do these cells constitute a fundamentally different, SoxB1-indpenendent, lineage of SoxB2+ cells? This would be noteworthy but is not mentioned or characterized. Figure 7 shows the effect of SoxB1 knockdown (by shRNA) on the number of Piwi-expressing cells, nematocytes, etc but why not show that SoxB2 and SoxB3 are also knocked down in these experiments? Figure S11 shows no effect of SoxB2 and SoxB3 knockdown on SoxB1 expression but why wasn't the reciprocal experiment performed? If SoxB2 and SoxB3 are really downstream of SoxB1, the authors should demonstrate that with the shRNA experiments.
2. Knockdown of SoxB genes resulted in complex defects in embryonic neurogenesis<br /> The manuscript aims to detail the roles of SoxB1, SoxB2, and SoxB3 in embryogenesis but only one of the main figures even shows pre-polyp life stages (Figure 7) and the results presented in in this figure are confusing. The authors suggest that knockdown of SoxB3 had no effect on embryonic neurogenesis but another interpretation of these data is that the SoxB3 shRNA simply did not work. The authors should provide additional support to show that this reagent is working as expected. Further, the results for SoxB1 and SoxB2 knockdown do not support the previous investigation of the role of SoxB2 in neurogenesis (Flici et al 2017). If SoxB1 is upstream of SoxB2, how does knockdown of SoxB1 have such a dramatic effect on RFamide neurons and nematocytes but knockdown of SoxB2 has an effect only on RFamide neurons? Is it possible the SoxB2 shRNA also wasn't working as expected? Can the results of the Flici et al 2017 paper showing SoxB2 knockdown in polyps be recapitulated using these shRNAs? If the point is to argue that embryos and adults (polyps) use fundamentally different mechanisms to drive neurogenesis, then the results presented in Figures 1-6 (which investigate SoxB genes in polyps) can't really be used to make inferences about embryonic neurogenesis. I think the authors have more work to do to demonstrate that embryonic and adult neurogenesis fundamentally differ.

Minor comments

Methods: A large bit of data from this manuscript relies on quantitative analysis of cell number but there's not enough information in the methods to understand how quantification was performed. How many slices from the z-stack were analyzed? Were counts made relative to the total tissue area in the X/Y dimension or relative to the number of total nuclei in the same section? How many individuals were examined for each analysis?

Page 11 - "Piwi2low cells, which are presumably i-cell progeny" - how were "high" and "low quantified?

Page 13 - "a role in maintaining stemness" - this comment is not totally clear to me. Why would the number of EdU+ cells increase if the role of SoxB1 is to maintain stemness? Wouldn't SoxB1 knockdown then force stem cells to exit their program, resulting in early differentiation of i-cell progeny? This should be clarified.

Page 13 - "if progenitors are limiting" - if progenitors are limited why would there be an increase in nematocytes?

Figures 1 and 2 claim to show "partial overlap" but they look perfectly overlapping to me. This makes the situation in Figure 4B difficult to interpret.

Figure 4 - No indication of which part of the animal or which stage is shown in these images.

Figure 5 - No indication of where these dissociated cells came from - polyps? Larvae? Panel D is a bit perplexing - what are the "progeny" of Piwi+ cells if not SoxB2+ cells and their derivatives? Why are nematocytes but not neurons indicated? Piwi seems to be maintained in Ncol-expressing cells but not in SoxB2- or RFamide-expressing cells? Does this suggest that Piwi is turned on in i-cells, off in SoxB2-expressing cells, and on again in terminally differentiating nematocytes? This would be quite surprising and should be verified with antibody labeling/imaging in Piwi transgenics to confirm the result. The resolution for Panel M is too low to evaluate this part of the figure.

Figure 7 - the low magnification images provide nice overall context but the authors should also provide high magnification panels for the same images. Without them it is not possible to assess "defects in ciliation" or to determine if there are defects in GLWamide neurons from these knockdowns (e.g., neurite vs cell body defects). There's no mention of the fact that SoxB1 knockdown resulted in complete loss of RFamide cells, which is strange. Are there SoxB2-independent populations of RFamide? Panel B could be interpreted multiple ways - downregulation of Piwi in SoxB1 shRNA or upregulation in SoxB2/B3. The authors should provide an image of control shRNA-injected larvae with the same co-labeling of Piwi/EdU for context. From the images, it's not clear that there were differential effects of SoxB2 and SoxB3 on nematocytes.

Figures 6 and S9 - why piwi2 and not piwi1?

Figure S1 - Kayal et al 2018 is the most recent phylogeny of cnidarians and should probably be cited in place of Zapata throughout the manuscript. Independent of this, the polytomy in Figure S1 panel A is not supported by either Zapata or Kayal and should be fixed.

Figure S3 - is this mRNA? Protein? Panels E-G are too small to interpret. Please provide stage/time for cartoons in panel H.

Figure S11 - please provide images of whole larvae as shown for Piwi knockdown in Fig S9 and some additional support (e.g., qPCR) to demonstrate the shRNAs are actually working.

Figure S12 - it's not clear what ciliary "defects" are being shown.

Significance

Generally, the results are either equivocal or the conclusions are not well supported by the results (as detailed above). The significance of this work to vertebrate neurobiology is somewhat weak. (Especially considering the orthology of these genes to bilaterian SoxB genes is not well supported.) Why not compare these results to other cnidarians - the expression patterns of SoxB1 and SoxB2 in corals and sea anemones seem to differ quite a lot (Shinzato et al 2008; Magie et al 2005), suggesting these genes are almost certainly not behaving in the same way across cnidarians. This is exciting! What's happening in Hydra? Seems like it should be possible to mine the single-cell data set from Siebert et al to test these hypothesized relationships between the Sox genes in another hydrozoan which constantly makes new neurons.

Transparent Peer Review

---

Download the complete Review Process [PDF] including:

• reviews
• authors' reply
• editorial decisions

</br>

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

Learn more at Review Commons

---

Reply to the reviewers

We are very grateful to the two referees for their constructive comments and suggestions which have helped improve the quality of our manuscript.

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

Ribes et al developed a FACS-based serological assay to detect antibodies against the SARS-CoV-2 spike protein in various hosts. The authors described an assay that is more sensitive and quantitative, allowing the detection of anti-spike antibodies with just a few ul of blood, and highlighted the potential of the assay as an alternative to commercial ELISA-based assays against SARS-CoV-2 spike protein.

Major concerns *

1. * On being quantitative analysis - the authors have used 20/130 reference serum from NIBSC as an example in figure 1. How does the RSS of the described assay compare/correlate with the Ab values in WHO standards? This should be included. * Response: We thank the referee for this helpful suggestion, and have now included the information on the IgG BAU in the legend of figure 1, and alluded to the characterisation of the 20/130 by the Expert Committee on Biological Standardization (Mattiuzzo et al., 2020) on lines 410-414 in the main text of the manuscript

* On sensitivity and specificity - AUC profiles should be performed and included. *

Response: If the Jurkat-flow test was intended for clinical use, the precise determination

of the sensitivity and specificity of the test would indeed be absolutely essential. As was already mentioned at the end of the introduction, the Jurkat-S&R-flow test is only destined to be used by research laboratories, for research purposes. This has now also been clarified at the end of the abstract : “Whilst the Jurkat-flow test is ill-suited and not intended for clinical use ….”

As suggested by the referee, to establish the sensitivity and specificity of a diagnostic test, it is indeed practical to use the Area Under the Receiver Operating Characteristic (ROC) curve (AUC). A ROC curve is a plot of the true positive rate (Sensitivity) in function of the false positive rate (100-Specificity) for different cut-off points of a parameter. Determining properly the sensitivity and specificity of a test thus requires large collections of samples which are known to be either certainly positive or certainly negative, which we did not have access to.

* Are there any cross-reactivity with the other spike proteins from other CoVs? If so, what is the level of cross-reactivity? *

Response: To assess cross-reactivity with other CoVs, we would have needed either Jurkat cells expressing the spike proteins from other CoVs, or sera with known reactivity against CoVs. Since we did not have access to such cells or sera, we were not in a position to address such a question.

* While the authors have showed that the flow-based assay has a more dynamic range, there is insufficient data showing that it is "more sensitive", as stated in the abstract. The authors should reflect this in the text. *

Response: In the abstract, we do not state that the Jurkat-S&R-flow test is more sensitive than the ELISA, but “at least as sensitive”. On the other hand, we state that it is more sensitive than the HAT test, which it clearly is since there are more than a dozen samples on figure 2 that were positive with either or both ELISA and Jurkat-S&R-flow but were negative by HAT.

Of note, we have recently described an improved protocol, called HAT-field, which significantly improves the sensitivity of HAT, albeit at the cost of decreased specificity (https://doi.org/10.1101/2022.01.14.22268980)

* Is trimer or monomer Spike expressed on the surface of the cells? *

Response: Several studies have shown that, when the spike protein is expressed in human cells after transfection or transduction, it is in its native trimeric form at the cells’ surface and can even cause fusion with cells expressing the ACE2 receptor. This has now been clarified in the introduction section.

* While there are significant advantages of the flow-based assay, the authors should discuss the limitations of a flow-based assay as a serological assay, especially for sero-surveillance and cohort studies. For instance, HTS application is usually limited for cell-based assays. In addition, while the assay is relatively cheap, it is worth nothing that the cytometer is an expensive equipment that not all laboratories have. *

Response: We bring the referee’s attention to the fact that those points are discussed at the end of the introduction (line 161-165) : “ Since the Jurkat-flow test calls for the use of both a flow cytometer and cells obtained by tissue culture, it is clearly not destined to be used broadly in a diagnostic context, but its simplicity, modularity, and performances both in terms of sensitivity and quantification capacities should prove very useful for research labs working on characterizing antibody responses directed against SARS-2, both in humans and animal models. “

Minor concerns*: *

1. * Figure 1 - text and numbers in the FACS plots are too small. Please adjust. In addition, for some of the FACS plots shown (eg. neg cont and serum 20/130), the population is right at the axis. Please pan the x-axis to allow better visualisation.
2. Figure 3A - please label axis.
3. Figure S2 - please label axis.
4. In general, please check through all figures for axis labels and also adjust the front size. For most, the text is too small. * Response: Sizes of numbers and text increased, and axis labels added in all figures*

Reviewer #1 (Significance (Required)):

As already discussed by the authors, there have already been quite a number of studies that have demonstrated the advantages of a flow-based assay for serological analysis for SARS-CoV-2. However, Ribes et al showed a new way to separate out alloreactivity from specific staining, which is important in reducing false positivity in serological assay. As more and more people receive their vaccination, there is a significant interest in immune-monitoring following vaccination. Given the more dynamic range of the flow-based assay, this is one good way to monitor antibody response. *

Expertise: My research interest focuses on the study of SARS-CoV-2 antibody responses following infection or vaccination. * *

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

In this paper, Joly and colleagues make use of a flow cytometry-based assay to measure in a reliable and sensitive manner the presence of IgG, IgA and IgM in blood samples from post-COVID human patients and also from laboratory (mouse and hamster) and domestic animals (dogs and cats). They find that the test is appropriate to detect the presence of humoral immunity in all species tested. The manuscript is clearly written and the Figures are clearly presented. The experiments with rodente deliberately infected with inactivated SARS-CoV-2 shows (Fig. 3) that the method is reliable and able to clearly discriminate positive from negative sera. Interestingly, dogs and cats were sampled from households in which the owners had been found to have passed COVID-19 by PCR. Among this cohort of house animals they find more than 90% seroconversion for dogs and slightly less than 30% of clear seroconversion in cats. We find however that the manuscript would benefit by establishing a clear cut-off value of "Specific Stain" for dogs and cats (Fig. 3). This could be implemented by including data from pre-COVID dog and cat sera or in its defect, sera from those species collected at households in which their owners were vaccinated and did not pass the infection. Another point of criticism that could be resolved is that the channels for flow cytometry in Figure 1 do not seem to be adequately compensated and there is evidence of some cross-contamination between FL1 and FL3. *

Responses: We thank the referee for bringing our attention to the fact that we had presented the data on sera from cats and dogs in a confusing manner, which led the referee to believe that the sets of samples presented were representative of the population of animals whose owner had tested positive for Covid-19. In fact, for this experiment, which was only ever intended as a preliminary proof of concept that the test could be adapted very simply to companion animals, we used sets of sera which we knew would contain approximately 50 % of positive and 50 % negative samples because they had previously been screened by sero-neutralisation (incidentally, a manuscript by Bessière et al., describing that work on sera from 131 cats and 156 dogs, has very recently been submitted for publication). To prevent possible confusions, we have now reworded the description of this proof of concept experiment, in the legend of figure 3, the text, and the methods section.

Regarding the question of a clear cut-off value, as when using human samples, we would suggest using a value of 40 for the instruments settings we used, corresponding to an RSS of 20 (i.e. 20 fold the value of the negative control). With such a value, it can be seen that one cat serum would be considered positive whilst showing no neutralising activity, but one dog serum which showed weak neutralising activity would be considered negative. If anything, this example highlights the difficulty in setting a precise cut off value for any biological test.

Regarding the question of inadequate compensation between channels 1 and 3, this is due to the fact that the Cellquest software does not allow for FL1/FL3 compensation, which is explained in the figure legend (see lines 208-210). We decided to simply draw the gates as they appear on figure 1 because attempts at post-acquisition compensation using the Flowjo software did not give satisfactory results. Incidentally, no compensation is required when samples are acquired on a Fortessa flow cytometer, where mCherry can be excited by a different laser (see figure S1) or if one uses the Jurkat-S&G-flow version of the test as in figure 3D for hamster sera (using Jurkat-GFP as negative control, and secondary antibodies conjugated to Alexa 488).

Minor points*: *

*-Figure 1.- Please describe the y- and x-axis. Such as they are is difficult to find out. *

Response: Done

* -It would be advisable to mention in Materials and Methods (page 22) how blood was collected from cats and dogs. *

Response: We thank the referee for highlighting this, and have now provided the information in the relevant method section.

* -Line 856, page 22, "ad libidum" should be "ad libitum" *

Response: We thank the referee for spotting this typo, which has been corrected

* Reviewer #2 (Significance (Required)):

This is another step in the implementation of flow cytometry tests, instead of ELISA or CLIA serological tests based on the use of recombinant proteins, as a more sensitive and reliable method. The description of the high frequency of human-domestic animal transfer of SARS-CoV-2 will also add to the idea that it is humans who transmit the virus to those animals. *

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 paper, Joly and colleagues make use of a flow cytometry-based assay to measure in a reliable and sensitive manner the presence of IgG, IgA and IgM in blood samples from post-COVID human patients and also from laboratory (mouse and hamster) and domestic animals (dogs and cats). They find that the test is appropriate to detect the presence of humoral immunity in all species tested.

The manuscript is clearly written and the Figures are clearly presented. The experiments with rodente deliberately infected with inactivated SARS-CoV-2 shows (Fig. 3) that the method is reliable and able to clearly discriminate positive from negative sera. Interestingly, dogs and cats were sampled from households in which the owners had been found to have passed COVID-19 by PCR. Among this cohort of house animals they find more than 90% seroconversion for dogs and slightly less than 30% of clear seroconversion in cats.

We find however that the manuscript would benefit by establishing a clear cut-off value of "Specific Stain" for dogs and cats (Fig. 3). This could be implemented by including data from pre-COVID dog and cat sera or in its defect, sera from those species collected at households in which their owners were vaccinated and did not pass the infection. Another point of criticism that could be resolved is that the channels for flow cytometry in Figure 1 do not seem to be adequately compensated and there is evidence of some cross-contamination between FL1 and FL3.

Minor points:

• Figure 1. Please describe the y- and x-axis. Such as they are is difficult to find out.
• It would be advisable to mention in Materials and Methods (page 22) how blood was collected from cats and dogs.
• Line 856, page 22, "ad libidum" should be "ad libitum"

Significance

This is another step in the implementation of flow cytometry tests, instead of ELISA or CLIA serological tests based on the use of recombinant proteins, as a more sensitive and reliable method. The description of the high frequency of human-domestic animal transfer of SARS-CoV-2 will also add to the idea that it is humans who transmit the virus to those animals.

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

Ribes et al developed a FACS-based serological assay to detect antibodies against the SARS-CoV-2 spike protein in various hosts. The authors described an assay that is more sensitive and quantitative, allowing the detection of anti-spike antibodies with just a few ul of blood, and highlighted the potential of the assay as an alternative to commercial ELISA-based assays against SARS-CoV-2 spike protein.

Major concerns:

1. On being quantitative analysis - the authors have used 20/130 reference serum from NIBSC as an example in figure 1. How does the RSS of the described assay compare/correlate with the Ab values in WHO standards? This should be included.
2. On sensitivity and specificity - AUC profiles should be performed and included.
3. Are there any cross-reactivity with the other spike proteins from other CoVs? If so, what is the level of cross-reactivity?
4. While the authors have showed that the flow-based assay has a more dynamic range, there is insufficient data showing that it is "more sensitive", as stated in the abstract. The authors should reflect this in the text.
5. Is trimer or monomer Spike expressed on the surface of the cells?
6. While there are significant advantages of the flow-based assay, the authors should discuss the limitations of a flow-based assay as a serological assay, especially for sero-surveillance and cohort studies. For instance, HTS application is usually limited for cell-based assays. In addition, while the assay is relatively cheap, it is worth nothing that the cytometer is an expensive equipment that not all laboratories have.

Minor concerns:

1. Figure 1 - text and numbers in the FACS plots are too small. Please adjust. In addition, for some of the FACS plots shown (eg. neg cont and serum 20/130), the population is right at the axis. Please pan the x-axis to allow better visualisation.
2. Figure 3A - please label axis.
3. Figure S2 - please label axis.
4. In general, please check through all figures for axis labels and also adjust the front size. For most, the text is too small.

Significance

As already discussed by the authors, there have already been quite a number of studies that have demonstrated the advantages of a flow-based assay for serological analysis for SARS-CoV-2. However, Ribes et al showed a new way to separate out alloreactivity from specific staining, which is important in reducing false positivity in serological assay. As more and more people receive their vaccination, there is a significant interest in immune-monitoring following vaccination. Given the more dynamic range of the flow-based assay, this is one good way to monitor antibody response.

My research interest focuses on the study of SARS-CoV-2 antibody responses following infection or vaccination.

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)):

Summary: This manuscript documents a very thorough biophysical, structural and functional dissection of interactions between the RNA-binding protein Rrm4 and the endosomal adaptor Upa1 in the filamentous fungus Ustilago maydis. It has been shown previously that the Rrm4-Upa1 interaction is critical for mRNA transport in this system as mRNAs hitchhike on motor-associated endosomes. Here, the authors reveal using modelling that Rrm4 has three MLLE domains, including a cryptic one that had not been identified previously. They then report the crystal structure of MLLE2 and analyze the distribution anf arrangement of the MLLE domains in the protein using SAXS. They then show using pulldowns and isothermal titration calorimetry that MLLE3 is critical for the Upa1 interaction (via the PAM2L domains of Upa1) and that MLLE2 contributes to Rrm4 localization in vivo when the MLLE3-Upa1 interaction is partially impaired. The study suggests that Rrm4 has a platform of MLLE domains for orchestrating Rrm4 function. Overall, this is technically a high quality study. However, a number of points (mostly minor) should be addressed.

Major comments:

__A key part of the study if the in vivo work illustrating a role for MLLE2 in regulating Rrm4 localization when the system is sensitized. Some aspects of this part of the work need clarifying.

a) The authors should show that the abberant staining is indeed microtubule-related with the benomyl experiment that they used in Jankowski et al. 2019. __

We included this important control in Figure EV5F demonstrating that the aberrant staining is no longer visible after the microtubule inhibitor benomyl treatment

b) The authors claim from these experiments that MLLE2 contributes to endosomal targeting (as there is ectopic protein on other structures (presumptive microtubules)). However, to make this claim, the authors would need to measure the intensity of the mutant Rrm4 protein on endosomes and/or the colocalization of these Rrm4 variants with endosomes, as they do in other experiments in this paper. Otherwise, it is possible that the MLLE2 deletion has another effect, e.g. increasing protein stability, and thus increasing the likelihood of binding to structures other than endosomes. If available, data on the relative abundance in the cell of the protein expressed from the wild-type control (rrm4-kat) and MLLE2 deletion constructs (e.g. rrm4-m1,2delta-kat) should be provided.

As indicated by the reviewer, a critical point is identifying a function of MLLE2. Surprisingly, the domain is conserved in evolution, but , we do not see a mutant phenotype under optimal culture conditions. Therefore, we challenged the system and observed the mislocalisation of Rrm4, if the MLLE2 domain is deleted. However, the overall amount of shuttling Rrm4-positive endosomes was not strongly affected according to our kymograph experiments. We observe aberrant staining, which is not seen with the Rrm4 wild-type protein. Thus, under challenging conditions, we do see a function of MLLE2.

To address the valid point of the reviewer, we quantified the signal intensities in kymographs of the most important Rrm4 variants. As indicated in Figure 5E, we observed that the maximum fluorescence intensity in kymograph signals was reduced when Rrm4 variants are mislocalised to microtubules while the minimum intensities were comparable in all strains. This underlines that a subset of Rrm4 molecules are no longer shuttling through the cell and most likely are attached to microtubules (to prove the involvement of microtubules, we did benomyl treatment which is now shown in Figure EV5F). We also included a Western Blot experiment (Figure EV5G) demonstrating that neither MLLE1 nor MLLE2 deletion impacts the total protein amount of Rrm4. These data support the notion that MLLE2 contributes to endosomal targeting.

c) Was the data in Figure 5D scored blind of the identity of the samples? Given that the classification has to be done manually, it is important to confirm the phenotypes are robust to blinding (at least for the key comparisons).

We agree entirely that manual evaluation of microscopic images has to be carried out with utmost care. The phenotype of aberrant microtubule staining is not easily detectable, and it needs an experienced person to quantify this. The data were analyzed by a second experimentalist with experience in evaluating microscopy images to validate the system’s robustness. Notably, the key findings were confirmed in both cases aberrant microtubule staining was only observed when the MLLE domain was mutated. However, the second person reported difficulties in differentiating a bundle of Rrm4 signals or stained microtubules. Therefore, this person quantified higher values with less experience in Rrm4 movement. In essence, we can rely on the key findings. We included the information in the section “Materials and methods” and gave the comparison in Figure EV5H.

If points b and c are addressed, it should be possible to draw an arrow between the gray question mark protein in Figure 6 and the endosome surface, which is what I assume the authors believe to be case based on their discussion.

Having addressed both points, we have also improved the model. To this end, we added a second unknown protein component (grey oval with a question mark) that interacts with MLLE2 and the endosomal surface. Thereby the hierarchical order with the accessory role of MLLE2 during endosomal attachment is stressed.

Minor comments:

1. The first line of the abstract is quite bold. It is hard to quantify the role of transport vs RNA stability for example, so I suggest this sentence is toned down. Correct, the first line now reads, “Spatiotemporal expression can be achieved by transport and translation of mRNAs at defined subcellular sites”.

Line 269: change "amount of motile Rrm4-M12delta-Kat positive signals" to "number of motile Rrm4-M12delta-Kat positive signals".

Changed as mentioned above.

Figure 3 legend: Insert "Variant" before "amino acids of the FxP and FxxP..." to indicate what is labeled in gray. Change "fond" to "font" in the same sentence.

Corrected as mentioned above.

The cartoons of the different protein variants are very helpful but I had problems spotting the Upa1-Pam2L deletions due to the similar gray to the background of the protein. This would perhaps be clearer if the gray used for the background was lighter than it currently is.

We improved the contrast by reducing the background of Upa1 to a lighter grey tone in all the corresponding figures.

The residual motility of wild-type Rrm4 when PAM2L1 and PAM2L2 are both mutated (Figure 5C) is reminiscent of what is seen in a complete Upa1 deletion in the group's previous work. It would be helpful to point this out to the reader, as well as the implication that other proteins are contributing to Rrm4's linkage to endosomes. After all, some of these other adaptors might contact MLLE2 of Rrm4.

We addressed this point by referring to our previous publication with the following sentence: “Comparable to previous reports, we observed residual motility of Rrm4-Kat on shuttling the endosomes if both PAM2L motifs are mutated or if upa1 is deleted. This indicates that additional proteins besides Upa1 are involved in the endosomal attachment of Rrm4 (Pohlmann et al., 2015).”

Some of the y-axes of the charts should be more descriptive so that the reader can understand the plots even before they consult the legends. For example, in Figure EV4A and EV5D and E, which protein is being to referred to in each 'number of signals' plot should be included. In Figure 5D, 'Hyphae [%]' would be clearer as 'Hyphae with MT staining of Rrm4 [%]'

We improved this in Figures EV4, 5D and EV5.

Figure EV5 legend title: this could be misleading as the authors are seeing ectopic MT localization rather than a deficit in microtubule association.

Corrected to “Deletion of MLLE1Rrm4 and -2 cause aberrant staining of microtubules”.

Reviewer #1 (Significance (Required)):

__The Feldbrugge group has previously mapped interactions between Upa1 and Rrm4 (Pohlmann et al., 2015) and some conclusions are corroborated in the paper by Boehm et al. The paper under review is, however, a significant advance due to the identification of the third MLLE domain, detailed biophysical characterization of the interactions, the structural insights, and evidence of a subsidiary role of MLLE2. The work would of course be stronger if the target of MLLE2 had been identified but I think this is beyond the scope of this initial work. To my knowledge, this is one of the most extensive analyses of the interactions mediated by MLLE and PAM domains and will be of interest to others working on these protein features. The work will also appeal to those interested in the links of localizing mRNAs with motor-associated membranes, which is an emerging field.

Reviewer expertise: I have a long-standing interest in molecular analysis of mRNA trafficking mechanisms. I do not have experience in fungal genetics. __

**Referee Cross-commenting**

It seems that we are in agreement that this is solid work and that biochemical and biophysical analysis of the MLLE-PAM interactions will be of significant interest to those working on those domains (or proteins containing those domains). I agree with the comments of the other reviewers and there are clearly some essential minor revisions needed to strengthen the evidence for their conclusions and some clarifications. I think it is a long shot that RNA binding to the RRMs will affect the MLLE-PAM interactions and would require quite a lot of work to show this conclusively. The study would, however, be more impactful if this was shown to be the case, or the target of MLLE2 was found. Nonetheless, I would not say these new avenues of research are necessary to find a home in one of the Review Commons journals.

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

Devan, Schott-Verdugo et al.

Summary

In this study the putative MLLE RNA-binding motifs of the endosomal RNA-binding protein, Rrm4, from Ustilago maydis were examined using structural and genetic analyses. MLLE motifs are conserved in polyA-binding proteins (Pab1/PABPC1) and found also in Rrm4, which was shown to reside on motile endosomes and deliver septin mRNAs for endosome-localized translation during polarized growth. Upa1 on the endosome interacts with Rrm4 via its PAM2L domain that itself interacts with the MLLE domains of proteins like Pab1. Mutations in the known MLLE domain of Rrm4 were earlier shown to affect localization to endosomes. Here, the C-terminal domain of Rrm4 was revealed to have three divergent MLLE motifs using comparative modeling; only two of which were previously predicted. Crystallization and X-ray diffraction analysis of a truncated version of bacterially produced Rrm4, showed MLLE2 is most similar to that of PABPC1 and UBR5, although MLLE1 and 2 are somewhat divergent in the key region of PAM2 binding. Small angle X-ray scattering of recombinant full-length or truncated Rrm4 revealed that the MLLE domains might form a platform that could allow for multiple contacts with different binding partners. In vitro binding studies with different N-terminal GST-tagged versions of the Rrm4 were used to examine for interactions with PAM2 sequences of Upa1 using N-terminal hexa-histidine-SUMO fusions. It was found that Pab1-MLLE interacts with the PAM2, but not PAM2L, domain of Upa1. In contrast, the complete Rrm4 MLLE region (G-Rrm4-NT4) interacted with the PAM2L domain, but not the PAM2 of Upa1. Notably, the interaction with PAM2L required the third MLLE and neither MLLE1 nor MLLE2, nor both. No significant differences in affinity were observed and were similar to that of the Pab1 MLLE. The results also show that the MLLE3 has a higher affinity for the PAM2L2 than PAM2L1 of Upa1.

To examine the biological role of the Rrm4 MLLEs, U. maydis strains bearing deletions in the domains of Rrm4 were examined for hyphal growth and endosomal transport (latter using Upa1-GFP and Rrm4-mKate2). Only the loss of the MLLE3 domain inhibited polarized growth (as seen with the full deletion of RRM4) and not the deletion of either MLLE1 or 2. Similar results were obtained regarding endosome shuttling. Thus, in line with the biochemical experiments performed the MLLE3 domain alone (of the three identified) is necessary for the biological actions of Rrm4. This suggested the MLLE1 and 2 are not necessary for function under these conditions.

To examine this further, Upa1 carrying mutations in the PAM2L 1or PAM2L2 domains were examined. It was found that the deletion of both PAM2L domains affected unipolar growth resulting in bipolar growth similar to the deletion of UPA1 alone. This phenotype was observed even upon the deletion of Rrm4 MLLE1 and 2 in the same background as the PAM2L mutants. The mutation of both PAM2L domains led to a reduction in Rrm4-labeled shuttling endosomes, which suggests that these domains help anchor Rrm4 to endosomes. When only the PAM2L1 domain is present in Upa1 there was a larger increase in hyphae with aberrant microtubule staining than upon the loss of PAM2L1. The authors suggest that this indicates PAM2L2 is more important and prescribes an accessory role for MLLE2 in endosome association.

Comments: Overall, the study seems well conducted. We cannot comment on the structural aspect of the work since this is not our field of expertise. That said, the biochemical and genetic/functional studies appear solid, well thought-out, and clearly presented. No new experiments are necessary to support the general claims of the paper, however, experiments suggested below might make it more revealing with regards to the connection between RNA binding and MLLE-PAM2L interactions (i.e. endosome localization and RNA binding functions).

1. Line 286 - It reads the they "Next, we investigated the association of Rrm4 -M12D-Kat in strains expressing PAM2L1. Thus, the endosomal attachment was solely dependent on the interaction of MLLE3 with the PAM2L2 sequence of Upa1." Unclear - wouldn't lacking PAM2L1 (and not expressing) fit the logic of the sentence? We corrected this with the sentence, “Next, we investigated the association of Rrm4-M1,2D-Kat in strains expressing Upa1 with mutated PAM2L1”.

Several questions regarding the specificity of PAM2 vs. PAM2L domains. What happens when you switch/replace the PAM2L1 or 2 of Upa1 with Upa1 PAM2 domains? Are they exclusive? What happens when the MLLE3 of Rrm4 is switched with that of Pab1? And if one does both - does that restore functionality to Rrm4?

These are very interesting suggestions. Previously, we have shown that a single PAM2L1 or PAM2L2 sequence of Upa1 is sufficient for unipolar growth and recruitment of Rrm4 to endosomes. Please note that Upa1 with mutated PAM2L1 and L2 still contains a PAM2 motif. Furthermore, mutating the PAM2 motif of Upa1 did not affect Rrm4 shuttling or unipolar growth. Thus, switching the domains would mostly address whether the precise location within Upa1 would be important. This is interesting but, unfortunately very labour-intensive and beyond the manuscript’s current scope.

Switching MLLE3 with MLLE of PAB1 is an interesting approach. One might expect that Rrm4 can be recruited to endosomes again. However, Rrm4 would also interact with numerous other proteins containing PAM2 motifs like deadenylase Not4. Here it would compete with the MLLE of Pab1. Thus, it would be expected that Rrm4 is on the surface, but the protein will be mistargeted to other proteins causing pleiotropic alterations. It will be difficult to judge whether Rrm4 functionality is restored or whether other processes are disturbed. In essence, these are stimulating ideas, but we believe that these experiments are beyond the scope of the current study. In the future, we might address this point by using a heterologous peptide-binding pocket or tethering approach.

Likewise, what happens if Upa1 only has PAM2L2 instead of only PAM2L1 domains? Does that alter function - perhaps now one can observe a contribution of MLLE1? If it it's there it's likely to have function. Anything known about the post-translational modification of these MLLE or PAM domains? Does it change during unipolar vs. bipolar growth? Perhaps the different MLLE domains are regulated in such a fashion?

Again also very valid points. Upa1 with two PAM2L2 motifs might interact stronger. The problem is that one PAM2L motif is sufficient for interaction, and we do not see a strong phenotype.

Currently, we do not know if post-translational modifications regulate the MLLE domains. This could alter the binding affinity or specificity, and by expressing fungal proteins in E. coli, we might have missed this type of regulation. However, we addressed the function of MLLE1 and MLLE2 in U. maydis using a genetic approach. We deleted the corresponding domains and interfered with potential regulation by posttranslational modification. Thus, we cannot exclude post-translational modification, but it appears to be not essential for function. We will address the posttranslational regulation of Rrm4 in more detail in the future.

Can the authors show whether the binding of mRNA cargo (e.g. Cdc3 mRNA) to the RRM motifs of Rrm4 affects the interaction between any of the MLLE-PAM2L pairs, or vice versa (i.e. does the MLLE-PAM2L interaction affect mRNA binding)?

In previous studies, we have investigated a version of Rrm4 carrying a mutation in the first RRM motif of Rrm4. According to RNA live imaging, the respective strains exhibit a loss of function phenotype and mRNA transport is strongly affected. However, the endosomal association of Rrm4-mR1-Gfp is not affected, indicating no direct cross-talk between RNA-binding via RRM1 and endosomal attachment via MLLE3. Also, a version of Rrm4 carrying a deletion of all three RRM domains is still shuttling on endosomes. The two functions, i.e. RNA binding and endosomal binding, appears to be carried out by two independent platforms, i.e. three RRMs and three MLLEs, respectively. The overall structure of the protein also reflects this. The RRM domains are structurally clearly separated from the flexible MLLE domains.

Discussion line 311 It is written that the three MLLE domains "collaborate for optimal functionality..." Perhaps there's a misunderstanding here, but the authors show that MLLE3 domain alone is necessary & sufficient for function, so where is the collaboration? MLLE2 may have an accessory role according to the authors, but we do not know if it is in collaboration with MLLE3 or independent thereof. Since the KD of MLLE3 is not affected by the presence or absence of MLLE1,2 in vitro at least, it may be that they have independent, and not collaborative, roles.

Correct, we rephrased this more carefully. We omitted the collaboration aspect. It now reads, ”but a sophisticated binding platform consisting of three MLLE domains with MLLE2 and MLLE3 functioning in linking the key RNA transporter to endosomes.”

Reviewer #2 (Significance (Required)):

This paper concerns functional domains found in an endosome-localized RNA binding protein, U. maydis Rrm4, which is necessary for localized translation on endosomes and subsequent unipolar growth. Here the authors show using structural, biochemical, and genetic studies that instead of one or two MLLE protein-protein interacting domain in Rrm4 there are three, although one (MLLE3) is necessary and sufficient for full function. This work is for an audience interested in those studying RNA trafficking and its role in cell physiology, which is our expertise. The work is interesting, but it could be made more so especially if a connection was established between the RNA-binding function of the RRM domains and the MLLE-PAM2L interaction(s). At this point it is solid technical work and could be published after minor revisions.

**Referee Cross-commenting**

I concur with the comments of the other reviewers in that the work is solid and necessitates minor revisions in order to be published. Clearly, establishing a connection between the RNA-binding function and the MLLE-PAM interactions of Rrm4 would be an interesting and worthy pursuit that might enhance the novelty of the work, but I agree that it could belong to future studies.

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

__ Summary: Long-distance subcellular transport of mRNAs is achieved through selective and dynamic interaction with the transport machinery. Using the highly polarized hyphae of Ustilago maydis, the authors previously showed i- that mRNAs can hitchhike on actively transported endosomes for proper distribution, and ii- that the connection between mRNAs and endosomes is mediated by the interaction between a C-terminal MademoiseLLE (MLE) domain of the RNA binding protein Rrm4 and the Upa1 adapter protein. In this study, the authors aimed at more precisely characterizing the structural and molecular bases underlying the Rrm4-Upa1 interaction. Combining structural modeling and X-ray analyses, they discovered a non-canonical, and previously missed, MLE domain (MLE1) in Rrm4, and characterized the structure of the second MLE domains (MLE2) of Rrm4. Through binding assays, they showed that the three MLE domains exhibit different binding properties, and that MLE3 is the only domain capable of binding to the PAM2 domain of Upa1. Consistent with this finding, functional assays performed in U. maydis revealed that MLE3 is the main domain involved in interaction with endosomes and trafficking, MLE1 and 2 having either no or minor functions in this process.

The manuscript is very-well written, the data are of high quality and clearly presented. A wide range of complementary approaches has been used to molecularly and functionally characterize the different MLE domains of Rrm4. From an "RNA transport" perspective, this manuscript falls short of a main novel findings as the domains characterized in this study (MLE1 and 2) don't have a clear function in connecting mRNAs to the transport machinery. From an "MLE domain" perspective, this work however provides interesting information about non-canonical domains and structures, and about binding and function specificity. As described below, my major concern relates to the role played by the ML2 domain of Rrm4, a role referred to as "accessory" by the authors. __

__

Major comments: __

The authors conclude from their results that ML2 has an accessory role in promoting association with endosomes.

1- This conclusion is made based on in vivo experiments showing that a form of Rrm4 lacking the M2 domain, in contrast to wild-type Rrm4, aberrantly attached to MTs in a context where the Rrm4-Upa1 interaction mediated by MLE3Rrm4 has been weakened (Upa1-pl2m). Although the results are convincing, their interpretation is less. The authors, indeed, claim that the observed phenotype results from "the static accumulation of Rrm4" due to reduced interaction with endosomes. Why then don't they see a decrease in the motility/transport properties of Rrm4-M2Δ in this context then? Also, do the authors see a decrease in the co-localization of Rrm4-M2Δ with endosomes (which would be expected if the interaction is decreased)? Can the authors perform IP or co-sedimentation experiments to strengthen their hypothesis?

This is a fair criticism that was also raised by reviewer 1. In the improved version of the manuscript, we now include important control experiments demonstrating that (i) the aberrant localisation is microtubule-dependent (Fig. EV5F) (ii) the mutations do not cause differences in protein amounts of Rrm4 (Fig. EV5G) (iii) the key findings of the aberrant microtubule staining, which were scored manually in microscopic images were verified independently by two persons (Fig. EV5H) and (iv) most importantly, Rrm4 signal intensity is decreased in processive signals of our kymograph analysis (Fig. 5E). We firmly believe that this set of experiments strengthens our conclusion that MLLE2 plays an accessory role in the endosomal attachment (Fig. 6).

2- Whether MLE2Rrm4 mediates interaction with endosomes through association with Upa1 is unclear, as the binding assays performed in Figure 3 test for association of Rrm4 variants with single isolated domains of Upa1, not with the full-length protein. Assessing the binding of Rrm4-M2Δ variants with Upa1-PL2m would help interpreting the phenotypes described in Figure 5.

Unfortunately, it is difficult to express full-length Upa1 protein in E. coli due to the presence of extended unstructured regions. To overcome this limitation, we performed yeast two-hybrid experiments with full-length proteins of Rrm4 and Upa1. We were able to recapitulate qualitatively the results observed in vitro using the individual domains.

Notably, the Rrm4 version carrying a deletion in MLLE1 and MLLE2 interacted with Upa1 versions carrying mutations in PAM2L1 or PAM2L2 (Fig. EV3C), suggesting that both MLLE domains of Rrm4 are dispensable for interaction with Upa1. MLLE3 is sufficient to interact with a single PAM2L sequence of Upa1. This suggests the presence of additional interaction partners for MLLE1 and MLLE2 and is entirely consistent with our genetic and cell biological analysis described in Fig. 5.

__

Minor comments: __

1- The authors have previously characterized the effect of a C-terminal deletion of Rrm4 on Rrm4 motility and binding to Upa1 (Becht et al., 2006; Pohlmann et al., 2015). How their previously-described construct compares to the Rrm4-M3Δ used in this study is unclear (is it the same?).

It is the identical mutation to allele rrm4GPD from Becht et al. 2006. We indicate the information in the text “(Fig. 4B-C; mutation identical to allele rrm4GPD in Becht et al., 2006).”

2- page 6, line 141: refer to Fig. 1B rather than Fig. EV1A ?

We included the reference to Fig. 1B.

3- page 10, line 274: "Rrm4-Kat was found"

We corrected this.

4- page 11, line 286: "in strains expressing Upa1-PAM2L1", replace by "in strains expressing Upa1 with mutated PAM2L1"?

We corrected this.

5- The Figures and accompanying legends are overall very clear and detailed. In Figures EV4A and EV5D-E, it would however help if the authors would indicate on the Figure itself, left to each panel which markers/signals is being analyzed (e.g Rrm4-Kat (top) and Upa1-GFP (down) for Figure EV4).

We clarified this.

Reviewer #3 (Significance (Required)):

Active transport of mRNAs along microtubule tracks has been shown to play a key role in the spatio-temporal control of gene expression in various cell types and species. How specific mRNAs mechanistically connect to molecular motors for their transport to their subcellular destination has however for long remained largely unclear. Recent work, including work from the authors, has uncovered that RNAs can hitchhike on membranous organelles through adapter proteins linking mRNAs and RNA binding proteins with trafficking membrane-bound organelles.

This study aimed at investigating the structural and molecular bases underlying the interaction between RNA binding proteins and endosomes. While their identification and characterization of the MLE1 and MLE2 domains of Rrm4 did not provide significant new insight into the mechanisms involved in the endosome-mediated transport of mRNAs, it uncovered interesting new properties of MLE domains, including structural variations, selective binding and functional specificity. This work should thus be of interest for structural biologists and researchers interested in protein-protein interaction platforms.

**Referee Cross-commenting**

Our comments all converge to the idea that this study is solid as it is and requires only minor revision work to support the authors conclusions. Although characterizing further MLE/PAM2 binding specificity and MLE2 interactors would be of great interest and indeed provide a more complete understanding of interaction networks at play, I feel that this is beyond expected revision work.

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:

Long-distance subcellular transport of mRNAs is achieved through selective and dynamic interaction with the transport machinery. Using the highly polarized hyphae of Ustilago maydis, the authors previously showed i- that mRNAs can hitchhike on actively transported endosomes for proper distribution, and ii- that the connection between mRNAs and endosomes is mediated by the interaction between a C-terminal MademoiseLLE (MLE) domain of the RNA binding protein Rrm4 and the Upa1 adapter protein.

In this study, the authors aimed at more precisely characterizing the structural and molecular bases underlying the Rrm4-Upa1 interaction. Combining structural modeling and X-ray analyses, they discovered a non-canonical, and previously missed, MLE domain (MLE1) in Rrm4, and characterized the structure of the second MLE domains (MLE2) of Rrm4. Through binding assays, they showed that the three MLE domains exhibit different binding properties, and that MLE3 is the only domain capable of binding to the PAM2 domain of Upa1. Consistent with this finding, functional assays performed in U. maydis revealed that MLE3 is the main domain involved in interaction with endosomes and trafficking, MLE1 and 2 having either no or minor functions in this process.

The manuscript is very-well written, the data are of high quality and clearly presented. A wide range of complementary approaches has been used to molecularly and functionally characterize the different MLE domains of Rrm4. From an "RNA transport" perspective, this manuscript falls short of a main novel findings as the domains characterized in this study (MLE1 and 2) don't have a clear function in connecting mRNAs to the transport machinery. From an "MLE domain" perspective, this work however provides interesting information about non-canonical domains and structures, and about binding and function specificity.

As described below, my major concern relates to the role played by the ML2 domain of Rrm4, a role referred to as "accessory" by the authors.

Major comments:

The authors conclude from their results that ML2 has an accessory role in promoting association with endosomes.

1- This conclusion is made based on in vivo experiments showing that a form of Rrm4 lacking the M2 domain, in contrast to wild-type Rrm4, aberrantly attached to MTs in a context where the Rrm4-Upa1 interaction mediated by MLE3Rrm4 has been weakened (Upa1-pl2m). Although the results are convincing, their interpretation is less. The authors, indeed, claim that the observed phenotype results from "the static accumulation of Rrm4" due to reduced interaction with endosomes. Why then don't they see a decrease in the motility/transport properties of Rrm4-M2Δ in this context then? Also, do the authors see a decrease in the co-localization of Rrm4-M2Δ with endosomes (which would be expected if the interaction is decreased)? Can the authors perform IP or co-sedimentation experiments to strengthen their hypothesis?

2- Whether MLE2Rrm4 mediates interaction with endosomes through association with Upa1 is unclear, as the binding assays performed in Figure 3 test for association of Rrm4 variants with single isolated domains of Upa1, not with the full-length protein. Assessing the binding of Rrm4-M2Δ variants with Upa1-PL2m would help interpreting the phenotypes described in Figure 5.

Minor comments:

1- The authors have previously characterized the effect of a C-terminal deletion of Rrm4 on Rrm4 motility and binding to Upa1 (Becht et al., 2006; Pohlmann et al., 2015). How their previously-described construct compares to the Rrm4-M3Δ used in this study is unclear (is it the same?).

2- page 6, line 141: refer to Fig. 1B rather than Fig. EV1A ?

3- page 10, line 274: "Rrm4-Kat was found"

4- page 11, line 286: "in strains expressing Upa1-PAM2L1", replace by "in strains expressing Upa1 with mutated PAM2L1"?

5- The Figures and accompanying legends are overall very clear and detailed. In Figures EV4A and EV5D-E, it would however help if the authors would indicate on the Figure itself, left to each panel which markers/signals is being analyzed (e.g Rrm4-Kat (top) and Upa1-GFP (down) for Figure EV4).

Significance

Active transport of mRNAs along microtubule tracks has been shown to play a key role in the spatio-temporal control of gene expression in various cell types and species. How specific mRNAs mechanistically connect to molecular motors for their transport to their subcellular destination has however for long remained largely unclear. Recent work, including work from the authors, has uncovered that RNAs can hitchhike on membranous organelles through adapter proteins linking mRNAs and RNA binding proteins with trafficking membrane-bound organelles.

This study aimed at investigating the structural and molecular bases underlying the interaction between RNA binding proteins and endosomes. While their identification and characterization of the MLE1 and MLE2 domains of Rrm4 did not provide significant new insight into the mechanisms involved in the endosome-mediated transport of mRNAs, it uncovered interesting new properties of MLE domains, including structural variations, selective binding and functional specificity. This work should thus be of interest for structural biologists and researchers interested in protein-protein interaction platforms.

Referee Cross-commenting

Our comments all converge to the idea that this study is solid as it is and requires only minor revision work to support the authors conclusions. Although characterizing further MLE/PAM2 binding specificity and MLE2 interactors would be of great interest and indeed provide a more complete understanding of interaction networks at play, I feel that this is beyond expected revision work.

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

Devan, Schott-Verdugo et al.

Summary

In this study the putative MLLE RNA-binding motifs of the endosomal RNA-binding protein, Rrm4, from Ustilago maydis were examined using structural and genetic analyses. MLLE motifs are conserved in polyA-binding proteins (Pab1/PABPC1) and found also in Rrm4, which was shown to reside on motile endosomes and deliver septin mRNAs for endosome-localized translation during polarized growth. Upa1 on the endosome interacts with Rrm4 via its PAM2L domain that itself interacts with the MLLE domains of proteins like Pab1. Mutations in the known MLLE domain of Rrm4 were earlier shown to affect localization to endosomes.

Here, the C-terminal domain of Rrm4 was revealed to have three divergent MLLE motifs using comparative modeling; only two of which were previously predicted. Crystallization and X-ray diffraction analysis of a truncated version of bacterially produced Rrm4, showed MLLE2 is most similar to that of PABPC1 and UBR5, although MLLE1 and 2 are somewhat divergent in the key region of PAM2 binding. Small angle X-ray scattering of recombinant full-length or truncated Rrm4 revealed that the MLLE domains might form a platform that could allow for multiple contacts with different binding partners. In vitro binding studies with different N-terminal GST-tagged versions of the Rrm4 were used to examine for interactions with PAM2 sequences of Upa1 using N-terminal hexa-histidine-SUMO fusions. It was found that Pab1-MLLE interacts with the PAM2, but not PAM2L, domain of Upa1. In contrast, the complete Rrm4 MLLE region (G-Rrm4-NT4) interacted with the PAM2L domain, but not the PAM2 of Upa1. Notably, the interaction with PAM2L required the third MLLE and neither MLLE1 nor MLLE2, nor both. No significant differences in affinity were observed and were similar to that of the Pab1 MLLE. The results also show that the MLLE3 has a higher affinity for the PAM2L2 than PAM2L1 of Upa1. To examine the biological role of the Rrm4 MLLEs, U. maydis strains bearing deletions in the domains of Rrm4 were examined for hyphal growth and endosomal transport (latter using Upa1-GFP and Rrm4-mKate2). Only the loss of the MLLE3 domain inhibited polarized growth (as seen with the full deletion of RRM4) and not the deletion of either MLLE1 or 2. Similar results were obtained regarding endosome shuttling. Thus, in line with the biochemical experiments performed the MLLE3 domain alone (of the three identified) is necessary for the biological actions of Rrm4. This suggested the MLLE1 and 2 are not necessary for function under these conditions.

To examine this further, Upa1 carrying mutations in the PAM2L 1or PAM2L2 domains were examined. It was found that the deletion of both PAM2L domains affected unipolar growth resulting in bipolar growth similar to the deletion of UPA1 alone. This phenotype was observed even upon the deletion of Rrm4 MLLE1 and 2 in the same background as the PAM2L mutants. The mutation of both PAM2L domains led to a reduction in Rrm4-labeled shuttling endosomes, which suggests that these domains help anchor Rrm4 to endosomes. When only the PAM2L1 domain is present in Upa1 there was a larger increase in hyphae with aberrant microtubule staining than upon the loss of PAM2L1. The authors suggest that this indicates PAM2L2 is more important and prescribes an accessory role for MLLE2 in endosome association.

Comments:

Overall, the study seems well conducted. We cannot comment on the structural aspect of the work since this is not our field of expertise. That said, the biochemical and genetic/functional studies appear solid, well thought-out, and clearly presented. No new experiments are necessary to support the general claims of the paper, however, experiments suggested below might make it more revealing with regards to the connection between RNA binding and MLLE-PAM2L interactions (i.e. endosome localization and RNA binding functions).

1. Line 286 - It reads the they "Next, we investigated the association of Rrm4 -M12D-Kat in strains expressing PAM2L1. Thus, the endosomal attachment was solely dependent on the interaction of MLLE3 with the PAM2L2 sequence of Upa1." Unclear - wouldn't lacking PAM2L1 (and not expressing) fit the logic of the sentence?
2. Several questions regarding the specificity of PAM2 vs. PAM2L domains. What happens when you switch/replace the PAM2L1 or 2 of Upa1 with Upa1 PAM2 domains? Are they exclusive? What happens when the MLLE3 of Rrm4 is switched with that of Pab1? And if one does both - does that restore functionality to Rrm4?
3. Likewise, what happens if Upa1 only has PAM2L2 instead of only PAM2L1 domains? Does that alter function - perhaps now one can observe a contribution of MLLE1? If it it's there it's likely to have function. Anything known about the post-translational modification of these MLLE or PAM domains? Does it change during unipolar vs. bipolar growth? Perhaps the different MLLE domains are regulated in such a fashion?
4. Can the authors show whether the binding of mRNA cargo (e.g. Cdc3 mRNA) to the RRM motifs of Rrm4 affects the interaction between any of the MLLE-PAM2L pairs, or vice versa (i.e. does the MLLE-PAM2L interaction affect mRNA binding)?
5. Discussion line 311 It is written that the three MLLE domains "collaborate for optimal functionality..." Perhaps there's a misunderstanding here, but the authors show that MLLE3 domain alone is necessary & sufficient for function, so where is the collaboration? MLLE2 may have an accessory role according to the authors, but we do not know if it is in collaboration with MLLE3 or independent thereof. Since the KD of MLLE3 is not affected by the presence or absence of MLLE1,2 in vitro at least, it may be that they have independent, and not collaborative, roles.

Significance

This paper concerns functional domains found in an endosome-localized RNA binding protein, U. maydis Rrm4, which is necessary for localized translation on endosomes and subsequent unipolar growth. Here the authors show using structural, biochemical, and genetic studies that instead of one or two MLLE protein-protein interacting domain in Rrm4 there are three, although one (MLLE3) is necessary and sufficient for full function. This work is for an audience interested in those studying RNA trafficking and its role in cell physiology, which is our expertise. The work is interesting, but it could be made more so especially if a connection was established between the RNA-binding function of the RRM domains and the MLLE-PAM2L interaction(s). At this point it is solid technical work and could be published after minor revisions.

Referee Cross-commenting

I concur with the comments of the other reviewers in that the work is solid and necessitates minor revisions in order to be published. Clearly, establishing a connection between the RNA-binding function and the MLLE-PAM interactions of Rrm4 would be an interesting and worthy pursuit that might enhance the novelty of the work, but I agree that it could belong to future studies.

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

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

Summary:

This manuscript documents a very thorough biophysical, structural and functional dissection of interactions between the RNA-binding protein Rrm4 and the endosomal adaptor Upa1 in the filamentous fungus Ustilago maydis. It has been shown previously that the Rrm4-Upa1 interaction is critical for mRNA transport in this system as mRNAs hitchhike on motor-associated endosomes. Here, the authors reveal using modelling that Rrm4 has three MLLE domains, including a cryptic one that had not been identified previously. They then report the crystal structure of MLLE2 and analyze the distribution anf arrangement of the MLLE domains in the protein using SAXS. They then show using pulldowns and isothermal titration calorimetry that MLLE3 is critical for the Upa1 interaction (via the PAM2L domains of Upa1) and that MLLE2 contributes to Rrm4 localization in vivo when the MLLE3-Upa1 interaction is partially impaired. The study suggests that Rrm4 has a platform of MLLE domains for orchestrating Rrm4 function. Overall, this is technically a high quality study. However, a number of points (mostly minor) should be addressed.

Major comments:

A key part of the study if the in vivo work illustrating a role for MLLE2 in regulating Rrm4 localization when the system is sensitized. Some aspects of this part of the work need clarifying.

a) The authors should show that the abberant staining is indeed microtubule-related with the benomyl experiment that they used in Jankowski et al. 2019.

b) The authors claim from these experiments that MLLE2 contributes to endosomal targeting (as there is ectopic protein on other structures (presumptive microtubules)). However, to make this claim, the authors would need to measure the intensity of the mutant Rrm4 protein on endosomes and/or the colocalization of these Rrm4 variants with endosomes, as they do in other experiments in this paper. Otherwise, it is possible that the MLLE2 deletion has another effect, e.g. increasing protein stability, and thus increasing the likelihood of binding to structures other than endosomes. If available, data on the relative abundance in the cell of the protein expressed from the wild-type control (rrm4-kat) and MLLE2 deletion constructs (e.g. rrm4-m1,2delta-kat) should be provided.

c) Was the data in Figure 5D scored blind of the identity of the samples? Given that the classification has to be done manually, it is important to confirm the phenotypes are robust to blinding (at least for the key comparisons).

If points b and c are addressed, it should be possible to draw an arrow between the gray question mark protein in Figure 6 and the endosome surface, which is what I assume the authors believe to be case based on their discussion.

Minor comments:

1. The first line of the abstract is quite bold. It is hard to quantify the role of transport vs RNA stability for example, so I suggest this sentence is toned down.
2. Line 269: change "amount of motile Rrm4-M12delta-Kat positive signals" to "number of motile Rrm4-M12delta-Kat positive signals".
3. Figure 3 legend: Insert "Variant" before "amino acids of the FxP and FxxP..." to indicate what is labeled in gray. Change "fond" to "font" in the same sentence.
4. The cartoons of the different protein variants are very helpful but I had problems spotting the Upa1-Pam2L deletions due to the similar gray to the background of the protein. This would perhaps be clearer if the gray used for the background was lighter than it currently is.
5. The residual motility of wild-type Rrm4 when PAM2L1 and PAM2L2 are both mutated (Figure 5C) is reminiscent of what is seen in a complete Upa1 deletion in the group's previous work. It would be helpful to point this out to the reader, as well as the implication that other proteins are contributing to Rrm4's linkage to endosomes. After all, some of these other adaptors might contact MLLE2 of Rrm4.
6. Some of the y-axes of the charts should be more descriptive so that the reader can understand the plots even before they consult the legends. For example, in Figure EV4A and EV5D and E, which protein is being to referred to in each 'number of signals' plot should be included. In Figure 5D, 'Hyphae [%]' would be clearer as 'Hyphae with MT staining of Rrm4 [%]'
7. Figure EV5 legend title: this could be misleading as the authors are seeing ectopic MT localization rather than a deficit in microtubule association.

Significance

The Feldbrugge group has previously mapped interactions between Upa1 and Rrm4 (Pohlmann et al., 2015) and some conclusions are corroborated in the paper by Boehm et al. The paper under review is, however, a significant advance due to the identification of the third MLLE domain, detailed biophysical characterization of the interactions, the structural insights, and evidence of a subsidiary role of MLLE2. The work would of course be stronger if the target of MLLE2 had been identified but I think this is beyond the scope of this initial work. To my knowledge, this is one of the most extensive analyses of the interactions mediated by MLLE and PAM domains and will be of interest to others working on these protein features. The work will also appeal to those interested in the links of localizing mRNAs with motor-associated membranes, which is an emerging field.

Reviewer expertise: I have a long-standing interest in molecular analysis of mRNA trafficking mechanisms. I do not have experience in fungal genetics.

Referee Cross-commenting

It seems that we are in agreement that this is solid work and that biochemical and biophysical analysis of the MLLE-PAM interactions will be of significant interest to those working on those domains (or proteins containing those domains). I agree with the comments of the other reviewers and there are clearly some essential minor revisions needed to strengthen the evidence for their conclusions and some clarifications. I think it is a long shot that RNA binding to the RRMs will affect the MLLE-PAM interactions and would require quite a lot of work to show this conclusively. The study would, however, be more impactful if this was shown to be the case, or the target of MLLE2 was found. Nonetheless, I would not say these new avenues of research are necessary to find a home in one of the Review Commons journals.

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

Reviewer #1: General comments:

Fujimoto and collaborators use Nanopore-based cDNA sequencing for genome-wide transcriptome analysis of a collection of hepatocellular carcinomas (HCCs) and matched normal liver tissues. To improve detection of alternatively spliced isoforms and hybrid transcripts potentially deriving from genomic rearrangements, they develop a dedicated pipeline SPLICE, which they benchmark against available software used for the same analysis. Besides having dual functionality (calls both alternative transcripts and fused transcripts), SPLICE seems to outperform previous software in calling alternative/fused transcripts and accuracy. They use the SPLICE pipeline to call isoforms and gene fusions in normal liver cells and HCCs and perform basic functional validations on novel fusions identified. The manuscript is well written, and the analyses are well performed. Perhaps the benchmarking of the SPLICE pipeline could have been more extensive (i.e., performed on additional independent datasets).

Major points: 1. Line 149-150: "We compared the results of mapping to the reference genome and the reference transcriptome sequences, and removed candidates if both were inconsistent (removal of mapping errors). " Please specify what "both were inconsistent" means.

Our reply; Thank you for this comment. The accuracy of fusion gene detection is influenced by mapping errors. To remove possible mapping errors, SPLICE aligned reads to the reference genome and the reference transcriptome sequences and compared the results. If the results are inconsistent (for example, GeneA-GeneB in the reference genome and GeneA-GeneB in the transcriptome genome, or GeneA-GeneB in the reference genome and GeneA in the transcriptome genome), SPLICE considers the candidates as false positive and removes them from the analysis.

          We changed the sentence “We compared the results of mapping to the reference genome and the reference transcriptome sequences, and removed candidates if both were inconsistent (removal of mapping errors).” to “we compared the results of mapping to the reference genome and the reference transcriptome sequences, and removed candidates if both results did not detect same fusion genes (removal of mapping errors).”  (line 150-152).

• Concerning TE-derived novel exons, in principle, this may lead to altered expression of the TE-transcript (as the Authors report for L1-MET) or to altered splicing of the transcript (i.e., other exon/introns could be retained or excluded). Can the Authors assess whether the inclusion of the TE in a transcript enhances its expression or affects the splicing of the "parental" transcript? If so, can they verify if the position of the insertion of the TE has any effect on expression and splicing?*

Our reply; Thank you very much for this important comment. As the reviewer mentioned, exonization of TE may affect the splicing patterns and gene expression levels of transcripts. To determine the effect of TE on expression levels, we compared the expression levels of transcripts with TE-derived novel exons with those of known transcripts of the gene. We found that the expression levels of transcripts with TE-derived novel exon were lower than those of known transcripts (Figure 1 in the reply). Since the same results were observed in all novel transcripts (Fig. 1E,F), most TE exonization would not affect the expression level of transcripts.

          We then analyzed the effects of TE in the splicing change, we compared the numbers of novel splicing junctions between transcripts with TE-derived novel exons and other transcripts in each gene. The proportions of genes with novel splicing junctions were not significantly different between the transcripts with TE-derived novel exons and others (transcripts with TE-derived novel exons; 9.1% and others; 11.9%)  (Figure 2 in the reply). As observed in L1-*MET* and L2-*RHR1*, transposons can affect expression levels and structures of transcripts, however, their effect would be limited to a part of genes.

Figure 1

Comparison of expression levels of transcripts with TE-derived novel exon and known transcripts. Only transcripts derived from genes with TE-derived novel exons were compared. The total number of transcripts is shown below the plot. Transcript abundance was measured in reads per million reads (RPM), and log10 converted values for RPM were shown in the violinplot. P-values were calculated by Wilcoxon rank-sum test.

Figure 2

Comparison of the percentage of novel splicing junction in transcripts with novel TE-derived exon and other transcripts. The total number of genes are shown below the plot. Transcripts with TE-derived novel exons and other transcripts were compared. P-value was calculated by Fisher’s exact test.

• Can the Authors explain why the NBEAL1-RPL12 was not detected by SPLICE?*

Our reply; Thank you for this comment. Although NBEAL1-RPL12 fusion was detected by SPLICE, mapping results to the reference genome and the reference transcriptome were inconsistent and removed from the final result. AsNBEAL1-RPL12 was not validated by PCR (Supplemental Fig. S4B) (line 183-184), we consider that this fusion-gene is a false positive, and filtering of SPLICE successfully removed false-positive fusions.

• Line 332: Can the Authors explain how the total amount of HVB mRNA was determined in each sample? Is it a relative amount calculated from the sequencing data? If so, it should be made clear in the text that this is a fractional measure.*

Our reply; Thank you very much for this comment. Expression levels were calculated by log10 converted reads per million reads (log10(RPM)) for each sample. We added the following sentences to the "Expression from HBV" subsection in the Results (line 337-338); “Expression levels were estimated by log10 converted support reads per million reads (log10(RPM)) for each sample.”.

• Fig4a: please specify if the y-axis "number of support reads" reports library normalized values.*

Our reply; Thank you for this comment. The values of the y-axis are row read counts. We added the following sentences to the Figure legend (line 348); “Y-axis shows the total number of support reads (raw counts).”.

• HCCs have more HBV-human genome fusion transcripts than normal liver. Could the authors clarify if these HCC transcripts are selectively found in tumors? or whether they are also expressed in normal liver samples? The paragraph starting from line 356 is confusing, and it is difficult to retrieve the above information for both HBs and HBx fusions.*

Our reply; We apologize for the confusing description. All HBV-human genome fusion transcripts were selectively expressed in tumor or normal liver. We added the following sentence to the "Expression from HBV" subsection in the Results (line 365-366); “All of these HBV-human genome fusion transcripts were selectively expressed in the HCCs and the livers.”.

• Figure 4C: what was the control used to calculate the relative viability in these analyses?*

Our reply; Thank you for this comment. Fig. 4C shows the number of HBV-human fusion transcripts in the six categories. If this comment refers to Fig. 4H, cell lines transfected with the empty vector (pIRES2-AcGFP1-Nuc) was used as controls. This has been described in the "Gene overexpression" subsection of Methods (line 716-717).

• MYT1L: the Authors report the identification of a novel MYT1L transcript downregulated in HCC, and argue it may have a potential tumor-suppressive function. For the sake of clarity, it will be advisable to show also the differential expression (HCC vs. Liver) of the other transcripts expressed from the same locus.*

Our reply; Thank you for this important comment. In HCCs and normal livers, only the novel MYT1L transcript was expressed from this locus, and no known transcript of MYT1L was expressed. We changed the sentence “In the MYT1Lgene, a highly-conserved novel exon was detected (Fig. 2E), and this transcript was significantly down-regulated in the HCCs” to “In the MYT1L gene, a highly-conserved novel exon was detected (Fig. 2E), and only a transcript with the novel exon was expressed.” (line 471-472).

• *

Minor points: 1. Table S4: there is a typo, correct “secific” in “specific”

Our reply; Thank you very much for this comment. We corrected the typo of Table S4.

• *

• *

*Reviewer #2: General comments:

Summary: This is both a presentation of a pipeline for analysis of Nanopore RNA-seq data, as well as an analysis of a cohort of 44 hepatocellular carcinomas against matched-normal liver tissue. It presents a number of quite intriguing results from the long-read RNA analysis, and suggests potential new targets for study in HCC. It is also worth noting that the current version of guppy (6) has functionality to detect primer sequences in the middle of reads and split those reads, which may obviate one of the steps in SPLICE.*

*Major comments:

1) The work done in this study used data that was basecalled using guppy 3.0.3. Since that version, I am aware of at least two major upgrades to the base caller accuracy, which would likely also improve the accuracy of isoform resolution. Given that the data is relatively low-coverage and that you have an automated workflow for the analysis, I would recommend re-basecalling using an updated basecaller and re-running your analysis using that. This is especially important given your comments in the paper about splice site misalignment.*

Our reply; Thank you very much for this important comment. We performed basecalling of a sequence data of MCF7 using the latest guppy v6.0.6 and compared the result with that by guppy v3.0.3. We randomly extracted 1M reads from MCF-7 reads that passed qscore filtering in guppy basecaller. The same reads were extracted and basecalled by guppy v3.0.3. These two data were analyzed by SPLICE.

The average error rate was 4.6 % for v6.0.6 and 6.8 % for v3.0.3. The number of transcripts was 9,674 for v6.0.6 and 9,329 for v3.0.3. Of these, the number of novel transcripts was 446 and 410, respectively. The number of fusion genes was 2 (BCAS3-BCAS4, and BCAS3-ATXN7) by v6.0.6 and one (BCAS3-BCAS4) by v3.0.3. As the reviewer mentioned, we found that using the latest version of guppy improved the accuracy and detected a larger number of transcripts.

We added the results to Supplemental Table S12. We also changed the sentences from “Second, our analysis removed the change of splicing sites within 5 bp to remove alignment errors (Fig. 1B). We consider that this cutoff value is necessary due to currently available high-error reads (S____upplemental Data S____2). However, sequencing technologies and basecallers are improving, and in the near future, we should be able to use a smaller cutoff value and identify larger numbers of splicing changes.” to “Second, the accuracy of the analysis depends on the sequencing error rate. Although several filters are used for currently available high-error reads (Fig. 1B and ____Supplemental____ Fig. S1), sequencing errors would affect the accuracy of the result. Sequencing technologies and basecallers are improving, and in the near future, we should be able to identify larger numbers of splicing changes with high accuracy (Supplemental Table S10).” (line 538-542).

2) You have compared your software to another tool for isoform analysis on Nanopore sequencing data, TALON. But a number of other tools exist for this purpose, including stringtie2, flair and bambu. My own testing has shown that stringtie2 outperforms TALON in terms of concordance with Illumina RNA-seq. It is quite important that you perform a complete comparison of your software to the state of the art for this purpose.

Our reply; Thank you very much for this important comment. We compared our tool with four tools (TALON, FLAIR, StringTie, and bambu). For this comparison, we used sequence data of MCF-7 and HCC (RK107C). We randomly extracted 1 M reads from MCF-7 and HCC (RK107C) sequence data using Seqtk (v1.3) (params: sample -s1 1000000). Reads were mapped to the reference genome sequence (hg38) with minimap2 (v2.17) (params: -ax splice --MD), and the output SAM files were converted to BAM files and sorted with samtools (v1.7) (Li et al. 2009).

For benchmarking of TALON (v5.0), we corrected aligned reads with TranscriptClean (v2.0.3) (Wyman and Mortazavi 2018). Next, we ran the talon_label_reads module to flagging reads for internal priming (params: --ar 20). TALON database was initialized by running the talon_initialize_database module (params: --l o --5p 500 --3p 300). Then, we ran the talon module to annotate the reads (params: --cov 0.8 --identity 0.8). To output transcript abundance, we first obtained a whitelist using the talon_filter_transcripts module (params: --maxFracA 0.5 --minCount 5), and then quantified transcripts using the talon_abundance module based on the whitelist. For FLAIR (v1.5), the sorted BAM file was converted to BED12 using bin/bam2Bed12.py. We then corrected misaligned splice sites with the flair-correct module. High-confidence isoforms were defined from the corrected reads using the flair-collapse module (params: -s 3 --generate_map). For benchmarking of StringTie (v2.2.1), Stringtie was performed with input files consisting of long-read alignment and reference annotation (params: -L -c 3). For benchmarking of bambu (v2.0.0), Bambu was performed with input files consisting of long-read alignment, reference annotation and reference genome (hg38) (params: min.readCount = 3). Candidates with low expression levels (support reads As a result, SPLICE identified the third-highest number of transcripts followed by FLAIR and StringTie (Supplemental Fig. S3A). In MCF-7 the concordance rate with IsoSeq MCF-7 transcriptome data was the highest in SPLICE for known transcripts and the second highest in SPLICE for novel transcripts (Supplemental Fig. S3B). These results indicate that SPLICE has sufficient accuracy for analyzing transcript aberrations.

We added the text to the "Comparison of SPLICE method with other tools" subsection of the Results (line 165-177) and the "Benchmarking" subsection of the Methods (line 640-679). We added the results to Supplemental Fig. S3.

3) Likewise, for fusion detection, you compare to LongGF. You should also compare to (and cite) JAFFAL.

Our reply; Thank you very much for this important comment. We compared our tool with the two tools (LongGF and JAFFAL). We used 1 M reads randomly extracted from MCF-7 and HCC (RK107C) sequence data as described above.

          For benchmarking of LongGF (v0.1.2), reads were mapped to the reference genome sequence (hg38) with minimap2 (v2.17) (params: -ax splice --MD), and the output SAM files were converted to BAM files and sorted with samtools (v1.7). We then ran the *longgf* module and obtained the list of fusion genes (params: min-overlap-len 100 bin_size 50 min-map-len 200 pseudogene 0 secondary_alignment 0 min_sup_read 3). For benchmarking of JAFFAL (v2.2), we ran the *JAFFAL.groovy* module with zipped fastq files.

          In this comparison, close gene pairs (We added the text to the "Comparison of SPLICE method with other tools" subsection in the Results (line 178-186) and the "Benchmarking" subsection in the Methods (line 667-679). We showed the results in Supplemental Fig. 4.

4) In terms of the source code, I have questions. Why did you use BASH to run the Python code, instead of making this into a Python package? Why did you not use the functionality already available in BioPython for a number of basic sequence data handling tasks? Why is there not even a single function defined anywhere, let alone classes?

At some level, if it works, it works. But I have serious concerns about the long-term maintainability of the code in its current state.

Our reply; Thank you very much for this critical comment. As the reviewer mentioned, we think it is better to make a python package and use BioPython for maintenance and long-term maintainability of the code. We have been building our analysis pipeline by trial and error, and at this stage, the current scripts are convenient for us (our group may need to learn software development). We provided a Docker package (see the reply to comment 5)), and this would promote usability.

5) Also related to the code, it is generally the standard now to create a BioConda package or Docker container for a bioinformatics package. BioConda has the advantage that the BioContainers project automatically generate Docker and Singularity containers from it. Please provide one of these.

Our reply; Thank you very much for this critical comment. We made a Docker file and provided it from our github page. It is available from the "Installation and usage via Docker" section.

6) There is some quite nice functional validation work done on some of the DE transcripts that would have been hidden in a gene-level analysis. There is also some nice work on detecting HBV fusion genes. These both contain important results which are not mentioned at all in the abstract. I feel like the abstract as it stands is selling the paper short.

Our reply; Thank you very much for this important comment. We added the following sentences to the abstract; “Comparison of expression levels identified 9,933 differentially expressed transcripts (DETs) in 4,744 genes. Interestingly, 746 genes with DETs, including LINE1-MET transcript, were not found by the gene-level analysis. We also found that fusion transcripts of transposable elements and hepatitis B virus (HBV) were overexpressed in HCCs. In vitro experiments on DETs showed that LINE1-MET and HBV-human transposable elements promoted cell growth.”.

7) Fig 5C shows a Venn diagram of fusions detected by short-read vs long-read sequencing, in which there is quite low overlap between these. You make the statement in the paper that "a combination of short- and long-reads can detect more fusion genes". I find it more likely that the short-read ICGC data had much greater depth of coverage than the MinION data you produced, which allowed for the detection of fusions that were expressed at much lower levels. This could be easily tested by downsampling the ICGC data to the same amount of sequence data as was generated on the MinION, and re-creating the Venn diagram with the fusions detected that way.

Our reply; Thank you very much for this very important comment. We compared the amount of data between our long-reads and the previous short-reads. However, the amounts of data were not quite different (Supplemental Fig. S14A). Therefore, differences in depth are not likely to be the cause of the low overlap. We considered that two possibilities could explain the low overlap. First, most of the fusion genes missed by short-read were very low expression levels, less than 1 reads per million reads (RPM) (Supplemental Fig. S14B), therefore, there are many fusion-genes with low expression levels, and they are difficult to be detected. Second, 28.9 % of transcripts in long-reads lacked 5' region (Supplemental Fig. S5 and Supplemental Fig. S14C,D). Therefore fusion-genes whose breakpoints are located in the 5' region were difficult to detect by long-read.

We added the following sentences to the "Fusion genes" subsection in the Results (line 400-405); “We considered that two possibilities could explain the low overlap. Since the most of the fusion genes missed by short-reads had very low expression levels (Supplemental Fig. S14B), many fusion-genes with low expression levels would be missed by a single approach. In addition, 28.9 % of transcripts in long-reads lacked 5' region (Supplemental Fig. S5 and Supplemental Fig. S14C, D). Therefore fusion-genes whose breakpoints are located in the 5' region would be difficult to detect by long-read.”. We also added a figure on the amount of data to Supplemental Information (Supplemental Fig. S14A).

8) Figure 5D is very interesting. What do you conclude from that result? Please comment in the manuscript.

Our reply; Thank you very much for this important comment. We used samples that used for whole-genome sequencing in our previous study. Therefore, a list of SVs is available. We classified fusion-gene to these supported by SVs (SV detected fusion-genes) and others (no SV detected fusion-genes), and compared the expression levels of them (Figure 5D).

Whole-genome sequencing can accurately identify clonal (high frequency) SVs, however, would miss sub-clonal (low frequency) SVs. Therefore, we considered that no SV detected fusion-genes were generated by sub-clonal SVs. This result suggests that there are a lot of sub-clonal fusion genes, and their expression levels are lower than clonal fusion genes. Although the functional importance of sub-clonal fusion genes is currently unknown, deeper RNA sequencing would detect a larger number of fusion genes.

          We added the following sentences to the “Fusion genes” subsection in the Results (line 410-412); “This result suggests that there are a lot of sub-clonal fusion genes, and their expression levels are lower than clonal fusion genes. Although the functional importance of sub-clonal fusion genes is currently unknown, deeper RNA sequencing would detect a larger number of fusion genes.”.

*Minor comments:

1) The manuscript has many small errors in English grammar, spelling and style. I would strongly recommend sending it for copy editing before submitting it to a journal.*

Our reply; Thank you very much for this comment. Due to the limitation of time, the current version has not been proofread by a native-English speaker. We are planning to review English grammar by a native-English speaker.

2) Neither the results section nor the methods section describing the sequencing that was performed specify whether it was done on a MinION or PromethION (or flongle). While this is implied elsewhere in the paper, it should definitely be specified in the methods at a minimum.

Our reply; Thank you for this comment. We used a MinION for sequencing. We added the following sentences to the Method section (line 579-580); “Libraries were sequenced on a SpotON FlowCell MKⅠ(R9.4) (Oxford Nanopore), using the MinION sequencer (Oxford Nanopore)”.

3) You also write in the introduction that your method, SPLICE, was developed for the MinION specifically. Please comment on its applicability to data generated on the PromethION and flongle Nanopore sequencers.

Our reply; Thank you very much for this comment. We consider that our method is applicable to data from MinION, PromethION, and flongle. We added the following sentence to the Methods section (line 592-593); “In the present study, we analyzed sequence data from MinION. We consider that our method is applicable to data from MinION, PromethION, and flongle.”.

4) The volcano plot in Fig 3A is missing its dots.

Our reply; Thank you very much for this comment. We modified the Fig. 3A.

*Reviewer #3: General comments:

Summary: In this manuscript, Kiyose et al have developed and tested a novel methodology for identifying splicing alterations, and fusions, from full-length transcript or long read sequencing data. They apply this approach to liver cancer and paired, non-cancerous liver tissue from a prior publication, and use wet-lab/experimental methods to validate their in silico findings. They conclude that their new methodology, SPLICE, outperforms one existing method, and is uniquely suitable to identifying fusion genes.*

Major Comments: 1) Figure 1B shows a schematic of common error patterns from MinION cDNA sequencing, and the text of the manuscript describes how the authors' new approach (SPLICE), overcomes several of these, e.g. sequencing errors, artificial chimeras, and mapping errors of highly homologous genes. However, there is a fundamental disconnect between the text and the graphic in Figure 1B. This should either be revised for clarity, or an additional graphic or flowchart placed in the supplementary materials to clearly show *how* SPLICE overcomes each of these limitations.

Our reply; We apologize for the insufficient explanation in Figure 1. We showed a detailed explanation of the data analysis procedure in Supplemental Fig. S1.

2) Why was TALON the only alternative approach chosen for validation of SPLICE performance? There are a number of other, more advanced pipelines such as SUPPA2, and IsoformSwitchAnalyzeR. It would strengthen the manuscript, and its conclusions, to incorporate at least one of these methods as a second comparator. This is particularly true for IsoformSwitchAnalyzeR, since Kiyose et al identify a number of differentially expressed transcripts (DETs) for genes that are not differentially expressed.

Our reply; Thank you very much for this important comment. Another reviewer also requested additional benchmarking, therefore we performed an additional performance comparison for the revised manuscript. As SUPPA2 and IsoformSwichAnalyzeR are used to analyze the annotated output GTF files, and direct comparison with SPLICE is difficult. Since IsoformSwichAnalyzeR recommends StringTie as an annotation soft, we compared using StringTie instead.

We compared the performance of SPLICE with that of four other methods (TALON, FLAIR, StringTie and Bambu) for splicing variant detection. SPLICE identified the third-highest number of transcripts followed by FLAIR and StringTie (Supplemental Fig. S3A). In MCF-7 the concordance rate with IsoSeq MCF-7 transcriptome data was the highest in SPLICE for known transcripts and the second highest in SPLICE for novel transcripts (Supplemental Fig. S3B).

We added the text to the "Comparison of SPLICE method with other tools" subsection of the Results (line 165-177) and the "Benchmarking" subsection of the Methods (line 640-665). We added the results to Supplemental Fig. 3.

3) The Venn diagram in Figure 5C appears to show that conventional short read sequencing identifies 46 fusion genes that are not also detected by long read sequencing. However, this result, and its implications are never addressed in the text.

Our reply; Thank you very much for this important comment. We apologize for the insufficient explanation. We considered that two possibilities could explain the low overlap. First, most of the fusion genes missed by short-read were very low expression levels, less than 1 reads per million reads (RPM) (Supplemental Fig. S14B), therefore these are many fusion-gene with low expression level and they are difficult to be detected. Second, 28.9 % of transcripts in long-reads lacked 5' region (Supplemental Fig. S5 and Supplemental Fig. S14C,D). Therefore fusion-genes whose breakpoints are located in the 5' region were difficult to detect by long-read.

          We added the following sentences to the "Fusion genes" subsection in the Results (line 400-405); “We considered that two possibilities could explain the low overlap. The most of the fusion genes missed by short-reads had very low expression levels (Supplemental Fig. S14B). This result suggests that there are many missed fusion-genes with low expression levels. In addition, 28.9 % of transcripts in long-reads lacked 5' region (Supplemental Fig. S5 and Supplemental Fig. S14C, D). Therefore fusion-genes whose breakpoints are located in the 5' region would be difficult to detect by long-read.”. We also added a figure on the amount of data to Supplemental Information (Supplemental Fig. S14A).

Minor Comments: 1) On pages 20-21, the language used to describe the HBV and/or HCV postive vs negative materials is very confusing. Please clarify that by "HBV- and HCV-related tissues" you in fact mean "HBV-and HCV-infected samples."

Our reply; We apologize for the confusing wording. We converted "HBV and HCV-related tissues" to " HBV and HCV-infected samples" in the manuscript.

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

Learn more at Review Commons

---

Referee #3

Evidence, reproducibility and clarity

Summary:

In this manuscript, Kiyose et al have developed and tested a novel methodology for identifying splicing alterations, and fusions, from full-length transcript or long read sequencing data. They apply this approach to liver cancer and paired, non-cancerous liver tissue from a prior publication, and use wet-lab/experimental methods to validate their in silico findings. They conclude that their new methodology, SPLICE, outperforms one existing method, and is uniquely suitable to identifying fusion genes.

Major Comments:

1. Figure 1B shows a schematic of common error patterns from MinION cDNA sequencing, and the text of the manuscript describes how the authors' new approach (SPLICE), overcomes several of these, e.g. sequencing errors, artificial chimeras, and mapping errors of highly homologous genes. However, there is a fundamental disconnect between the text and the graphic in Figure 1B. This should either be revised for clarity, or an additional graphic or flowchart placed in the supplementary materials to clearly show how SPLICE overcomes each of these limitations.
2. Why was TALON the only alternative approach chosen for validation of SPLICE performance? There are a number of other, more advanced pipelines such as SUPPA2, and IsoformSwitchAnalyzeR. It would strengthen the manuscript, and its conclusions, to incorporate at least one of these methods as a second comparator. This is particularly true for IsoformSwitchAnalyzeR, since Kiyose et al identify a number of differentially expressed transcripts (DETs) for genes that are not differentially expressed.
3. The Venn diagram in Figure 5C appears to show that conventional short read sequencing identifies 46 fusion genes that are not also detected by long read sequencing. However, this result, and its implications are never addressed in the text.

Minor Comments:

1. On pages 20-21, the language used to describe the HBV and/or HCV postive vs negative materials is very confusing. Please clarify that by "HBV- and HCV-related tissues" you in fact mean "HBV-and HCV-infected samples."

Significance

There is somewhat strong significance to this advance. As promising as long read, full-transcript sequencing is for the field, current limitations such as its high error rate have limited applicability, and most of the current analytic pipelines require complementary short read RNA sequencing to be performed in parallel for error correction. The authors assert that SPLICE overcomes these limitations, and to some extent demonstrates this. As a predominantly wet-lab experimentalist in the area of RNA processing, I have the relevant expertise to most rigorously assess the downstream impacts of findings from pipelines such as SPLICE, e.g. the validation experiments shown in the latter portion of the manuscript. These are uniformly strong. Where I was challenged some is in the authors' explanations of how and why SPLICE's specific design, as an algorithm, overcomes the known limitations in current analytic pipelines for long-read sequencing.

Referees cross-commenting

I concur with Reviewer 2. I think the 3 of us were broadly enthusiastic, yet raised some of the same concerns. In my view, these concerns should be able to be readily addressed by the authors.

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

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

Summary:

This is both a presentation of a pipeline for analysis of Nanopore RNA-seq data, as well as an analysis of a cohort of 44 hepatocellular carcinomas against matched-normal liver tissue. It presents a number of quite intriguing results from the long-read RNA analysis, and suggests potential new targets for study in HCC. It is also worth noting that the current version of guppy (6) has functionality to detect primer sequences in the middle of reads and split those reads, which may obviate one of the steps in SPLICE.

Major comments:

1. The work done in this study used data that was basecalled using guppy 3.0.3. Since that version, I am aware of at least two major upgrades to the base caller accuracy, which would likely also improve the accuracy of isoform resolution. Given that the data is relatively low-coverage and that you have an automated workflow for the analysis, I would recommend re-basecalling using an updated basecaller and re-running your analysis using that. This is especially important given your comments in the paper about splice site misalignment.
2. You have compared your software to another tool for isoform analysis on Nanopore sequencing data, TALON. But a number of other tools exist for this purpose, including stringtie2, flair and bambu. My own testing has shown that stringtie2 outperforms TALON in terms of concordance with Illumina RNA-seq. It is quite important that you perform a complete comparison of your software to the state of the art for this purpose.
3. Likewise, for fusion detection, you compare to LongGF. You should also compare to (and cite) JAFFAL.
4. In terms of the source code, I have questions. Why did you use BASH to run the Python code, instead of making this into a Python package? Why did you not use the functionality already available in BioPython for a number of basic sequence data handling tasks? Why is there not even a single function defined anywhere, let alone classes?

At some level, if it works, it works. But I have serious concerns about the long-term maintainability of the code in its current state. 5. Also related to the code, it is generally the standard now to create a BioConda package or Docker container for a bioinformatics package. BioConda has the advantage that the BioContainers project automatically generate Docker and Singularity containers from it. Please provide one of these. 6. There is some quite nice functional validation work done on some of the DE transcripts that would have been hidden in a gene-level analysis. There is also some nice work on detecting HBV fusion genes. These both contain important results which are not mentioned at all in the abstract. I feel like the abstract as it stands is selling the paper short. 7. Fig 5C shows a Venn diagram of fusions detected by short-read vs long-read sequencing, in which there is quite low overlap between these. You make the statement in the paper that "a combination of short- and long-reads can detect more fusion genes". I find it more likely that the short-read ICGC data had much greater depth of coverage than the MinION data you produced, which allowed for the detection of fusions that were expressed at much lower levels. This could be easily tested by downsampling the ICGC data to the same amount of sequence data as was generated on the MinION, and re-creating the Venn diagram with the fusions detected that way. 8. Figure 5D is very interesting. What do you conclude from that result? Please comment in the manuscript.

Minor comments:

1. The manuscript has many small errors in English grammar, spelling and style. I would strongly recommend sending it for copy editing before submitting it to a journal.
2. Neither the results section nor the methods section describing the sequencing that was performed specify whether it was done on a MinION or PromethION (or flongle). While this is implied elsewhere in the paper, it should definitely be specified in the methods at a minimum.
3. You also write in the introduction that your method, SPLICE, was developed for the MinION specifically. Please comment on its applicability to data generated on the PromethION and flongle Nanopore sequencers.
4. The volcano plot in Fig 3A is missing its dots.

Significance

Nature and significance of the advance: The paper presents several exciting advances in terms of tumour biology. The authors demonstrate how alternative splicing can drive liver cancer, while being undetectable by short-read sequencing. They also show a large number of fusion transcripts that were validated by RT-PCR but were undetectable with short-read sequencing. The analysis method they present, SPLICE, contains a number of smaller advances, but raises major concerns about its capacity to act as a maintainable piece of bioinformatics software.

Comparison to existing published knowledge: The authors compare the software they present to a single tool in the same class for the two functions it performs (isoform analysis and fusion detection). A more thorough comparison to a broader range of available tools would be better.

In terms of biology, the authors extensively cite related literature to place their discoveries in context.

Audience: Cancer researchers, anyone interested in doing isoform-level differential expression analysis or gene fusion detection using Nanopore RNA-seq data.

My expertise: I am a staff scientist working on developing and testing tools for Nanopore sequencing analysis at a cancer research centre.

Referees cross-commenting

I fully agree with all of the comments by the other two reviewers.

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

Fujimoto and collaborators use Nanopore-based cDNA sequencing for genome-wide transcriptome analysis of a collection of hepatocellular carcinomas (HCCs) and matched normal liver tissues. To improve detection of alternatively spliced isoforms and hybrid transcripts potentially deriving from genomic rearrangements, they develop a dedicated pipeline SPLICE, which they benchmark against available software used for the same analysis. Besides having dual functionality (calls both alternative transcripts and fused transcripts), SPLICE seems to outperform previous software in calling alternative/fused transcripts and accuracy. They use the SPLICE pipeline to call isoforms and gene fusions in normal liver cells and HCCs and perform basic functional validations on novel fusions identified. The manuscript is well written, and the analyses are well performed. Perhaps the benchmarking of the SPLICE pipeline could have been more extensive (i.e., performed on additional independent datasets).

Major points:

1. Line 149-150: "We compared the results of mapping to the reference genome and the reference transcriptome sequences, and removed candidates if both were inconsistent (removal of mapping errors). " Please specify what "both were inconsistent" means.
2. Concerning TE-derived novel exons, in principle, this may lead to altered expression of the TE-transcript (as the Authors report for L1-MET) or to altered splicing of the transcript (i.e., other exon/introns could be retained or excluded). Can the Authors assess whether the inclusion of the TE in a transcript enhances its expression or affects the splicing of the "parental" transcript? If so, can they verify if the position of the insertion of the TE has any effect on expression and splicing?
3. Can the Authors explain why the NBEAL1-RPL12 was not detected by SPLICE?
4. Line 332: Can the Authors explain how the total amount of HVB mRNA was determined in each sample? Is it a relative amount calculated from the sequencing data? If so, it should be made clear in the text that this is a fractional measure.
5. Fig4a: please specify if the y-axis "number of support reads" reports library normalized values.
6. HCCs have more HBV-human genome fusion transcripts than normal liver. Could the authors clarify if these HCC transcripts are selectively found in tumors? or whether they are also expressed in normal liver samples? The paragraph starting from line 356 is confusing, and it is difficult to retrieve the above information for both HBs and HBx fusions.
7. Figure 4C: what was the control used to calculate the relative viability in these analyses?
8. MYT1L: the Authors report the identification of a novel MYT1L transcript downregulated in HCC, and argue it may have a potential tumor-suppressive function. For the sake of clarity, it will be advisable to show also the differential expression (HCC vs. Liver) of the other transcripts expressed from the same locus.

Minor points:

1. Table S4: there is a typo, correct "secific" in "specific"

Significance

The Authors show that applying long-reads sequencing to the study of the transcriptome, combined with their improved in-house analyses pipeline, leads to the identification of novel transcripts, which are alternative splicing isoforms and transcripts originating from novel gene fusions with potential oncogenic function. This provides a proof of principle study which show the advantages of long-reads sequencing and offers a solid data for further mechanistic studies on liver cancer.

Referees cross-commenting

I also agree with the other reviewers. All the concerns expressed by the reviewers seem addressable in a reasonable timeframe.

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

1. General Statements [optional]

Overall we were elated to have received such positive comments on the manuscript, with requests for only minor changes. We have made all suggested changes to clarify or tone down the language as suggested.

We would like to thank each of the three reviewers for their assessment of our work. We note that all three reviewers agreed the phylogenetic analysis was interesting and convincing. Two of the three reviewers felt the study sufficiently demonstrated roles for Baramicin in the nervous system. We have responded to comments from Reviewer 2 to draw attention to some aspects of the data that they may have been overlooked, which we hope reassures them that our proposal of BaraB and BaraC involvement in the nervous system is robust, coming from different approaches that show consistent results.

Reviewer 1 and Reviewer 3 compliment the study as being very worthwhile, and for suggesting concrete routes for how an AMP evolved non-immune functions. Both compliment its comprehensiveness, and describe the study as having striking findings that should have broad appeal to audiences interested in the crosstalk between the nervous system and the innate immune system.

2. Point-by-point description of the revisions

In the revised manuscript file, we have highlighted all text where changes were made.

---

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

The authors provide convincing evidence for an evolutionary scenario in which duplications of an AMP gene with ancestral immune function led to paralogs specialist for neural functions. They focus on the Baramicin genes, coding for major Toll signalling targets in the context of antifungal defence. Their study uses infection experiments in several Drosophila species, a careful annotation of the Baramicin genes of D. melanogaster, the demonstration of neural expression of BaraB and BaraC, the KD analysis of Bara B revealing lethality and neurological phenotypes, a reconstruction of the evolutionary history of Baramicn genes in Drosophilids and an analysis of the sequence evolution of the IM24 domain providing the neural functions. In general the paper is well written. There are a few places in the manuscript where the language can be improved and one point, which needs clarification: - ine 297: ...,which did not present with... - line 314/315: ...to just 14% that of...to 63% that of - line 459: ..., we this motif... - line 518: What does "... genomic relatedness (by speciation and locus)..." mean? - line 527/528: ...drive behaviour or disease through interactions... - line 532: ... ancestrally encodes distinct peptides involved with either the nervous system or the immune response... line 535: ...with either the nervous system (IM24) or.... Do the data provide enough evidence suggesting that IM24 had a neural function in the ancestor? Ideally the authors should look at neural expression of the Baramicin gene in the ourgroup, S. lebanonensis. The authors later (line571) admit, that they cannot rule out that IM24 is also antimicrobial.

We thank reviewer #1 for drawing attention to these points. We have made changes to each line to be more concise, clarify our meaning, or fix typos.

Reviewer #1 (Significance (Required)):

This is a very comprehensive study, which, to my knowledge for the first time, suggests concrete routes of how an AMP evolved non-immune functions. One of the striking findings of this paper is that duplications and subsequent truncations of the ancestral Baramicin locus linked to specialisation for neural functions occurred independently in different Drosophila lineages.

We thank reviewer #1 for their very positive comments. We also agree with all suggested changes, including more careful phrasing to emphasize that we have not described a mechanism, just an involvement in the nervous system. For instance, see lines 556-568 are reworked to soften language and explicitly state the ancestral function of IM24 is unknown, and our suggestion that IM24 could underlie Dmel\BaraA interactions with the nervous system is speculation that should be tested.

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

Hanson and Lemaitre present a genomic and phylogenetic characterization of the Baramicin family of antimicrobial peptide genes in different species. They discover new Baramicin paralogs, united by the presence of an IM24 domain at the N-terminus. They show that among Baramicins, those that are not inducible by infection (which they improperly call non-immune since a protein can be non-inducible by infection and have very important immune functions), are truncated. They propose that an ancestor peptide with immune functions evolved into a neuronal regulator/effector via truncation.

Although the hypothesis is interesting, the data do not really support it. This manuscript is rather descriptive at this point. The demonstration that IM24 is necessary for neural function is very tenuous. For example, in the paragraphs titled Dmel\BaraB is required in the nervous system during development and Baramicin B plays an important role in the nervous system, I did not find convincing data demonstrating that BaraB is required in the nervous system. The only data that links BaraB to the nervous system is a weak locomotion defect observed in the BaraB mutant. But how many genes, when inactivated, give a locomotion defect? This remains totally unexplained at the molecular level. The authors also mentioned that BaraB is expressed in a subset of mechanosensory neuron cells in the wing. What is the link between this expression and the nubbin phenotype? The authors also mention that data in the literature indicate that BaraC is expressed in glial cells but also in other tissues. Finally, we have no idea what role, if any, these peptides have in the nervous system.

While the characterization of the Baramicin gene family and its evolution across species is convincing, the link between these AMPs and the nervous system is really too preliminary to be convincing. The manuscript would greatly benefit from being more concise.

Reviewer #2 (Significance (Required)):

see above

We thank reviewer #2 for their fair assessment. We have made edits to soften our phrasing, and to emphasize that we have not described a mechanism, just an involvement, in the nervous system.

Examples:

line 270: “integral development role” -> “important for development”

line 277: “Baramicin B plays an important role in the nervous system“ -> “Baramicin B suppression in the nervous system mimics mutant phenotypes”

line 532: “Here we demonstrate that the Baramicin antimicrobial peptide gene of Drosophila ancestrally encodes distinct peptides involved with either the nervous system or the immune response.“ -> “Here we demonstrate that the Baramicin antimicrobial peptide gene of Drosophila ancestrally encodes distinct peptides that may interact with either the nervous system (IM24) or invading pathogens (IM10-like, IM22).”

line 562 new text: “Thus while our results suggest that IM24 of different Baramicin genes might underlie Baramicin interactions with the nervous system, we cannot exclude the possibility that IM24 is also antimicrobial, or even that antimicrobial activity is IM24’s ancestral purpose. Future studies could use tagged IM24 transgenes or synthetic peptides to determine the host binding partner(s) of secreted IM24 from the immune-induced Dmel\BaraA, and/or to see if IM24 binds to microbial membranes.”

We have also changed all instances of “non-immune Baramicins” to “Baramicins lacking immune induction” or something to that effect (e.g. new Lines 25,464, 469,478-82).

We also made some small changes to be more concise (e.g. line 387, 447, cut lines 492-495 from previous version, cut lines 506-507 from previous version).

We have responded below in the reviewer-to-reviewer comments for a few of the specific points raised there, which we hope further assuage some of Reviewer 2’s concerns.

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

Antimicrobial peptides are main effectors in (insect) immune defenses. It is becoming more and more clear, that AMPs can have pleiotropic effects or even acquire new functions. In the present paper, the authors investigate Baramicin, an antifungal AMP that they described first in publication last year. Here they show that in Drosophila melanogaster Baramicin A, which they described before, has paralogs, that are not immune-inducible. They then show that these paralogs, named BarB and BarC, which are truncated versions of BarA, are expressed in the head and neural tissues. That they have neural functions is supported by targeted gene-silencing experiments. They go on to show, using a comparative approach across Drosophila, that Baramicin A with its antimicrobial function constitutes the ancestral state. Moreover, Baramicin is also enriched in head samples of some of the other Drosophila species they study. This manuscript, which according to the acknowledgements has already been seen by reviewers, is in a very good shape.

I have only a number of minor points, that might help to clarify the presentation.

Lines 34-36: I would delete this sentence and replace it with a statement based on the main findings of the manuscript

We now conclude the abstract with “As many AMP genes encode polypeptides, a full understanding of how immune effectors interact with the nervous system will require consideration of all their peptide products.”

Lines 56-60. May be tone down a bit. Anti-inflammatory activities of AMPs have been known for a long time. I think the next paragraph makes a very good case what is already known and is hence a nice motivation for the current study.

Toned down. This part now reads: “However AMPs and AMP-like genes in many species have recently been implicated in non-immune roles in flies, nematodes, and humans, suggesting non-immune functions might help explain AMP evolutionary patterns.”

Line 125: classical instead of classically

done

Line 200: what is a 'novel' time course? I would just describe what has been done.

Now reads: “We next measured Baramicin expression over development from egg to adult.”

Line 268: hypomorph, I guess in the literature usually hypomorphic is used.

done

Line 279: I would suggest to tone this headline down. This is not a criticism of the paper, but the actual mechanisms of the roles in the nervous system are not studied here.

Done. Now reads: “Baramicin B suppression in the nervous system mimics mutant phenotypes”

Line 505: what does not really become clear is whether IM24 plays an important role in the nervous system of fly species that only have BarA.

Edits from lines 556-568 now help highlight this question.

Line 540-549. This comparison I find a bit far-fetched, or maybe it needs clarification how doublesex expression is related to Baramicins.

Being completely honest: the doublesex discussion was requested during previous review at another journal. We agree that it is a bit of a tangent, and so we have removed these sentences.

Line 584-585. I think that this has been known for much longer from studies in frogs and beetles.

Our use of “in vivo” might have been a bit squishy here. We have edited this to reflect endogenous loss-of-function study, rather than simply “in vivo,” to clarify our intended sentiment.

Reviewer #3 (Significance (Required)):

Overall, I think that this is a very worthwhile and convincing story about the evolution AMPs and how they can acquire new functions. All the main statements are supported by careful experiments and data analysis. The paper does not go into any detail, of how the neurological role of BarB and BarC is achieved, but I think this is beyond the scope of the current manuscript. In short, this is a very worthwhile contribution to the growing literature of the role of AMPs in the nervous system. The authors provide the context of the main published papers in the area in the introduction. As opposed to most papers on this so far, the current manuscript also provides very interesting data on the evolutionary history of the Baramicin genes, both within the main study species, and within other Drosophila species. This paper should appeal to a rather broad audience of researchers interested in innate defenses, AMPs and the crosstalk between the nervous system and the innate immune system.

My background is insect immunology with a focus on AMPs and evolutionary approach.

We thank reviewer #3 for their very positive comments. We agree with all suggested changes.

**Referees cross-commenting**

This session contains the comments of all reviewers

Reviewer 3

Reviewer 2 and I share the view, that the evidence for the effects of BarB and C on the nervous system is rather limited. But I still think, that the paper provides enough new and interesting data that make it a very useful contribution. Though not a neurobiologist, I would assume that providing functional evidence for the role of BarA and B in the nervous system would justify a paper on its own. I agree though, that the relevant sections should be toned down.

Reviewer 2

As I mentioned in my review, I found the genomic and phylogenetic analysis interesting and convincing. I therefore totally agréé with reviewers 2 and 3 on that. Whether BarA and B are playing a role in the nervous system and how it does remain speculative. BaraB mutants show locomotion defects. But mutants in mitochondrial genes have locomotion defects. Can we conclude that mitochondria play a role in the nervous system? If I understand correctly, downregulating Bara in neurons only (With Elav-Gal4 driver) does not show the locomotion phenotype. it induces early lethality. How many genes when inactivated in neurons will give rise to such a phenotype? A lot. I really think that the implication of Bara in the nervous system should be seriously toned done and more presented as an hypothesis than a validated fact.

We would like to note for Reviewer 2 here that it is specifically elav> BaraB-IR that results in lethality, and in weaker gene silencing experiments, adult elav>BaraB-IR flies emerge, and they do suffer locomotor defects. Often, they got stuck in the food shortly after emerging, or would move haphazardly (which was common in flies with nubbin-like wings). We have added explicit mention that elav>BaraB-IR also results in locomotor defects (Line 288-289).

Our private speculation is that the reason flies fail to emerge from their pupae is because they are so uncoordinated that they sometimes cannot wriggle out of the pupal case before their cuticle hardens. In some instances, both using mutants and RNAi, we observed fully developed adults with mature abdominal pigmentation that died trapped inside their pupal cases.

We’d also like to emphasize here that despite testing many other Gal4 drivers, including mef2-Gal4 (muscle/myocytes), nubbin-like wings and lethality were only found using elav-Gal4. A role interacting with mitochondria would likely have been revealed using mef2-Gal4, given the importance of mitochondrial function in muscle.

For BaraC: expression in other tissues (like the rectal pad) could nevertheless be from e.g. nerves innervating the muscles controlling the sphincter. Or it could indeed be entirely unrelated to the nervous system. However we feel the nearly perfect overlap with Repo-expressing cells is a strong argument for a neural role. We also made an effort using RNAi to validate this pattern suggested by scRNAseq, which confirmed a strong knockdown of BaraC-IR with Repo-Gal4 (Fig. 3, Fig. S4).

We hope these comments clarify for Reviewer 2 why we feel confident in proposing a role for Baramicins in the nervous system, even if we do not investigate a mechanism in this study.

Reviewer 1

I agree with reviewer 3 that the main message of the paper providing a concrete scenario of how non-immune functions of AMPs may evolve is an important contribution. A deep investigation of the neural function is definitely going beyond the scope of the paper. Indeed this might be quite tricky. But it would help if the authors could clarify their idea about the ancestral condition. Is there the possibility that IM24 had ancestrally already non-immune function? They are not really clear about this point.

Reviewer 2

I agree with the other reviewers that determining the exact role of Bara peptides could be complicated. I just ask that the authors limit themselves to proposing that the peptides have lost their immune function. I stress that this argument is not very strong. It relies solely on the lack of inducibility of these peptides following infection. I still think that the demonstration of the role of Bara in the nervous system is not provided.

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

Antimicrobial peptides are main effectors in (insect) immune defenses. It is becoming more and more clear, that AMPs can have pleiotropic effects or even acquire new functions. In the present paper, the authors investigate Baramicin, an antifungal AMP that they described first in publication last year. Here they show that in Drosophila melanogaster Baramicin A, which they described before, has paralogs, that are not immune-inducible. They then show that these paralogs, named BarB and BarC, which are truncated versions of BarA, are expressed in the head and neural tissues. That they have neural functions is supported by targeted gene-silencing experiments. They go on to show, using a comparative approach across Drosophila, that Baramicin A with its antimicrobial function constitutes the ancestral state. Moreover, Baramicin is also enriched in head samples of some of the other Drosophila species they study. This manuscript, which according to the acknowledgements has already been seen by reviewers, is in a very good shape.

I have only a number of minor points, that might help to clarify the presentation.

Lines 34-36: I would delete this sentence and replace it with a statement based on the main findings of the manuscript

Lines 56-60. May be tone down a bit. Anti-inflammatory activities of AMPs have been known for a long time. I think the next paragraph makes a very good case what is already known and is hence a nice motivation for the current study.

Line 125: classical instead of classically

Line 200: what is a 'novel' time course? I would just describe what has been done.

Line 268: hypomorph, I guess in the literature usually hypomorphic is used.

Line 279: I would suggest to tone this headline down. This is not a criticism of the paper, but the actual mechanisms of the roles in the nervous system are not studied here.

Line 505: what does not really become clear is whether IM24 plays an important role in the nervous system of fly species that only have BarA.

Line 540-549. This comparison I find a bit far-fetched, or maybe it needs clarification how doublesex expression is related to Baramicins.

Line 584-585. I think that this has been known for much longer from studies in frogs and beetles.

Significance

Overall, I think that this is a very worthwhile and convincing story about the evolution AMPs and how they can acquire new functions. All the main statements are supported by careful experiments and data analysis. The paper does not go into any detail, of how the neurological role of BarB and BarC is achieved, but I think this is beyond the scope of the current manuscript.

In short, this is a very worthwhile contribution to the growing literature of the role of AMPs in the nervous system. The authors provide the context of the main published papers in the area in the introduction. As opposed to most papers on this so far, the current manuscript also provides very interesting data on the evolutionary history of the Baramicin genes, both within the main study species, and within other Drosophila species.

This paper should appeal to a rather broad audience of researchers interested in innate defenses, AMPs and the crosstalk between the nervous system and the innate immune system.

My background is insect immunology with a focus on AMPs and evolutionary approach.

Referees cross-commenting

This session contains the comments of all reviewers

Reviewer 3

Reviewer 2 and I share the view, that the evidence for the effects of BarB and C on the nervous system is rather limited. But I still think, that the paper provides enough new and interesting data that make it a very useful contribution. Though not a neurobiologist, I would assume that providing functional evidence for the role of BarA and B in the nervous system would justify a paper on its own. I agree though, that the relevant sections should be toned down.

Reviewer 2

As I mentioned in my review, I found the genomic and phylogenetic analysis interesting and convincing. I therefore totally agréé with reviewers 2 and 3 on that. Whether BarA and B are playing a role in the nervous system and how it does remain speculative. BaraB mutants show locomotion defects. But mutants in mitochondrial genes have locomotion defects. Can we conclude that mitochondria play a role in the nervous system? If I understand correctly, downregulating Bara in neurons only (With Elav-Gal4 driver) does not show the locomotion phenotype. it induces early lethality. How many genes when inactivated in neurons will give rise to such a phenotype? A lot. I really think that the implication of Bara in the nervous system should be seriously toned done and more presented as an hypothesis than a validated fact.

Reviewer 1

I agree with reviewer 3 that the main message of the paper providing a concrete scenario of how non-immune functions of AMPs may evolve is an important contribution. A deep investigation of the neural function is definitely going beyond the scope of the paper. Indeed this might be quite tricky. But it would help if the authors could clarify their idea about the ancestral condition. Is there the possibility that IM24 had ancestrally already non-immune function? They are not really clear about this point.

Reviewer 2

I agree with the other reviewers that determining the exact role of Bara peptides could be complicated. I just ask that the authors limit themselves to proposing that the peptides have lost their immune function. I stress that this argument is not very strong. It relies solely on the lack of inducibility of these peptides following infection. I still think that the demonstration of the role of Bara in the nervous system is not provided.

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

Hanson and Lemaitre present a genomic and phylogenetic characterization of the Baramicin family of antimicrobial peptide genes in different species. They discover new Baramicin paralogs, united by the presence of an IM24 domain at the N-terminus. They show that among Baramicins, those that are not inducible by infection (which they improperly call non-immune since a protein can be non-inducible by infection and have very important immune functions), are truncated. They propose that an ancestor peptide with immune functions evolved into a neuronal regulator/effector via truncation.

Although the hypothesis is interesting, the data do not really support it. This manuscript is rather descriptive at this point. The demonstration that IM24 is necessary for neural function is very tenuous. For example, in the paragraphs titled Dmel\BaraB is required in the nervous system during development and Baramicin B plays an important role in the nervous system, I did not find convincing data demonstrating that BaraB is required in the nervous system. The only data that links BaraB to the nervous system is a weak locomotion defect observed in the BaraB mutant. But how many genes, when inactivated, give a locomotion defect? This remains totally unexplained at the molecular level. The authors also mentioned that BaraB is expressed in a subset of mechanosensory neuron cells in the wing. What is the link between this expression and the nubbin phenotype?

The authors also mention that data in the literature indicate that BaraC is expressed in glial cells but also in other tissues.

Finally, we have no idea what role, if any, these peptides have in the nervous system.

While the characterization of the Baramicin gene family and its evolution across species is convincing, the link between these AMPs and the nervous system is really too preliminary to be convincing. The manuscript would greatly benefit from being more concise.

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

The authors provide convincing evidence for an evolutionary scenario in which duplications of an AMP gene with ancestral immune function led to paralogs specialist for neural functions. They focus on the Baramicin genes, coding for major Toll signalling targets in the context of antifungal defence. Their study uses infection experiments in several Drosophila species, a careful annotation of the Baramicin genes of D. melanogaster, the demonstration of neural expression of BaraB and BaraC, the KD analysis of Bara B revealing lethality and neurological phenotypes, a reconstruction of the evolutionary history of Baramicn genes in Drosophilids and an analysis of the sequence evolution of the IM24 domain providing the neural functions. In general the paper is well written. There are a few places in the manuscript where the language can be improved and one point, which needs clarification:

• line 297: ...,which did not present with...
• line 314/315: ...to just 14% that of...to 63% that of
• line 459: ..., we this motif...
• line 518: What does "... genomic relatedness (by speciation and locus)..." mean?
• line 527/528: ...drive behaviour or disease through interactions...
• line 532: ... ancestrally encodes distinct peptides involved with either the nervous system or the immune response... line 535: ...with either the nervous system (IM24) or.... Do the data provide enough evidence suggesting that IM24 had a neural function in the ancestor? Ideally the authors should look at neural expression of the Baramicin gene in the ourgroup, S. lebanonensis. The authors later (line571) admit, that they cannot rule out that IM24 is also antimicrobial.

Significance

This is a very comprehensive study, which, to my knowledge for the first time, suggests concrete routes of how an AMP evolved non-immune functions.<br /> One of the striking findings of this paper is that duplications and subsequent truncations of the ancestral Baramicin locus linked to specialisation for neural functions occurred independently in different Drosophila lineages.

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

We thank the reviewers for carefully reading our manuscript. We found their comments to be incredibly thoughtful and constructive and greatly appreciate their feedback. We are confident that addressing the reviewers’ concerns has strengthened our manuscript.

Reviewer #1 (Evidence, reproducibility and clarity (Required)): Camuglia, Chanet and Martin investigate the mechanisms that control cell division orientation in vivo, using the mitotic domains (MDs) in the head of the Drosophila embryo as their main model system. They find that cells in the head mitotic domains rotate and align their spindles within 30 degress of the anterior-posterior axis of the embryo. The Pins protein, implicated in spindle orientation in other systems, is planar polarized in mitotic cells. Pins polarization precedes spindle rotation and is correlated with the division angle (but cell shape is not, violating Hertwig's rule). Overexpression of myristoylated Pins results in uniform Pins distribution on the membrane and affects spindle orientation. alpha-catenin RNAi (but not canoe RNAi) disrupts Pins polarity and spindle orientation in MDs 1, 3 and 5. Low dose CytoD injections (which should disrupt force transmission) also result in defective Pins polarity and spindle orientations. Finally, mechanical isolation by laser ablation also disrupts spindle orienttion. The authors find that preventing mesoderm invagination by snail dsRNA disrupts Pins polarity and spindle orientation in the head. MAJOR 1. Is there a certain chirality in the rotation of the spindles? From Movie 1, it seems like in MDs 1 and 3 at least, a majority of spindles on the right side of the embryo rotate clockwise, while spindles on the left side rotate counter-clockwise? Is that so, and in that case, are there geometric/molecular considerations that could explain that chirality?

We thank the reviewer for pointing this out. They are correct in that there is a tilt to the spindle orientation relative to the AP axis. To illustrate this tilt, we performed our spindle analysis separately on the right and left sides of MD1 and found that spindles on the left side align with an average division angle of about 30 from the AP axis whereas spindles on the right side align with an average division angle of -30 from the AP axis. To determine whether spindles on either side rotated with a certain chirality, we found there was no preference in rotating clockwise or counterclockwise on the left and right sides (on the left side of MD1 53% of measured spindles rotated counterclockwise and 47% rotated clockwise, on the right side 46% rotated counterclockwise and 54% clockwise). We have added this data as Fig. 1I-J and discussed in the Results lines 134-145.

1. The authors are experts in mesoderm invagination, and understandably concentrate on the role that forces from that process may have in the orientation of head MD divisions. However, the cephalic furrow forms much closer to the head MDs, and in an orientation that might also explain the alignment of spindles in the head. Is cephalic furrow formation important for Pins polarity and spindle orientation in the head MDs?

This was certainly a possibility, but our experimental results strongly argues that mesoderm invagination is most relevant.

1) Perturbing the ventral furrow (e.g. by Snail depletion) does not block the cephalic furrow (Vincent et al., 1997; Leptin and Grunewald, 1990), but does block mesoderm invagination. Snail depletion strikingly disrupted spindle orientation and Pins localization, which suggests mesoderm is most important.

2) In addition, depletion of -catenin blocks ventral furrow invagination but not cephalic furrow formation. We see a disruption in spindle orientation and Pins localization in -catenin RNAi, which suggests cephalic furrow itself cannot orient spindles.

3) Furthermore, light sheet imaging of the Drosophila embryo has shown that the head region of the embryo undergoes tissue movement in the direction of the cell division and that this is associated with mesoderm invagination (Streichan et al., 2018; Stern et al., 2022).

See movies here: https://www.youtube.com/watch?v=kC11Upr30JY

To further test the importance of mesoderm invagination, we will perform additional ablation experiments trying to disrupt forces transmitted to the mitotic domains from distinct directions. Once we get this experimental result we will include language in the Discussion that will summarize the experimental results and the weight of the evidence for the roles of either ventral or cephalic furrow.

1. Does expression of myristoylated Pins affect mesoderm invagination (or cephalic furrow formation)? From Table S1 it seems that a maternal Gal4 driver was used to express myristoylated Pins, which could affect other tissues in the embryo. So it is in principle possible that effects of myristoylated Pins on mesoderm internalization/cephalic furrow formation could affect cell division orientation much like sna loss of function does, but in a mechanism that does not depend on Pins polarity. There is definitely an effect on mesoderm invagination in alpha-catenin RNAi (but not in canoe RNAi) embryos, so I wonder if the effect could be consistently through defects in mesoderm invagination (or cephalic furrow formation), and Pins polarity is really dispensable for spindle orientation. Are there head-specific Gal4 drivers that could be used to drive myristoylated Pins exclusively in the head?

We apologize that we did not clarify this in the text. Maternal overexpression of myr-Pins does not obviously disrupt mesoderm internalization/cephalic furrow formation. But, we do see that targeted disruption of mesoderm internalization via a Snail depletion affects the orientation of division. Note that our paper demonstrates the effect of force transmission on Pins polarity and division orientation, which is new and the main conclusion. The role of these divisions in morphogenesis is more complicated and is beyond the scope of this study.

In response to this comment we: 1) added language in the Results that states that gastrulation proceeds in myr-Pins expressing embryos (lines 206-208), 2) Added to the Discussion of the role of these oriented divisions to morphogenesis (lines 443-449), and 3) will add a figure showing ventral furrow and cephalic furrow formation in embryos ectopically expressing the myr-Pins.

1. Related to the previous point, does mechanical isolation by laser ablation (Figure 6I-N) affect Pins polarity? This experiment could alleviate some of my concerns above, as it certainly does not (should not?) disrupt neither mesoderm invagination nor cephalic furrow formation.

We agree that it would be useful to look at Pins polarity in laser ablated embryos. Currently, we have been unable to analyze Pins polarity after laser ablation, because the ablation to fully isolate the mitotic domain has bleached our Pins::GFP signal. Also, we have shown that Pins polarity is disrupted by 1) alpha-catenin-RNAi, 2) low dose CytoD injection, and 3) Snail depletion, all of which are expected to disrupt force generation and transmission through tissues.

In response to the reviewer comment, we will determine if Pins::GFP can be analyzed in less aggressive (directional) laser ablations. Again, remember that myr-Pins does not affect mesoderm internalization and that Snail depletion affects Pins polarity.

MINOR 1. Figure S5: I am a bit confused about the role of Toll 2, 6, 8 in orienting spindle orientation. In Figure S5D it seems that dsRNA treatment against these genes does not disrupt spindle orientation, but Figure S5F shows quite a significant (p=0.0057) effect in triple mutants. The authors favor the idea that Toll receptors do not affect spindle orientation, but the difference with the mutant should be addressed. Furthermore, what happens in MDs 3, 5 and 14 (if the germband extension defect does not affect those divisions)? Is there a difference between dsRNA and triple mutant embryos in these other MDs?

We think this is a great point. We stated in the text that TLRs are not solely responsible (line 247) for spindle orientation as they do not recapitulate the random pattern of division seen in the myr-Pins expression condition. We acknowledge the differences between the dsRNA injection and TLR triple mutant in the manuscript (lines 242-247), but our data show a greater importance for the role of force transmission. We favor the idea that other mechanisms contribute to spindle orientation because of the small effect of mutating all three Tolls and the dramatic effects of depleting AJs, inhibiting actin (with CytoD), laser ablation, and blocking mesoderm invagination. The planned laser ablation experiments (described above) will also contribute to addressing this point.

1. No statistical analysis is provided for any of the differences in polarity between Pins and Gap43, and this should be done to demonstrate the significance of the polarization of Pins. Also, particularly for MD14, they should compare anterior vs. posterior polarity, as based on the images in Figure 2H it is not clear that there is a difference between the anterior and posterior side of cells.

We thank the reviewer for this point. We have added the statistical comparison.

1. Figure 2A-D: the authors propose that Pins localizes preferentially to the posterior end of cells (instead of both anterior and posterior ends) in MDs 1, 3 and 14 (and anterior in MD 5). How is the asymmetry in the distribution of Pins along the AP axis accomplished, and is there any significance to it? This should be discussed in a bit more detail (currently no potential mechanisms provided in the discussion, just an acknowledgment of the question).

__We agree the localization of Pins to the posterior end of cells in MDs 1, 3, and 14 and anterior end in MD 5 is of great interest. The details and further mechanism of this preferential localization are beyond the scope of this paper, but we have added an acknowledgment of the question and discuss possible models that could explain the result (lines 458-460). __TYPOS 1. Line 49: "one daughter cells" should be "one daughter cell". 2. Line 193: "rotation. (Figure 3E-F)." should be "rotation (Figure 3E-F)." 3. Lines 232-237: please review. 4. Line 238: "epithelia cells" should be "epithelial cells".

We thank the reviewers for carefully reading our manuscript. We have fixed the typos mentioned.

Reviewer #1 (Significance (Required)): This is the first study to my knowledge that demonstrates the role of mechanical forces in polarizing Pins, and provides a nice model to further investigate how mechanical forces generated in one tissue may affect cell division orientation in distant ones. The paper is clear, well written, and quantitative analysis is present for most results. I have some issues with the statistics (or lack thereof) for a couple of results, and potential alternative interpretations for some experiments that in my opinion should be addressed prior to publication. Specifically, it is not clear to me if Pins polarity is at all necessary for spindle orientation in any of the examined MDs.

Reviewer #2 (Evidence, reproducibility and clarity (Required)): Overview: In this manuscript, Camuglia et al. show Pins/LGN, which is understood to drive spindle orientation, can localize asymmetrically (with respect to the tissue plane) in the Drosophila embryo. Experimental work (including drug treatments, laser ablation, and knockdowns) lead the authors to propose that this asymmetry is driven by tissue-level tension. The findings are quite interesting and the manuscript is well-written overall. Major Comments: • The authors propose that localization is driven by tissue-level tension, but the direction of the tension isn't clear from the experimental work. For example, the laser ablation experiments cut around the entire perimeter of the mitotic domain, rather than along just one tension axis. Similarly, the finding that disruption of the ventral furrow (by Snail RNAi) interferes with spindle orientation in the head is very puzzling; the furrow is A) outside the embryonic head and B) runs in the parallel direction to the divisions considered. The authors need to address the directionality of tension experimentally.

We thank the reviewer for this comment and agree that better defining the direction of tension would strengthen our manuscript. We showed that blocking mesoderm invagination with Snail depletion disrupts spindle orientation, despite Snail not being required for cephalic furrow formation (refs). Recent light sheet data has shown that mesoderm invagination is associated with global movements throughout the embryo. Furthermore, the ventral furrow extends into the head region just past the anterior of MD5. To address the reviewer’s comments, we plan to: 1) Perform directional laser ablations to determine the directionality of the tension that orients the spindle, 2) Analyze strain rates in the mitotic domains prior to and during division, and 3) Add to our Discussion more about what is said in the literature about the movements that occur in the head during mesoderm invagination.

• As acknowledged in the text, the asymmetric enrichment of Pins in MD14 is fairly weak. Since the cells being examined here border a divot in the tissue, and might therefore be curving relative to the focal plane, it would be good to rule out the possibility that some of the asymmetry in Pins intensity is just a consequence of cell/tissue geometry. One way this could be achieved is by showing multiple focal planes.

Good point. We do not think that the asymmetric Pins enrichment in MD14 is due to tissue geometry or junction tilt. 1) MD14 divides ~10-15 minutes after mesoderm invagination is completed, so the cells do not border a divot (as seen with Gap43::mCh, Fig. 2I). The cells do round up, which can be seen as gaps between cells (Fig. 3E). 2) We compare Pins to GapCh and only see an enrichment with Pins (Fig. 2H-K). If the enrichment was due to tissue curvature or junction orientation relative to imaging axis, we would see the same enrichment in GapCh. 3) Expression of myr-Pins randomizes spindle orientation in MD14 (Fig. 3M, N).

• In Figure 3I (and 3M?), it appears that there are fewer cell divisions in the presence of myr-Pins. Is this the case? Since cell shapes change during division, and cell shapes influence tissue tension, an increase in cell divisions could lead to a change in tissue tension. This would be important to address, since tissue tension plays an important role in the proposed model.

These images are not taken at the same point of MD1 division ‘wave’, there are the same number of divisions in each condition. These mitotic domains exhibit a ‘wave’ of cell division (Di Talia and Wieschaus, 2012), and so the number of divisions in each image reflect the timing at which we captured the image. Quantifications involved divisions throughout this wave, but we have chosen images for figures which are most representative of what we see. We will add this to the text in the final version of the manuscript.

• The alpha-catenin and Canoe results are a bit confusing: - The rose plot in Figure 4D doesn't show a random distribution of spindle angles, but rather a modest change; most spindles still orient in the normal range. The p value in the figure legend (0.0012) is very different from the one in the figure (5.8284e-04). - Alpha-catenin is the strongest way to disrupt AJs, but A) the epithelium appears to be intact in the knockdown condition and B) spindle orientation is impacted but not randomized. Does this mean that the knockdown is incomplete? Or is Cadherin-mediated adhesion (in which alpha-catenin participates) only partially responsible for force transduction?

We acknowledge that perturbation using ____alpha-cat RNAi does not recapitulate the complete disruption of division orientation seen in embryos expressing myr-Pins. This is likely due to the variability in the strength of RNAi knockdown, which is observed for most RNAi lines that we use. To address the reviewer’s comment, we have added rose plots for individual embryos showing extremes in the severity of division orientation disruption (Fig. 4E and F). For the main plot (Fig. 4D), we have included all the data that we took because we obviously did not want to pick and choose which embryos were used for analysis. So Fig. 4D includes all the variability.

• Given that previous studies implicate Canoe in Pins localization, it seems important to lock down the question of whether Canoe is participating in the mechanism described in this paper. How do the authors know the extent of Canoe knockdown? As suggested by the alpha-catenin results (described above), is it possible that Canoe knockdown is simply not strong enough to impact spindle orientation? Aren't there genetic nulls available? We thank the reviewer for bringing these points to our attention. There are certainly genetic nulls available (Sawyer et al., 2009), but the experiment suggested by the reviewer would not establish the necessity of Canoe in mitotic domain cells. This is because Canoe nulls severely disrupt mesoderm invagination (Sawyer et al., 2009; Jodoin et al., 2015), as well as affecting junctions in the ectoderm during germband extension (Sawyer et al., 2011). Therefore, we would not be able to distinguish what effect of Canoe would be responsible for the spindle orientation using a null mutation. We did better experiments, we used 1) a mutant which specifically compromised mesoderm invagination (snail), 2) laser isolation to show the importance of external force transmission in orienting mitotic domain divisions, and 3) RNAi to deplete Canoe so that mesoderm invagination initiates and pulls on the ectoderm, but where there is clearly compromised Canoe function. This treatment did not cause any effect on spindle orientation arguing against a role of Canoe in this case. In response to the reviewers comment, we added language to the Results to indicate that it is possible that the Canoe knockdown is not strong enough and our rationale for why we did not perform the experiment in a Canoe null (lines 279-282).

Minor Comments:

• It can be difficult to interpret some of the spindle orientation data since the AP axis is vertical in the diagrams but horizontal in the rose plots. Can one of these be flipped so they go together?

We thank the reviewer for this suggestion and have flipped the rose plots so they match the images. Note that because of the large size of the figures, we have had to consistently orient anterior towards the top, which we establish at the beginning of the Results.

• Figure S3 is important information for the reader and should be ideally moved into the main paper. - Protein localizations referred to in text should be annotated on images, as they can be hard to see.

We disagree that S3 should be included in the main paper. The myr-Pins reagent has been used previously so the information in S3 is not new (Chanet et al., 2017).

• There are some discrepancies between figures, legends and text. - p-values differ between figures, legends, and/or text. - Fluorescent markers are labelled differently in figures and legend (CLIP170 in Figure 1) - Graphs appear to show that MD3 polarizes on posterior side, but figure legend says anterior in Figure S1. Vice versa for MD5.

We thank the reviewer for catching these typos. We have fixed these issues.

• Ideally, multichannel image overlays should be shown along with individual channels (b/w). However, it is appreciated that the fluorescent signals are exceptionally weak in this study, presenting a challenge to presentation and to quantification.

We agree the overlays would be nice. However, the Pins::GFP signal is weak compared to the tubulin and Gap43 signals, the merge does not provide more clarity, and the figures are already quite large. Therefore, we have only included the separated the images.

• Graph axes depicting spindle orientation would be more clear if shown in degrees, instead of normalized or in radians.

We thank the reviewer for this suggestion. We have changed the graph axes to be in degrees.

Reviewer #2 (Significance (Required)): Several recent studies have demonstrated that division orientation (in the tissue plane) is governed by tissue level tension. Remarkably, it appears that diverse mechanisms link tension with spindle orientation. Here the authors provide the first in vivo evidence connecting tension to the asymmetric localization of Pins, an important and evolutionarily conserved spindle orientation factor.

Reviewer #3 (Evidence, reproducibility and clarity (Required)): This beautiful manuscript uncovers a role for planar polarized PINS/LGN in orienting the mitotic spindle in Drosophila epithelia. In response to morphogenetic forces acting on adherens junctions, PINS/LGN localises to junctions in a planar polarized fashion to orient the spindle, and de-polarization of PINS/LGN prevents planar spindle orientation. The experiments are very well performed and the findings are robust. The conclusions are well supported by the data. Reviewer #3 (Significance (Required)): These important findings mirror previous work in human cell culture, but crucially reveal that the same phenomenon occurs in vivo in the Drosophila embryo. Thus, the findings underscore the highly conserved nature and in vivo relevance of this phenomenon.

We thank this reviewer for reading the manuscript and their encouraging words.

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 beautiful manuscript uncovers a role for planar polarized PINS/LGN in orienting the mitotic spindle in Drosophila epithelia. In response to morphogenetic forces acting on adherens junctions, PINS/LGN localises to junctions in a planar polarized fashion to orient the spindle, and de-polarization of PINS/LGN prevents planar spindle orientation. The experiments are very well performed and the findings are robust. The conclusions are well supported by the data.

Significance

These important findings mirror previous work in human cell culture, but crucially reveal that the same phenomenon occurs in vivo in the Drosophila embryo. Thus, the findings underscore the highly conserved nature and in vivo relevance of this phenomenon.

Referees cross-commenting

this session contains comments of all reviewers

Reviewer 2

My biggest concern was that the direction of tension isn't obvious. I was particularly puzzled over the ventral furrow experiments, since I'm not clear on how that manipulation impacts the head. I agree with Reviewer #1 that it makes more sense to disrupt the cephalic furrow, but I'm not sure how to do that.

Reviewer 1

Agreed. I guess the question is whether there are cephalic furrow mutants in which mesoderm invagination is not affected. If so, those would be ideal.

Reviewer 3

Hi both. I understand your comments, but I felt that the direction of tension was apparent from the spindle orientation and the cell division axis itself. So, I wasn't concerned about using the snail mutant to prevent gastrulation and thus abolish forces generally.

Reviewer 2

I see. Well I certainly suspect that you and the authors are correct - and I'm enthusiastic about that! - but I'm concerned that using the direction of division to define the direction of tension is getting a little bit circular with the argument. I noticed that their ablation experiments aren't directional; instead they isolate the entire MD. Reviewer 1, as an expert in ablations, do you think it would make sense to make cuts that are only AP or DV?

Reviewer 1

I agree with Reviewer 2 about the circularity of the argument. I was going to propose AP vs DV cuts in sna mutants,with the idea that the wound healing response to those would pull in specific directions. My concern is that It won't be an effect of the same magnitude as the entire mesodermal placode going in, but maybe worth trying?

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

Overview:

In this manuscript, Camuglia et al. show Pins/LGN, which is understood to drive spindle orientation, can localize asymmetrically (with respect to the tissue plane) in the Drosophila embryo. Experimental work (including drug treatments, laser ablation, and knockdowns) lead the authors to propose that this asymmetry is driven by tissue-level tension. The findings are quite interesting and the manuscript is well-written overall.

Major Comments:

• The authors propose that localization is driven by tissue-level tension, but the direction of the tension isn't clear from the experimental work. For example, the laser ablation experiments cut around the entire perimeter of the mitotic domain, rather than along just one tension axis. Similarly, the finding that disruption of the ventral furrow (by Snail RNAi) interferes with spindle orientation in the head is very puzzling; the furrow is A) outside the embryonic head and B) runs in the parallel direction to the divisions considered. The authors need to address the directionality of tension experimentally.
• As acknowledged in the text, the asymmetric enrichment of Pins in MD14 is fairly weak. Since the cells being examined here border a divot in the tissue, and might therefore be curving relative to the focal plane, it would be good to rule out the possibility that some of the asymmetry in Pins intensity is just a consequence of cell/tissue geometry. One way this could be achieved is by showing multiple focal planes.
• In Figure 3I (and 3M?), it appears that there are fewer cell divisions in the presence of myr-Pins. Is this the case? Since cell shapes change during division, and cell shapes influence tissue tension, an increase in cell divisions could lead to a change in tissue tension. This would be important to address, since tissue tension plays an important role in the proposed model.
• The alpha-catenin and Canoe results are a bit confusing:
  • The rose plot in Figure 4D doesn't show a random distribution of spindle angles, but rather a modest change; most spindles still orient in the normal range. The p value in the figure legend (0.0012) is very different from the one in the figure (5.8284e-04).
  • Alpha-catenin is the strongest way to disrupt AJs, but A) the epithelium appears to be intact in the knockdown condition and B) spindle orientation is impacted but not randomized. Does this mean that the knockdown is incomplete? Or is Cadherin-mediated adhesion (in which alpha-catenin participates) only partially responsible for force transduction?
  • Given that previous studies implicate Canoe in Pins localization, it seems important to lock down the question of whether Canoe is participating in the mechanism described in this paper. How do the authors know the extent of Canoe knockdown? As suggested by the alpha-catenin results (described above), is it possible that Canoe knockdown is simply not strong enough to impact spindle orientation? Aren't there genetic nulls available?

Minor Comments:

• It can be difficult to interpret some of the spindle orientation data since the AP axis is vertical in the diagrams but horizontal in the rose plots. Can one of these be flipped so they go together?
• Figure S3 is important information for the reader and should be ideally moved into the main paper.
  • Protein localizations referred to in text should be annotated on images, as they can be hard to see.
• There are some discrepancies between figures, legends and text.
  • p-values differ between figures, legends, and/or text.
  • Fluorescent markers are labelled differently in figures and legend (CLIP170 in Figure 1)
  • Graphs appear to show that MD3 polarizes on posterior side, but figure legend says anterior in Figure S1. Vice versa for MD5.
• Ideally, multichannel image overlays should be shown along with individual channels (b/w). However, it is appreciated that the fluorescent signals are exceptionally weak in this study, presenting a challenge to presentation and to quantification.
• Graph axes depicting spindle orientation would be more clear if shown in degrees, instead of normalized or in radians.

Significance

Several recent studies have demonstrated that division orientation (in the tissue plane) is governed by tissue level tension. Remarkably, it appears that diverse mechanisms link tension with spindle orientation. Here the authors provide the first in vivo evidence connecting tension to the asymmetric localization of Pins, an important and evolutionarily conserved spindle orientation factor.

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

Camuglia, Chanet and Martin investigate the mechanisms that control cell division orientation in vivo, using the mitotic domains (MDs) in the head of the Drosophila embryo as their main model system. They find that cells in the head mitotic domains rotate and align their spindles within 30 degress of the anterior-posterior axis of the embryo. The Pins protein, implicated in spindle orientation in other systems, is planar polarized in mitotic cells. Pins polarization precedes spindle rotation and is correlated with the division angle (but cell shape is not, violating Hertwig's rule). Overexpression of myristoylated Pins results in uniform Pins distribution on the membrane and affects spindle orientation. alpha-catenin RNAi (but not canoe RNAi) disrupts Pins polarity and spindle orientation in MDs 1, 3 and 5. Low dose CytoD injections (which should disrupt force transmission) also result in defective Pins polarity and spindle orientations. Finally, mechanical isolation by laser ablation also disrupts spindle orienttion. The authors find that preventing mesoderm invagination by snail dsRNA disrupts Pins polarity and spindle orientation in the head.

Major

1. Is there a certain chirality in the rotation of the spindles? From Movie 1, it seems like in MDs 1 and 3 at least, a majority of spindles on the right side of the embryo rotate clockwise, while spindles on the left side rotate counter-clockwise? Is that so, and in that case, are there geometric/molecular considerations that could explain that chirality?
2. The authors are experts in mesoderm invagination, and understandably concentrate on the role that forces from that process may have in the orientation of head MD divisions. However, the cephalic furrow forms much closer to the head MDs, and in an orientation that might also explain the alignment of spindles in the head. Is cephalic furrow formation important for Pins polarity and spindle orientation in the head MDs?
3. Does expression of myristoylated Pins afect mesoderm invagination (or cephalic furrow formation)? From Table S1 it seems that a maternal Gal4 driver was used to express myristoylated Pins, which could affect other tissues in the embryo. So it is in principle possible that effects of myristoylated Pins on mesoderm internalization/cephalic furrow formation could affect cell division orientation much like sna loss of function does, but in a mechanism that does not depend on Pins polarity. There is definitely an effect on mesoderm invagination in alpha-catenin RNAi (but not in canoe RNAi) embryos, so I wonder if the effect could be consistently through defects in mesoderm invagination (or cephalic furrow formation), and Pins polarity is really dispensable for spindle orientation. Are there head-specific Gal4 drivers that could be used to drive myristoylated Pins exclusively in the head?
4. Related to the previous point, does mechanical isolation by laser ablation (Figure 6I-N) affect Pins polarity? This experiment could alleviate some of my concerns above, as it certainly does not (should not?) disrupt neither mesoderm invagination nor cephalic furrow formation.

Minor

1. Figure S5: I am a bit confused about the role of Toll 2, 6, 8 in orienting spindle orientation. In Figure S5D it seems that dsRNA treatment against these genes does not disrupt spindle orientation, but Figure S5F shows quite a significant (p=0.0057) effect in triple mutants. The authors favor the idea that Toll receptors do not affect spindle orientation, but the difference with the mutant should be addressed. Furthermore, what happens in MDs 3, 5 and 14 (if the germband extension defect does not affect those divisions)? Is there a difference between dsRNA and triple mutant embryos in these other MDs?
2. No statistical analysis is provided for any of the differences in polarity between Pins and Gap43, and this should be done to demonstrate the significance of the polarization of Pins. Also, particularly for MD14, they should compare anterior vs. posterior polarity, as based on the images in Figure 2H it is not clear that there is a difference between the anterior and posterior side of cells.
3. Figure 2A-D: the authors propose that Pins localizes preferentially to the posterior end of cells (instead of both anterior and posterior ends) in MDs 1, 3 and 14 (and anterior in MD 5). How is the asymmetry in the distribution of Pins along the AP axis accomplished, and is there any significance to it? This should be discussed in a bit more detail (currently no potential mechanisms provided in the discussion, just an acknowledgment of the question).

Typos

1. Line 49: "one daughter cells" should be "one daughter cell".
2. Line 193: "rotation. (Figure 3E-F)." should be "rotation (Figure 3E-F)."
3. Lines 232-237: please review.
4. Line 238: "epithelia cells" should be "epithelial cells".

Significance

This is the first study to my knowledge that demonstrates the role of mechanical forces in polarizing Pins, and provides a nice model to further investigate how mechanical forces generated in one tissue may affect cell division orientation in distant ones. The paper is clear, well written, and quantitative analysis is present for most results. I have some issues with the statistics (or lack thereof) for a couple of results, and potential alternative interpretations for some experiments that in my opinion should be addressed prior to publication. Specifically, it is not clear to me if Pins polarity is at all necessary for spindle orientation in any of the examined MDs.

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

Learn more at Review Commons

---

Reply to the reviewers

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

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

Learn more at Review Commons

---

Referee #3

Evidence, reproducibility and clarity

Hattori et al. assessed the role of astrocytic CD38 by generating astrocyte-specific conditional CD38 knockout mice and discovered defects in social memory, synapse, and spine density in the mPFC. They further showed that conditioned media from CD38-deficient astrocytes are defective in promoting synapse formation. A known astrocyte-derived synapse promoting protein, Sparcl1, is reduced in the conditioned medium from CD38 KO astrocytes and pharmacological experiments suggest that CD38 and calcium signaling regulates Sparcl1 secretion by astrocytes.

The discoveries are novel and important and will be of broad interest to readers. However, the following concerns need to be addressed to improve the manuscript.

Major comments:

1. It's unclear if experiments were conducted while the experimenters are blinded to the genotype of the mice. This is essential for behavior tests.
2. Hippocampus is also important for memory formation. Do synapse and spine densities change in the hippocampus?
3. The proposed model of CD38 inducing Ryr3-mediated calcium release from internal stores is interesting. However, the Barres database showed that Ryr3 is not expressed by mouse astrocytes. Could the authors demonstrate the presence of Ryr3? That's a key link in their model that hasn't been demonstrated to operate in astrocytes.
4. The authors demonstrated reduced synapse and spine density in mPFC. Interestingly, a battery of behavior tests showed no defect, except for the social memory test. Reducing synapses in mPFC should affect a range of behaviors. Why that is not the case here?
5. The authors only tested very short-term memory (30 minutes delay). Does CD38 regulate long-term memory? It would be important to know but I realize that a single paper cannot address all questions and therefore do not think addressing this point is a prerequisite for publication.

Minor comments:

1. Fig. 2F, multiple comparison adjustment is needed.
2. Fig. 3A, scale bar is 10 micrometers, not millimeters
3. Fig. 4C, D, it is unclear if the quantification is normalized to actin loading control. BDNF levels appear lower in KO, though not significantly different, raising the question of whether an equal amount of samples was loaded.
4. Need to validate whether CD38 levels are reduced in P42-46-injected adult knockout before concluding that CD38 is required only during development

Significance

Astrocytic contribution to social memory has not been reported. This study is thus the first report on the role of astrocytes in social memory. Their discovery of CD38-regulation of Sparcl1 release is also novel and important for synapse formation, although more evidence is needed to support this point (see major comments above). This study will be of broad interest to neuroscientists. I have expertise in cellular and molecular neurobiology and can evaluate all parts of the paper.

Referees cross-commenting

I agree with the issues that the other reviewers pointed out, especially the need for improving data reporting and consistency/accuracy. Overall, I think this manuscript has potential and the issues are addressable.

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

Learn more at Review Commons

---

Referee #2

Evidence, reproducibility and clarity

Summary

In their manuscript, Hattori et al., put forward evidence that the knock-out of CD38 expression in astrocytes at approximately post-natal day 10 (referred to as CD38 AS-cKO P10) leads to a specific deficit in social memory in adult mice, while other types of memory remain unaltered. Using immunohistochemistry (IHC), the authors found a reduced number of excitatory synapses in the medial prefrontal cortex (mPFC) of CD38 AS-cKO P10 mice. Switching to in vitro primary cell culture models, the authors identify the astrocyte secreted protein SPARCL1 as a relevant synaptogenic factor. Using pharmacological dissection of relevant signaling pathways, Hattori et al., propose that cADPR formation and calcium released from intracellular stores, is essential for SPARCL1 secretion from astrocytes. Finally, the authors analyzed the transcriptome of primary CD38 KO astrocytes using bulk mRNA sequencing, and found that genes related to calcium signaling were downregulated in these cells.

Major commments:

• Are the key conclusions convincing?
  1. From a global perspective, the multiple lines of evidence provided by the authors strongly suggest that expression of CD38 in astrocytes is important for synaptogenesis in the mPFC of P10 mice, with ablation of CD38 and reduced synapse formation leading to social memory deficits at P70. However, the data concerning the role of astrocyte-secreted SPARCL1 is not particularly strong: further experiments are needed to support this claim (see below).
• Are the claims preliminary or speculative?
  1. As it stands, there is no proof that the claimed astrocyte-specific deletion of CD38 is actually astrocyte specific. This evidence is crucial: without it the reported effects could be due to non-specific CD38 knock-out in other CNS cells. In this respect, the Western Blot in Supplementary Figure 1A does not provide information on astrocyte-specific deletion, merely that CD38 was globally reduced in the mPFC. Interestingly, the authors have previously published data (Hattori et al., 2017, 10.1002/glia.23139) showing that CD38 expression is mostly astrocyte-specific, peaking at p14, which coincides with the peak period of synaptogenesis. The degree of CD38 heterogeneity is also an issue that I think the authors need to consider. Do they information on this? Is CD38 expressed in every astrocyte of the CNS, or are there some astrocytes that are CD38 negative at P14? Is the mPFC a region specifically enriched in CD38 positive astrocytes and does this explain the observed behavioral deficit? I think if this is known, the authors should mention it in the "Introduction" or "Discussion". If this is not known, maybe the authors could provide data addressing the issue.
  2. I think the authors should take more caution in claiming that SPARCL1 is the main factor secreted through the CD38 signaling pathway and responsible for increased synaptogenesis. This is for several reasons, all centered on data displayed in Figure 4 and Supplementary Figure 6:
     • a) Western Blot (WB) data: The "Materials and Methods" section for WB does not indicate how protein loading and transfer efficiency were controlled for. Normalizing to β-Actin levels is an acceptable way to control for loading and transfer efficiency when using cell lysates. However, in the absence of such an abundant structural protein in conditioned media it is unclear how loading and transfer was controlled for under these conditions. Do the authors normalized the CD 38 KO AS ACM data by expressing protein levels relative to those from WT AS ACM? Is BDNF being used as a control, based on proteomics data? If so, why is proteomics data not given in the manuscript and why is this control not shown for all ACM blots? I realize that (quantitative) blotting using ACM is difficult, but I am also not convinced that the methodology used is sufficiently rigorous. Simple steps to give confidence would be Coomassie staining of gels both before and after membrane transfer, to show that i) the total protein amount loaded was the same in each lane of the gel and ii) the transfer to the nitrocellulose membrane was complete. In addition, Ponceau S staining of the nitrocellulose membrane should also have been performed and displayed, to show (roughly) equal amounts of protein were transferred for each lane. In summary, the WB data quantification needs to be better controlled. The values of the Y axis in these graphs (and throughout the manuscript) are simply too small to be read properly. Finally, I want to highlight the general lack of precision regarding the nature of the replication unit (the "n"). For example, the legend of Figure4C-D states "n = 6", but we have no idea if these are 6 independent primary cultures originating from 6 mice, 6 independent cultures from the same mouse, 6 repeats of the Western Blot using the same sample etc. This issue is valid for the whole manuscript: in my opinion, the authors should be more much careful when it comes to these crucial elements of scientific reporting.
     • b) While the data hint at an important role of SPARCL1 in synapse formation, when the authors tested if ACM from CD38 KO astrocytes supplemented with exogenous SPARCL1 could rescue synapse formation, the effect was incomplete, with only a trend to an increase in synapse number (Figure 4J-K). Perhaps the authors simply forgot to indicate the statistical significance of differences between the experimental groups (Figure 4K)? However, if there really were no statistically significant differences observed, the authors should reduce the strength of their conclusions regarding SPARCL1. This protein may well be pro-synaptogenic but, as it stands, other factors could well be in play. Perhaps the authors should have tried higher concentrations of SPARCL1 to further boost synaptogenesis? In this respect, the SPARCL1 knockdown (KD) experiment in Supplementary Figure 6B-D is an important addition, but should be supplemented by rescue with an siRNA-resistant recombinant SPARCL1? If SPARCL1 is a major player in synaptogenesis, the prediction is that synapse numbers would be close to wild type levels with this approach.
     • c) In my opinion, there are also issues with the data displayed in Figure 4H-I. The authors want to convince the reader that SPARCL1 is mostly an astrocytic protein using immunohistochemistry on mouse mPFC sections, co-labelled with antibodies against neuronal and astrocytic markers. In these panels, we are presented with images showing a few cells, in which it seems SPARCL1 is absent from NeuN positive cells, present in WT astrocytes and reduced in CD38 AS-cKO P10 astrocytes. However, the numbers of cell counted and lack of quantification severely impact on the strength of this conclusion. In my opinion, the authors should have quantified their IHC data by counting cells and establishing the ratios of SPARCL1 positive over NeuN or S100β positive cells, in both control and CD38 AS-cKO P10 animals. This experiment would provide critical information that the conditional gene targeting strategy is robust. The authors should also consider quantifying the intensity of the SPARCL1 signal in astrocytes. This is recommended as the image displayed in Figure 4I for the CD38 AS-cKO is problematic: are the authors really claiming that the reduction in SPARCL1 expression following cKO of CD38 in astrocytes is at best only partial? Is 11 days between the first tamoxifen injection and tissue fixation actually sufficient to allow for CD38 turnover? With low levels of protein turnover, the possibility exists that residual levels of CD38 are still sufficient to impact SPARCL1 levels. What would happen if there is a greater interval between tamoxifen administration and tissue recovery? Would levels of synaptogenesis be further reduced? Is this an issue of production versus secretion or a combination of factors?
  3. The heatmap (Figure 5E-F) is simply too small to interpret. The color choice is also not accessible for colorblind readers. The authors might consider displaying this heatmap in a separate figure. The authors should also provide a supplementary table where all the genes detected are listed along with their respective counts. Furthermore, it is surprising that the authors only found genes being downregulated in CD 38 KO astrocytes. Were they really no genes up-regulated? The authors might also want to indicate the genes belong to each of the ontological categories listed in Figure 5F. On p. 11, Figure 5E: The authors should indicate in the main text they performed bulk RNA-sequencing and not another type of RNA sequencing (like single cell RNA sequencing for instance). The authors indicate n = 2 but we have no indications of the nature of the replicate (also see earlier comments). Please amend.
• Are additional experiments necessary? I think supplementary experiments are essential to support the claims of the paper. Most are described in the section above, but to summarize:
  1. Show data to prove that the CD38 AS-cKOP10 model is astrocyte-specific and leads to a total loss of CD38 in these cells.
  2. WB data: The issue of protein loading and transfer efficiency should be dealt with. Quantifications should be revisited.
  3. The authors should quantitatively analyze the different IHC performed in Figure 4H-I.
  4. The authors should provide more information on their RNA sequencing data: list of genes detected with their FPKM values etc. The authors should display the RNA sequencing data in a separate figure, allowing the heatmap to be enlarged.
  5. LC-MS/MS data: the authors should provide the list of all the proteins they identified in their LC-MS/MS experiment. As a supplementary table for instance? The majority of these experiments should be able to be performed with pre-existing samples/tissue slices. If not, the experimental pipeline necessary exits and these supporting experiments should not be too burdensome.
• Data and methods presentation Methods: The authors need to work on this aspect of the manuscript. Most of the important details are already described, but some crucial ones are missing, while the phrasing used to describe methods is sometimes misleading. I will give some examples here, but this is not an exhaustive list. The fact that the manuscript is riddled with small mistakes, inconsistencies and/or oversights makes it difficult to read and creates a negative impression. The whole manuscript would benefit from a thorough proof-reading, preferably by a native speaker.
  1. in the "Immunohistochemistry and Synaptic Puncta Analysis" section on p. 21-22, we have no indication of which antibodies against "GFAP, NDRG2, VGlut1, PSD95, S100β, NenN(?) and SPARCL1" were used. It is standard practice to indicate the company, product number and lot number. The authors must also indicate the dilution at which they use these antibodies. On p.22, the authors write the cells were incubated with "Alexa- or Cy3-conjugated secondary antibodies". The excitation wavelengths of the Alexa dyes used need to be given.
  2. The authors need to provide more details on the microscope they used. Merely writing "using a 63× lens on a fluorescence microscope" (p.23) is insufficient.
  3. In the "LC-MS/MS" method the authors wrote: "Briefly, these proteins were reduced, alkylated, and digested by trypsin". I think that in the reduction and alkylation steps, chemicals other than trypsin were actually used. This sentence should be modified to reflect this.
  4. p.19: "uM" is written when the authors very likely mean "µM". Please check the whole manuscript for repeat examples. I know this is often lab "short-hand", but it should be avoided in scientific publications.
  5. The authors should be careful when describing their data to always indicate whether they referring to experiments performed using cultured astrocytes or not. As it stands, the text is confusing: for instance, when describing RNA-sequencing data in Figure 5, the main text appears to indicate that these astrocytes were acutely isolated from adult mice, when in fact they were obtained from primary cultures. Given concerns in the literature about potential differences between acutely isolated and cultured astrocytes (Foo et al., Neuron, 2011), this is essential. Data presentation: The figures appear to have been produced in a rush - and almost have a "screenshot" feel to them. This is not a scientific issue per se, but does impact on the overall impression given by the manuscript. The following is a non-exhaustive list of issues with the figures. I list the major ones that the authors should correct.
  6. Almost all Y axis labels are too small. The authors should comply to the basic journal requirements in terms of font sizes. Some axes do not end on a tick (e.g. Figure 3R). This is not dramatic, but should be corrected. Globally, the authors need to display bigger bar plots - most of them are extremely hard to read. Labeling should also be checked: Figure 4K, the Y axis label indicates values displayed are in %, when I think the axis graduation displays ratio values. Some of the IHC pictures are also too small to be easily interpreted.
  7. The heatmap in Figure 5E is impossible to read and, as such, has little or no value for the manuscript.
  8. Scale bars: where is the scale bar in Figure 2A? Figure 3A-H: Is the scale bar really representing 10 millimeters? Supplementary Figure 3A: scale bar is missing. Please check for similar issues throughout the manuscript.
  9. Figure Legends are problematic, and often contain incorrect or incomplete information. Examples include: Supplementary Figure 1: The description of panels J, L and N appears to be missing. Please also use the Greek letter beta and not 'b' for S100β. Supplementary Figure 5: I think the term "KO" is missing after CD 38 in the legend title. Figure 3: why state that nuclei were counterstained with DAPI in Figure 3P,Q, when this precision is not given for panels Figure 3A-H? Figure 3A-H: If the authors choose to explicitly state PSD95 is a post-synaptic marker, why not indicate that VGlut1 is a pre-synaptic marker? Same issue in Supplementary Figure 4.
  10. There are multiple instances of panels being wrongly referred to in the main text. On p.10, Figure 4H is referenced, when I think the authors mean Figure 4I; on p.10, Figure 4I-J are referred to when the authors clearly describe data found in Figure 4J-K. These types of mistakes are problematic and recur throughout the manuscript.
• Statistical analysis As mentioned above, the exact nature of the replicates is often not stated, when the "n" number is indicated. The authors must correct this issue and give the information either at the appropriate point in the main text or in the figure legend.

The authors should also be more consistent in the way they indicate which statistical tests were performed. This should also be indicated either at the appropriate point in the main text or in the figure legend. Furthermore, care should be taken to ensure statistics are presented in an appropriate manner: at the end of legend for Figure 4, it is indicated #p < 0.05 vs. CD38 KO ACM. This hashtag symbol is completely absent from the figure. In Figure 4F-G, the lack of statistical symbols seems to indicate no statistical tests were performed on these data, when the legend covering these panels states "*p < 0.05 versus P70", indicating some tests were done. We cannot interpret this panel without knowing which comparisons were done exactly and which were significant.

In the "Materials and Methods", the authors give no indication that the assumptions of the statistical test they used were met (normality of data distribution for t-tests, homogeneity of variances for ANOVA...). This needs to be checked, and if not met, appropriate non-parametric tests should be used instead.

Minor commments:

• Specific experimental issues that are easily addressable. Most of the experimental issues that need to be addressed are given in previous sections and should be easily addressable.
• Citation of previous studies? Adequate
• Clarity and accuracy of text and figures There are issues with the clarity and accuracy of text and figures - which are described above. The text is also often problematic in its phrasing and other, more fundamental aspects. For instance, the authors spent a considerable amount of time speaking about the role of oxytocin, when they only performed one measurement of oxytocin levels in mice.
• Suggestions to improve the presentation of data and conclusions? All my suggestions to improve the presentation of data can found in previous sections. As for improving the authors presentation of their conclusions, the authors should make a considerable re-drafting effort, particularly for the "Discussion", which lacks clarity in how supporting arguments are built and presented. For example, on p.13, I am confused with the argument made by the authors. Their data are focused on synapses onto pyramidal neurons of the mPFC, but here the discussion states that the behavioral phenotype they observed in CD38 AS-cKOP10 might be explained by a lack of mPFC neurons synapsing onto neurons in the Nucleus Accumbens (assuming that "NAc" really refers to this brain region, as the definition is missing from the text). I think the authors should make it clear if this is their interpretation of their own result, which essentially renders their focus on mPFC pointless, or a speculation on possible other mechanisms that could also explain their behavioral results. Personally, given the data shown, I believe the authors should focus on explaining how their data in mPFC might explain the behavioral output observed. The authors could also provide perspectives on how the hypothesis laid down in this paragraph would be tested. When the authors write on p.14 "We identified SPARCL1 as a potential molecule for synapse formation in cortical neurons" why use the word "potential"? Does this mean the authors consider their data on SPARCL1 (one of the key messages of the paper) invalid? If the authors themselves think the role of SPARCKL1 is ambiguous based on their own data, they should perform further experiments. P. 13, the authors write: "Moreover, many studies have shown that astrocyte-specific molecules, including extracellular molecules such as IL-6, are involved in memory function"; Interleukin 6 (Il-6, abbreviation not defined in the manuscript) is definitely not an astrocyte-specific molecule (see, for example, Erta et al., 2021 10.7150/ijbs.4679).

Significance

NATURE AND SIGNIFICANCE OF THE ADVANCE: I think that despite the issues described above, this manuscript, once revised, could have a strong impact in the field. It would fuel the current paradigm shift which puts astrocytes at the forefront of neuronal circuit wiring during development with links to adult behavior. By identifying clear molecular targets involved in astrocyte-driven synaptogenesis, this article could help the clinical field to find new druggable targets, which may help reverse aging-related cognitive decline.

COMPARISON TO EXISTING PUBLISHED KNOWLEGDE: This work adds new data in the specific and growing line of research that study how astrocytes control synaptogenesis. Recent reviews have summarized advances in this field (Shan et al., 2021, 10.3389/fcell.2021.680301; Baldwin et al., 2021, 10.1016/j.conb.2017.05.006).

AUDIENCE: Neuroscientists in general, clinicians interested in cellular and molecular causes of neurodevelopmental disorders leading to social dysfunctions.

REVIEWER EXPERTISE: Astrocyte biology; Astrocyte-neuron interactions and synapse assembly; Neuronal circuit formation and plasticity

Referees cross-commenting

After careful reading of the other comments, I feel that there is considerable agreement/overlap between the reviewers on the main issues with this manuscript. Perhaps the major difference relates to the amount of further work necessary for the manuscript to be publication ready.

As Reviewer 3 rightly points out, this is always a moot point: how much is it reasonable for reviewers to ask authors to do? While I agree with all of Reviewer 1's comments regarding the rigour of the mass-spec/western blot analysis, it seems to me that from a molecular/cell biological point of view, the key issue is whether Sparcl1 is a synaptogenic factor released from astrocytes following CD38/cADPR/calcium signaling (irrespective of whether other factors may be in play); and whether raising Sparcl1 levels is sufficient to recover spine morphology and synapse numbers. Of course, if these experiments were performed in vivo using AAV-mediated overexpression of Sparcl1, it is also reasonable to think that the deficit in social memory may be reversed on testing.

The issues of whether there is a difference in observable behavioral phenotypes between the astrocyte-specific and constitutive CD38 knock-outs is an interesting one, as is why there is only a deficit in social memory seen following astrocyte-specific CD38 ablation. These issues should at least be discussed.

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

Learn more at Review Commons

---

Referee #1

Evidence, reproducibility and clarity

Summary

In their work submitted for review, Hattori et al. identify an astrocyte enriched protein (CD38) as important for social memory tasks in mice. The authors developed a conditional KO model to remove CD38 specifically in astrocytes using the GLAST-CreERT2 line crossed to a CD38 floxed line. The investigators use a three-chamber social approach test to show that loss of CD38 leads to reduced interaction time with a novel social stimulus only when the animal is given a break between test periods. The authors test whether changes in neuronal morphology or synapses in the medial prefrontal cortex (mPFC), a region important for social memory, can account for their behavioral phenotype. The researchers found that mPFC neurons in their conditional CD38 KO (cKO) animals have significantly less mature spines than wild-type (WT) controls. The authors then claim that this reduction in mature spines correlates with a reduction in VGluT1 positive excitatory synapse density in mPFC of cKO vs. WT. Next, the investigators use mass spectrometry of astrocyte conditioned media and neuronal cultures treated with astrocyte conditioned media to test whether a known astrocyte secreted synaptogenic factor, Sparcl1/Hevin, could underlie their reported changes in synapse density in their cKO animals. Finally, the authors use pharmacological inhibitors against different components of the CD38 signaling pathway to test whether CD38 regulates Hevin secretion by astrocytes. While the reported behavioral phenotype is interesting, this reviewer has several major concerns with the data claiming that reduction in Sparcl1/Hevin is underlying synaptic phenotypes in the CD38 cKO. Therefore, the paper is not suitable for publication without addressing the concerns listed below.

Major Concerns:

Synapse analysis in vivo: For the analysis of VGluT1 excitatory synapses in mPFC, it is not clear how the statistical analysis was performed. From the plotted error bars, it seems that the investigators used individual z-projections as the n for a t-test. This is inappropriate for this analysis as it would overinflate the N and down the p-value. It would be more appropriate to plot and compare animal averages between conditions or use a test that can account for the fact that there are repeated measures taken from the same animal. Additionally, the authors note a decrease in VGlut1+ puncta in the global CD38 KO but no change in the protein levels in both the global and cKO.

Synapse analysis in vitro: The authors are missing key experimental controls for their analysis of synapse induction by astrocyte conditioned media. Firstly, the authors do not include a condition of neurons cultured alone without astrocytes or astrocyte conditioned media treatments. This is critical to this experiment because, without this control, it is impossible to assess the effectiveness of the astrocyte conditioned media or any recombinant protein treatments on synapse formation. Secondly, the authors give very few details and no supporting data about the purity of their neuronal cultures. This is critical to this experiment because any contaminating astrocytes in their cultures could severely skew the data for any given condition. Finally, the authors do not specify how they determined the doses for astrocyte conditioned media and Hevin treatments. The researchers give no details on how the astrocyte conditioned media was collected or treated before adding onto neurons. For this experiment to be viable, the researchers must collect the conditioned media in serum-free media, determine the protein concentration of their samples, and the dose-response to the astrocyte conditioned media must be performed to determine the optimal dose for each batch. When comparing between genotypes, this type of quality control is critical to assess whether there is, in fact, a difference in their synaptogenic capacity.

Western blots: All western blot quantification of astrocyte conditioned media should include total protein normalization. The authors do not describe how they normalize the astrocyte conditioned media blots, but without a total protein stain to normalize, it is impossible to be sure the same amount of protein was loaded into the gel for each lane. In Figure 3L, the western blot data showing the expression of VGluT1 and PSD95 should be improved, and a better representation is recommended. It is also strange that the CD38 cKO has no expression because CD38 is also expressed in endothelial cells. Why not isolate astrocytes from CD38 KO? Also, for VGluT1 and PSD95 western blots, it would be better to test mPFC lysates rather than whole cortical lysates. Astrocyte morphogenesis: Since the astrocyte-specific deletion of CD38 from P10 impairs postnatal development of astrocytes, the authors should investigate if the impaired synaptogenesis seen in later stages is due to impaired astrocyte morphogenesis or the defect in the secretion of synaptogenic proteins like Sparcl1/Hevin or thrombospondins.

Mass spectrometry: There is no information about how many samples were used for mass spectrometry. This reviewer is concerned that this experiment may be underpowered given that other published datasets have identified significantly more proteins in wild-type ACM (about double than what was identified here). There needs to be a quality assessment of the ACM to help ensure the production protocol can capture the full extent of proteins secreted by cultured astrocytes.

RNA sequencing: RNA sequencing results seem underpowered, and an accurate description of their collection methods is missing. It also seems to this reviewer that any prolonged culturing of the astrocytes would lead to additional transcriptional changes independent of their genetic manipulation. To avoid confounds due to culture artifacts, it might be cleaner to FACS sort astrocytes using a fluorescent reporter such as the Aldh1l1-eGFP line or RTM in their GLAST-creERT2 model. In the latter case, this could also provide data on the specificity of their recombination, which is lacking elsewhere in the manuscript.

Comparison between astrocyte-specific cKO and global KO: Considering the abundant expression of CD38 in astrocytes compared to other cell types in the brain, I am wondering whether the comparison between the current astrocyte-specific CD38 cKO and the previous constitutive CD38 KO mice would provide a different phenotype with respect to its importance in synaptic function in neural circuits that mediate social behaviors in various brain regions. The authors note the importance of CA1, CA2, and NAC in social memory, but they only assessed synapses in mPFC. Multiple studies, including one from the authors, have reported that constitutive CD38 KO mice exhibit impaired social behaviors. Expanding beyond what is already known would require better spatial and temporal regulation of CD38 expression than presented here.

Rescue experiments: The authors claim that reduced levels of Hevin secretion are responsible for reducing intracortical synapses in mPFC and the inability of their CD38 KO ACM to stimulate synapse formation. However, Hevin has primarily been linked to the formation of VGluT2+ synapses with only a transient effect on VGluT1+ synapses. Furthermore, Hevin's synaptogenic effect in astrocyte conditioned media is masked by its homolog Sparc. To claim that Hevin is responsible for reducing VGluT1+ synapses in mPFC the authors need to do a rescue experiment by expressing hevin in CD38 KO through AAVs brains or intracortical injections of recombinant Hevin.

Other synaptogenic factors: The authors focus on Sparcl1/Hevin; however, other synaptogenic factors have been reported to affect VGluT1+ excitatory synapse formation and development directly. Notably, thrombospondins have been shown to regulate the formation of this specific synapse type through their receptor a2d1. The authors do not report any investigation into this family of factors despite their clear link to VGluT1+ synapse development.

Effect of CD38 cKO on astrocyte numbers: The authors note that CD38 cKO alters GFAP expression; however, they also report a decrease in the number of GFAP+ and S100ꞵ+ cells without a change in NDRG2+ cells. The authors should address this discrepancy in astrocyte numbers with additional known markers such as Sox9.

MBP quantification: The authors previously reported changes in MBP expression and oligodendrocyte maturation in the global CD38 KO animals. However, there is no quantification of the MBP staining in the cKO in supplementary figure 1. It would be important to verify that white matter structures developed properly in their cKO model, especially in mPFC.

Minor Concerns:

1. SPARCL1 annotation should be Sparcl1.
2. Avoid repetition of the same sentences in multiple places. E.g., The sentence- "Social behavior is essential for the health, survival, and reproduction of animals" is repeated both in the abstract and introduction.
3. The introduction needs to be thoroughly revised. In the first paragraph, a description of various studies(Fmr/Mecp2) which indicated the importance of synaptic function in neural circuits that mediate social behaviors in various brain regions could be presented later part of the introduction in a very concise manner since the article doesn't cover anything related to these genes. This part can be presented along with the narration of CD38, where authors described its importance in social behavior. Introduce the importance of social behavior and their behavioral paradigm, especially what social memory is and what brain regions are important for it.
4. Introduction feels too short and abrupt.
5. In Figures 2 and 3, Are the spine numbers/density/synapses affected in the p42 ctrl/CD38 AS-cKO group compared to the p10 ctrl/CD38 AS-cKO group?
6. In Figure 2; The authors should compare both the behavioral phenotype seen in two different tamoxifen injection/time points with the respective constitutive CD38 KO mice data.
7. In Figures 3 and 4, the authors should analyze the spine numbers/density both in WT or CD38 KO ACM treated experiments and Sparcl1 KD/Sparcl1 treated rescue experiments?
8. The discussion section needs to be revised to reflect better the conclusions drawn from the data without overstatement.

Significance

Understanding the mechanisms underlying control of behaviors is important and linking non-neuronal cell types to behavioral processes is novel and timely. However, the study at its current state lacks important controls, and interpretations are overstated and often too targetted to a favorite mechanism. These concerns limit the impact of the study and reduces its significance.

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

RC-2022-01245 Willemsen et al., 2022

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

Summary:

Willemsen et al studied the contribution of PSMB4 and PSMB8 on proteasomal activity in adipocyte tissue/cells. Mutations in PSMB4/8 have been associated with metabolic diseases that lead to inflammation and proteotoxicity. They used mice and murine cell lines to assess the abundance and activity of the proteasome as well as the stress response upon depletion of psmb4 and regulatory factors. The study is interesting and could provide new insight into the role of specific proteasomal subunits of the immunoproteasome on the metabolism and associated diseases. However, there are a number of issues on the presented data that should be addressed. See below.

RESPONSE: We thank the reviewer for her/his positive remarks.

Major comments:

1. Most of the presented data are secondary data (graphs) and the differences between the experimental conditions (e.g. siRNA) are sometimes only minor and statistical analyses lacking or only indicated with a letter (a, b,..), but its meaning not explained. RESPONSE: We apologize if some of the statistics were difficult to see. Using letter designations for groups of indifference declutters the figure compared to using asterisks to indicate significant differences between groups. We have now made sure the statistical analyses are emphasized in the methods and figure legends.

The authors should introduce all tested marker genes e.g. the genes that were analyzed in figures 2H/I, 3D-F, 4A/D. What was the hypothesis and do they represent all or a selected set of genes of the integrated stress response?

RESPONSE: We apologize if the relevance of these markers was left unclear. We have now introduced the marker genes in the text.

Figure 1A: RNA levels were analyzed yet not protein levels. Why not?

RESPONSE: As we performed loss of function experiments later anyway, we decided not to venture into more descriptive analyses. The fact that Psmb4 and Psmb8 are robustly expressed in adipocytes was enough for us to justify studying their function further.

Figure 2C: This figure is problematic. Apart from the smear on the right side, a loading control is missing. This is essential to quantify signal intensities. Moreover, on the left side, the intensities of lanes 1 and 2 are very different, yet are both controls and were used for the conclusion that proteasomal activity is reduced upon siPsmb4 (lanes 3 and 4 - that do not differ from lane 2). In addition: from which day were the data collected? This information is missing, but important as Figure 2E shows opposing proteasomal activity on day 3 and day 5.

RESPONSE: We agree with the reviewer that the duplicate of the scrambled control cells showed variation. Therefore, we have now repeated the experiment and replaced the figure (now figure 2A). The outcome is not affected, as knockdown reduces proteasomal activity and leads to abnormal proteasome formation in the native PAGE. Regarding the internal loading, it is common practice to display both in-gel activity and immunoblot on the membrane as is. For recent examples in the literature please see VerPlank et al., PNAS 2019 (10.1073/pnas.1809254116) or Yazgili et al., Cell Press STAR Protocols 2021 (10.1016/j.xpro.2021.100526). Obviously, the same amount of protein was loaded, and this is also seen in the immunoblot. As this is a native PAGE, there is no beta-tubulin or other commonly used loading controls for immunoblots. Furthermore, we apologize for the missing information, this experiment was performed using day 5 mature cells. This information is now included in the figure legend.

Figure 2D: tubulin was used as loading control, yet the signal of tubulin in lane 1 is by far weaker compared to the other lanes. How does that affect the quantification (missing) of Nfe2I1?

RESPONSE: We have now included the quantifications, which do not affect the outcome of the experiments and the conclusions drawn.

Figure 3C and Fig 2E (control vs siPsmb4) contradict each other. Please clarify.

RESPONSE: We respectfully point out that the reviewer might have overlooked that 2E shows the time course and 3C only shows day 5 data – both in 2E and 3C, day 5 total proteasomal activity is (insignificantly) increased. Hence, the panels do not contradict each other.

Figure 3B: Issues with the loading control: tubulin signals are in first 3 lanes much weaker. Where is the quantification for that data set that takes the fluctuations of the tubulin signals into account?

RESPONSE: We have now included the quantification, which does not affect the outcome of the experiments and the conclusions drawn. The quantifications can be found in figure S3.

Minor comments:

Why did the authors not use human adipocyte cells and performed all experiments in murine cells?

RESPONSE: The advantage of the cell lines used lies in the ability to study both brown and white features as well studying aspects of adipogenesis and thermogenesis simultaneously. Based on this comment and the comment of reviewer #2, we have reproduced our findings in 3T3-L1 adipocytes, in the hope of strengthening our study. These data are shown in the new Supplementary figure 4.

In which cell/tissue is Psmb4 expressed?

RESPONSE: Thank you for this question, we have now measured Psmb4 expression in a panel of mouse tissues. As shown in figure S1, Psmb4 is ubiquitously expressed in all tissues measured with the highest levels in kidney and liver, followed by brown fat.

Figure 4G: information on the different colors is missing.

RESPONSE: Thank you for bringing this to our attention. We have now included a legend.

The result section appears to have been restructured as sections do not build up on each other well. This should be corrected.

RESPONSE: We appreciate this critical feedback. We have now improved the flow for an enhanced reading experience.

There are a number of grammatical errors or doubling of words/phrases e.g. bottom of page 1: "In addition, PRAAS patients display suffer from..." or on page 2: "Adapting proteasomal activity to the needs of the UPS..." This statement does not make sense. Maybe the authors mean "proteolytic demands"?

RESPONSE: Thank you, we have fixed the remaining typos.

Although, the UPS is a proteostasis node, the authors should avoid statements such as "We show that proteostasis and lipid metabolism are intricately linked..." Better is "UPS activity and lipid metabolism..." Or the authors should expand their analysis to protein synthesis, folding and additional clearance pathways.

RESPONSE: Thank you, we have specified our statement regarding proteostasis and UPS.

Reviewer #1 (Significance (Required)):

This study links clinical research with basic science and if the authors address the above mentioned issues this work will provide new insight into the role of the UPS in the lipid metabolism.

target audience: clinical scientist on lipid metabolism and basic researchers on the UPS and associated pathologies

my expertise: UPS

RESPONSE: We are very thankful for these positive concluding remarks.

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

The molecular mechanism by which proteasome mutations cause lipodystrophy in PRAAS, which is caused by proteasome dysfunctions, has not been well understood. However, it was shown that Psmb4 (β4), a component without enzymatic activity, is required for the formation and maintenance of adipocyte function. In the proteasome dysfunction state of Psmb4 deficiency, the expression of Nfe2l1 was enhanced for proteostasis, but it could not complement the adipocyte formation defect. We showed that repression of Arf3, which is associated with stress response and is markedly expressed in this situation, resulted in the recovery of inflammation and adipogenesis.

Major comments

1. The Graphical abstract seems to indicate that Loss of PSMB4 activates NFE2L1 and ATF3, resulting in the suppression of proteotoxicity and Inflammation. In the case of NFE2L1, this is correct, but in the case of ATF3, as shown in Fig. 4D, ATF3 acts in a promotive manner on Inflammation, causing misleading to the reader. RESPONSE: Thank you, we agree that this aspect of the graphical abstract was partially misleading. We hope the new version now makes more sense.

Fig. 2 shows the rise of Nfe2l1 and the restoration of proteasomal capacity on Day 5 (Fig. 2E). [Nfe2l1 is cleaved, and initiates the transcription of proteasome subunits, which results in restoration or heightening of proteasomal capacity (16,17).] It is known that the brown adipocyte mount an adaptive response to overcome UPS dysfunction, and the transcription of proteasome subunits was increased in this experimental system. However, there are no results showing that the transcription of proteasome subunits is actually increased in this experimental system.

RESPONSE: Thank you for pointing this out. In Psmb4 KD cells, we see an increase in Psmd2 protein levels (Fig. 3B). In addition, we see a small increase in expression levels of various proteasome subunits. We have now included a graph showing these expression levels (Fig 3C).

Are Psmb4KO mice available? If yes, are there any symptoms? Is there any change in proteasome activity, etc.?

RESPONSE: We do not have a Psmb4 KO mouse model, yet, and to the best of our knowledge, none is available. We agree with the reviewer that it would be insightful to study Psmb4 in an in vivo model, but in this project, we have used a cell model to study the cell intrinsic mechanisms of Psmb4.

4A. In PRAAS patients, most of the lipodystrophy occurs in white adipocytes, but if Psmb4 deficiency is induced in white adipocytes, do they show the same dynamics?

RESPONSE: Thank you for your stimulating question. We have repeated our Psmb4 KD experiments in 3T3-L1 cells, to study the dynamics in a white adipocyte model. We found that also in 3T3-L1 cells, Psmb4 knockdown disrupts adipogenesis. The results can be found in Supplemental figure 4.

4B. The first mention of heat production was made, but it was not clear how much the patient's cyclic fever symptoms were related to changes in brown adipocyte function.

RESPONSE: Our data suggest that aberrant brown adipose tissue does not contribute to cyclic fevers in PRAAS patients. We elaborate on this in the Discussion.

minor points

fig2; The numbering of the figure is not correct.

RESPONSE: Thank you, we have corrected the error.

Fig. 2: (day 5) in the figure legend of (E) is unnecessary.

RESPONSE: Thank you but based on the other reviewers’ questions we think it is important to indicate the stage of the differentiation.

Fig. 4 The figure legend in (A) and (B) are switched.

RESPONSE: Thank you, we have corrected the error.

In Fig. 4 (G), there are n = 8 and n = 6 in the figure legend, which is difficult to understand.

RESPONSE: Thank you, we have corrected the error.

Reviewer #2 (Significance (Required)):

It has been thought that the accumulation of defective proteins caused by proteasome dysfunction stresses cell metabolism and leads to lipodystrophy, but the detailed mechanism has not been understood. In this paper, we have clarified a part of the mechanism that links the accumulation of ubiquitinated proteins caused by proteasome dysfunction to the disruption of proteostasis, inflammation and adipogenesis. The results of the study, which showed the relationship between intracellular proteostasis, inflammation and lipid metabolism, will help us understand not only PRAAS patients but also abnormal lipid metabolism, obesity, induction of inflammation, and chronic inflammation with persistent inflammation.

RESPONSE: We are very thankful for these positive concluding remarks.

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

In this study Willemsen et al. investigated the role of proteasome subunit beta 4 and 8 (PSMB4/8) in immortalized brown (pre)adipocytes regarding adipogenesis ability, inflammation, function, and proteostasis. The group showed that Psmb4/8 are expressed in brown adipose tissue and adipocytes but that they are differently regulated. The loss of PSMB8 had no effect on brown adipose tissue/adipocyte function. In contrast, knock-down of PSMB4 altered proteostasis, which was partially compensated by NFE2L1, as well as reduced adipocyte differentiation, lipid accumulation, and beta adrenergic-stimulated glycerol release in immortalized brown adipocytes. The group further demonstrated that the effect of PSMB4 knock-down on impaired adipogenesis was mediated via Atf3 activation.

The manuscript is well-written with clearly structured text and figures. The data and methods are presented in a way that makes it easy to reproduce the experiments. The statistical analysis is adequate.

Some suggestions:

1. Since you stated in 3.1. Result section that Psmb4/8 are robustly expressed in BAT, it would be interesting to directly compare the expression of Psmb4/8 in BAT. RESPONSE: We thank the reviewer for this interesting suggestion. The comparison is now included in figure S3.

Please normalize the glycerol release to protein content (Fig 2J, 4F). It would be also interesting to show whether the reduced glycerol release is due to reduced TG content and/or lipolytic activity. Therefore, you should determine the expression of lipases (e.g. Atgl, Hsl) in adipocytes.

RESPONSE: We thank the reviewer for this interesting suggestion. We have looked at the expression of lipases. Psmb4 knockdown did not alter the expression of lipases, which indicates that the reduced glycerol release is rather due to reduced TG content than due to the absence of lipases. The comparison is now included as Figure 4G. For the glycerol assays, we have normalized glycerol release to protein content. Comparing an undifferentiated (in this case the cells with silencing of Psmb4) vs differentiated cells (in this case the scrambled siRNA control cells) will result in many fundamental differences. Specifically, the lipid to protein ratio is very different, much higher in mature adipocytes, obviously. This obscures some if the differences in lipolysis when glycerol release is normalized to protein levels. Therefore, we have included a figure, in which we show the fold change. Interestingly, this way in the Psmb4 knockdown cells, it is evident that they become refractory to norepinephrine stimulation, and this is rescued when Atf3 is silencing, too.

Please define early/late transfection - on which day of differentiation was Psmb8 silenced? (Fig S2)

RESPONSE: Psmb8 was silenced on day(-1). We have now added this information.

Reviewer #3 (Significance (Required)):

This study clearly demonstrated that proteasome dysfunction via impaired PSMB4 action modulates brown adipocytes differentiation, function, and health. In this study a novel link between dysfunctional proteostasis and impaired lipid metabolism was identified.

RESPONSE: We are very thankful for these positive concluding remarks.

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

Learn more at Review Commons

---

Referee #3

Evidence, reproducibility and clarity

In this study Willemsen et al. investigated the role of proteasome subunit beta 4 and 8 (PSMB4/8) in immortalized brown (pre)adipocytes regarding adipogenesis ability, inflammation, function, and proteostasis. The group showed that Psmb4/8 are expressed in brown adipose tissue and adipocytes but that they are differently regulated. The loss of PSMB8 had no effect on brown adipose tissue/adipocyte function. In contrast, knock-down of PSMB4 altered proteostasis, which was partially compensated by NFE2L1, as well as reduced adipocyte differentiation, lipid accumulation, and beta adrenergic-stimulated glycerol release in immortalized brown adipocytes. The group further demonstrated that the effect of PSMB4 knock-down on impaired adipogenesis was mediated via Atf3 activation.

The manuscript is well-written with clearly structured text and figures. The data and methods are presented in a way that makes it easy to reproduce the experiments. The statistical analysis is adequate.

Some suggestions:

Since you stated in 3.1. Result section that Psmb4/8 are robustly expressed in BAT, it would be interesting to directly compare the expression of Psmb4/8 in BAT. Please normalize the glycerol release to protein content (Fig 2J, 4F). It would be also interesting to show whether the reduced glycerol release is due to reduced TG content and/or lipolytic activity. Therefore, you should determine the expression of lipases (e.g. Atgl, Hsl) in adipocytes.

Please define early/late transfection - on which day of differentiation was Psmb8 silenced? (Fig S2)

Significance

This study clearly demonstrated that proteasome dysfunction via impaired PSMB4 action modulates brown adipocytes differentiation, function, and health. In this study a novel link between dysfunctional proteostasis and impaired lipid metabolism was identified.

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 molecular mechanism by which proteasome mutations cause lipodystrophy in PRAAS, which is caused by proteasome dysfunctions, has not been well understood. However, it was shown that Psmb4 (β4), a component without enzymatic activity, is required for the formation and maintenance of adipocyte function. In the proteasome dysfunction state of Psmb4 deficiency, the expression of Nfe2l1 was enhanced for proteostasis, but it could not complement the adipocyte formation defect. We showed that repression of Arf3, which is associated with stress response and is markedly expressed in this situation, resulted in the recovery of inflammation and adipogenesis.

Major comments

1. The Graphical abstract seems to indicate that Loss of PSMB4 activates NFE2L1 and ATF3, resulting in the suppression of proteotoxicity and Inflammation. In the case of NFE2L1, this is correct, but in the case of ATF3, as shown in Fig. 4D, ATF3 acts in a promotive manner on Inflammation, causing misleading to the reader.
2. Fig. 2 shows the rise of Nfe2l1 and the restoration of proteasomal capacity on Day 5 (Fig. 2E). [Nfe2l1 is cleaved・・・・, and initiates the transcription of proteasome subunits, which results in restoration or heightening of proteasomal capacity (16,17).] It is known that the brown adipocyte mount an adaptive response to overcome UPS dysfunction, and the transcription of proteasome subunits was increased in this experimental system. However, there are no results showing that the transcription of proteasome subunits is actually increased in this experimental system.
3. Are Psmb4KO mice available? If yes, are there any symptoms? Is there any change in proteasome activity, etc.? In PRAAS patients, most of the lipodystrophy occurs in white adipocytes, but if Psmb4 deficiency is induced in white adipocytes, do they show the same dynamics? The first mention of heat production was made, but it was not clear how much the patient's cyclic fever symptoms were related to changes in brown adipocyte function.

Minor points

fig2; The numbering of the figure is not correct.

Fig. 2: (day 5) in the figure legend of (E) is unnecessary.

Fig. 4 The figure legend in (A) and (B) are switched.

In Fig. 4 (G), there are n = 8 and n = 6 in the figure legend, which is difficult to understand.

Significance

It has been thought that the accumulation of defective proteins caused by proteasome dysfunction stresses cell metabolism and leads to lipodystrophy, but the detailed mechanism has not been understood. In this paper, we have clarified a part of the mechanism that links the accumulation of ubiquitinated proteins caused by proteasome dysfunction to the disruption of proteostasis, inflammation and adipogenesis. The results of the study, which showed the relationship between intracellular proteostasis, inflammation and lipid metabolism, will help us understand not only PRAAS patients but also abnormal lipid metabolism, obesity, induction of inflammation, and chronic inflammation with persistent inflammation.

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:

Willemsen et al studied the contribution of PSMB4 and PSMB8 on proteasomal activity in adipocyte tissue/cells. Mutations in PSMB4/8 have been associated with metabolic diseases that lead to inflammation and proteotoxicity. They used mice and murine cell lines to assess the abundance and activity of the proteasome as well as the stress response upon depletion of psmb4 and regulatory factors. The study is interesting and could provide new insight into the role of specific proteasomal subunits of the immunoproteasome on the metabolism and associated diseases. However, there are a number of issues on the presented data that should be addressed. See below.

Major comments:

• Most of the presented data are secondary data (graphs) and the differences between the experimental conditions (e.g. siRNA) are sometimes only minor and statistical analyses lacking or only indicated with a letter (a, b,..), but its meaning not explained.
• The authors should introduce all tested marker genes e.g. the genes that were analyzed in figures 2H/I, 3D-F, 4A/D. What was the hypothesis and do they represent all or a selected set of genes of the integrated stress response?
• Figure 1A: RNA levels were analyzed yet not protein levels. Why not?
• Figure 2C: This figure is problematic. Apart from the smear on the right side, a loading control is missing. This is essential to quantify signal intensities. Moreover, on the left side, the intensities of lanes 1 and 2 are very different, yet are both controls and were used for the conclusion that proteasomal activity is reduced upon siPsmb4 (lanes 3 and 4 - that do not differ from lane 2). In addition: from which day were the data collected? This information is missing, but important as Figure 2E shows opposing proteasomal activity on day 3 and day 5.
• Figure 2D: tubulin was used as loading control, yet the signal of tubulin in lane 1 is by far weaker compared to the other lanes. How does that affect the quantification (missing) of Nfe2I1?
• Figure 3C and Fig 2E (control vs siPsmb4) contradict each other. Please clarify.
• Figure 3B: Issues with the loading control: tubulin signals are in first 3 lanes much weaker. Where is the quantification for that data set that takes the fluctuations of the tubulin signals into account?

Minor comments:

• Why did the authors not use human adipocyte cells and performed all experiments in murine cells?
• In which cell/tissue is Psmb4 expressed?
• Figure 4G: information on the different colors is missing.
• The result section appears to have been restructured as sections do not build up on each other well. This should be corrected.
• There are a number of grammatical errors or doubling of words/phrases e.g. bottom of page 1: "In addition, PRAAS patients display suffer from..." or on page 2: "Adapting proteasomal activity to the needs of the UPS..." This statement does not make sense. Maybe the authors mean "proteolytic demands"?
• Although, the UPS is a proteostasis node, the authors should avoid statements such as "We show that proteostasis and lipid metabolism are intricately linked..." Better is "UPS activity and lipid metabolism..." Or the authors should expand their analysis to protein synthesis, folding and additional clearance pathways.

Significance

This study links clinical research with basic science and if the authors address the above mentioned issues this work will provide new insight into the role of the UPS in the lipid metabolism.

target audience: clinical scientist on lipid metabolism and basic researchers on the UPS and associated pathologies

my expertise: UPS